[{"content":"The Fediverse is quite nice! For my main account, I\u0026rsquo;ve been relying on the fantasticly run chaos.social instance. I\u0026rsquo;ve played with running gotosocial and Mastodon instances in the past, but did not run them for long.\nNow had a look at Misskey, and it\u0026rsquo;s quite nice! Some thoughts, service operation wise:\nMisskey is heavier than gotosocial, it needs redis and Postgresql, nodejs. So heavy that I want to lock it away into a container, but not as heavy as Mastodon, less spinning wheels. Misskey comes with a web interface: quite different from Mastodon, I\u0026rsquo;m missing Mastodons multi-tab-view. User differences Some Misskey benefits relevant for me:\ndefault config allows not just 500 characters as Mastodon does, but 3000 Reactions are not just \u0026rsquo;likes\u0026rsquo;, but freely selectable emojis.\nOwn emojis possible. Of course, when \u0026lsquo;reactions\u0026rsquo; between Mastodon and Misskey are done, they are down to plain \u0026rsquo;likes\u0026rsquo; again. Much more formatting freedom than for Mastodon! Huge Japanese community. Docs partly Japanese. Users can mark their account as \u0026lsquo;cat\u0026rsquo;. :D You can use Misskey much more like a chat client: you see each others online status, can do chats. Playful in many places. For example: you get badges for various actions. My test instance, not sure how long I\u0026rsquo;ll run it: https://halcyon.neon.jp.net.\nMy setup notes My style for complex software like this is:\nI run an nginx on a server, that\u0026rsquo;s starting to process incoming connections and handling https Misskey is running in a podman container on the system. Service listening on a port, and communication between the nginx and Misskey are plaintext. Container setup I\u0026rsquo;m using a plain Debian Trixie in the container, and manage services with systemd inside the container. The below setup instructions are not automated down to the last piece: that\u0026rsquo;s making future adaptions easier when the setup instructions change.\nThe steps are quite basic.. but expecially for the config files I have created a repo. Starting with building the container image:\ngit clone https://codeberg.org/globalc/smallhelpers cd smallhelpers/misskey-podman-setup # build locally podman build -t misskey build-trixie/ # run locally podman run -p 127.0.0.1:8085:8085 --name misskey -d \\ --security-opt seccomp=unconfined --hostname misskey \\ localhost/misskey /lib/systemd/systemd # run a shell podman exec -it misskey bash With that, we should end up with the image created, an instance started, and a shell open into the instance.\nSetting up Misskey This is loosely following the manual. We start installing Nodejs:\n# install nodejs # https://github.com/nodesource/distributions#installation-instructions mkdir -p /etc/apt/keyrings curl -fsSL https://deb.nodesource.com/gpgkey/nodesource-repo.gpg.key | \\ gpg --dearmor -o /etc/apt/keyrings/nodesource.gpg echo \u0026#34;deb [signed-by=/etc/apt/keyrings/nodesource.gpg] https://deb.nodesource.com/node_24.x nodistro main\u0026#34; \u0026gt; \\ /etc/apt/sources.list.d/nodesource.list apt update apt install -y nodejs vim Now PostgreSQL and redis, preparing database. Customize the postgres passowrd.\n# install further packages apt install -y \\ ffmpeg redis-server redis-tools postgresql systemctl start redis postgresql # Don\u0026#39;t cut\u0026#39;n\u0026#39;paste the following in bulk, not working as new # console is entered. Paste this line by line. sudo -u postgres psql CREATE USER misskey CREATEDB; alter user misskey with password \u0026#39;misskeypqpassword\u0026#39;; CREATE DATABASE misskey OWNER = misskey; \\q Creating app user, installing the code:\n# First actions as misskey user adduser --disabled-password --disabled-login misskey usermod -s /bin/bash misskey su - misskey npx get-pnpm source ~/.bashrc # Setup Misskey itself git clone --recursive https://github.com/misskey-dev/misskey.git cd misskey git checkout master git submodule update --init NODE_ENV=production pnpm install --frozen-lockfile cp .config/example.yml .config/default.yml vim .config/default.yml The last step started the configuration. Go trough the pieces carefully, in my case I customize this:\nmisskey@misskey:~/misskey$ diff .config/example.yml .config/default.yml 60c60 \u0026lt; # publishTarballInsteadOfProvideRepositoryUrl: true --- \u0026gt; publishTarballInsteadOfProvideRepositoryUrl: true 74a75 \u0026gt; setupPassword: insertpostgresspassword 80c81 \u0026lt; url: https://example.tld/ --- \u0026gt; url: https://halcyon.neon.jp.net/ 102c103 \u0026lt; port: 3000 --- \u0026gt; port: 8085 167,168c168,169 \u0026lt; user: example-misskey-user \u0026lt; pass: example-misskey-pass --- \u0026gt; user: misskey-user \u0026gt; pass: misskey-pass With that, we can build Misskey and start it for the first time:\nNODE_ENV=production pnpm run build pnpm run init NODE_ENV=production pnpm run start It\u0026rsquo;s now running just in the terminal, commands below configure it as proper service. To access, we need to tell nginx on the host about Misskey. In my config, nginx runs on my server and does the https connections. Misskey runs in a container, connection nginx\u0026lt;-\u0026gt;Misskey is not encrypted. So now deploying the additional nginx config on the host, not in the container.\n# ( https://blog.vyvojari.dev/mastodon-behind-a-reverse-proxy-without-ssl-https ) sudo cp misskey-nginx.conf /etc/nginx/sites-available/misskey sudo ln -s /etc/nginx/sites-available/misskey /etc/nginx/sites-enabled/misskey systemctl daemon-reload systemctl restart nginx You can now access the Misskey web interface and go from there. For proper operation, inside the container, to start Misskey as systemd service:\ncp misskey.service /etc/systemd/system/ systemctl daemon-reload systemctl start misskey systemctl enable redis-server postgresql misskey Corrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-08-22T19:32:27+09:00","image":"/2026/08/installing-misskey-fediverse-instance/headline.png","permalink":"/2026/08/installing-misskey-fediverse-instance/","title":"Installing Misskey Fediverse Instance"},{"content":"I got free time to spend with HAM gear in the middle of Germany, on a house with 3 floors and a garden behind the house. After doing some connections with FT8, a digital mode, I started to wonder: which of the 2 antennas is best? Huge disclaimer: \u0026ldquo;best\u0026rdquo; is relative, I\u0026rsquo;m trying to optimize for \u0026ldquo;connections to far away stations\u0026rdquo;, and did just test for the 20m band (~15Mhz).\nWhat to test, and how? The top three ways of communicating via HAM radio equipment are right now\nMorse Code Voice transmission Data transmission I\u0026rsquo;m focusing on the third one right now, so I connect a computer or mobile phone to the radio transmitter and have it communicate via digital protocols, a bit like modems over phone lines in the past. I\u0026rsquo;m using the FT8 protocol. Effectively, one the software I see short messages from others, and send out messages myself.\nFor the comparisons I use WSPR, pronounced \u0026ldquo;whisper\u0026rdquo;, an acronym for \u0026lsquo;Weak Signal Propagation Reporter\u0026rsquo;. It is a protocol meant to be sent out with low power, and to be usable even under bad conditions.\nAll over the world, HAMs are running setups with configs which listen to WSPR signals from others, and over the internet the listen reports are then gathered.\nWe could use this to compare receive as well as sending capabilities of our setups, but as listening is rather easy, I\u0026rsquo;m focusing on sending out WSPR for 30min with each configuration, and then compare how I\u0026rsquo;m getting received. For this comparison, it\u0026rsquo;s important that the radio wave propagation conditions are stable over all test setups.\nThe Antennas/scenarios to compare 1: Wire antenna, short counterpoise EFHW (End Feed Half Wave) antenna with a long wire and a 1:49 balun, link. EFHW means that the antenna connector (= feed point) is attached to the antenna at one end of the cable, and the cable length is half of the wave length to be transmitted. Additionally, a counterpoise needs to be connected to the antenna. For this scenario, the antenna feed point is at the roof of the house, and the antenna wire is stretched downwards. A string at the end of the wire is tied on in the garden. Not in the graphic: a 2m flexible cable connected to the box/balun, as counterpoise. The wire antenna does not directly allow tuning for a good SWR, for my tests here I had an SWR of 1.5 on the 20m band.\n2: Wire antenna, long counterpoise This setup is identical to the first setup, just a 4m counterpoise is getting used, instead of the 2m one. It\u0026rsquo;s located on the right side of the roof, while the antenna wire is up on the other side.\n3: Comet antenna, long counterpoise This setup is using the Comet HFJ-350 antenna, attached at the top of the roof. The Comet is a telescope stick antenna, up to 1.2m long. This is a EFQW antenna, End feed Quarter Wave. So quite short, but also covering various bands. The bottom of the antenna contains a coil, electricly \u0026rsquo;extending\u0026rsquo; the antenna for these bands. Being just quarter wave gives the antenna disadvantage over the half-wave antenna, but then it\u0026rsquo;s setup higher. The Comet needs tuning for a good SWR, that\u0026rsquo;s done in selecting the best coil connect point, and then tuning the telescope extension. For the tests here I tuned to SWR 1.1 for the 20m band.\n4: Comet antenna below the roof This setup is like #3, but with the Comet below the roof. Counterpoise again the wide 4m cable which I had obtained when purchasing a 6m telescope antenna. SWR tuned to 2.1 for this setup, I did not get it lower.\n5: Comet antenna on middle floor of the house Seeing also reception for setup #4, I got curious how the Comet would perform from inside the house, at the 2nd floor. This time with just short 2m counterpoise cable, tuned to SWR 1.6.\nTests For the tests, I had the wsjtx software send out WSPR for 30min, with 20dbm/100mW. That\u0026rsquo;s a fraction of the normal 3W/4W which the QMX+ transceiver is outputting, but the 100mW are a typical value for measuring antennas. The propagation conditions of the 20m band were for all 5 tests \u0026rsquo;excellent\u0026rsquo;, according to the HF prop lab page which I\u0026rsquo;m using.\nResults Scenario received # TX # rec/run 1: Wire antenna, short counterp. 149 5 30 2: Wire antenna, long counterp. 70 4 18 3: Comet on roof, long counterp. 193 5 39 4: Comet below roof, long counterp. 137 4 34 5: Comet 2nd floor, short counterp. 38 5 8 Column \u0026lsquo;received #\u0026rsquo; has the number of receive-reports from other stations. I noticed that within the 30min, I had sometimes 4 and sometimes 5 transmissions, that\u0026rsquo;s reflected in column \u0026ldquo;TX #\u0026rdquo;. The last column has then the number of receive reports per single transmission - so that\u0026rsquo;s the most interesting value.\nLet\u0026rsquo;s not read to much out of this: this is just a back-of-the-envelope comparison. For example, if by coincidence for one of the test runs just a few receiving WSPR reporters were available, also the number of reports would be lower..\nTake aways Measuring and comparing require a considerable amount of effort. Even with WSPR and a network of HAMs constantly receiving WSPR and offering the data - that helps greatly! I can\u0026rsquo;t make sense from getting for scenario 2 fewer receive reports than for scenario 1. I expected the hand soldered/crimped short counterpoise would be worse than the long one. Needs more research. The Comet performs better than the wire antenna. As expected, the Comet below the roof is worse than ontop of the roof, but not much: below the roof got 87% of the receive reports of over the roof. I might consider a stationary deployment, but I also need another radio transceiver then. Even with the Comet deployed inside the house we got receive reports! Corrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-08-11T02:37:36+09:00","image":"/2026/08/comparing-antennas-with-wspr/05-wsprglobe.com-scenario3.png","permalink":"/2026/08/comparing-antennas-with-wspr/","title":"Comparing Antennas with WSPR"},{"content":"POTA is \u0026lsquo;Parks On The Air\u0026rsquo;. You go to a park, deploy antenna and radio transceiver, and to \u0026lsquo;activate the park\u0026rsquo; you have then exchanges with at least 10 folks over the air.\nSOTA is \u0026lsquo;Summits On The Air\u0026rsquo;: you go to a mountain summit, and for an \u0026lsquo;activation of the summit\u0026rsquo; you need at least 4 connections over the air. I\u0026rsquo;m in Germany right now, and this is about my first successful SOTA activation. Fits perfectly: I have call signs in both Japan and Germany. My first POTA activation was in Japan, first SOTA in Germany - covering both countries where I feel \u0026lsquo;at home\u0026rsquo;.\nThe tech parts are almost unchanged since my descriptions in Starting POTA and optimizing equipment. Only important change: I got a new power bank. To power the QMX+ transceiver, it needs to offer 12V DC via power delivery. After some research, I found a fitting power bank: 12000mAh, weight 243g. Much lighter than the power bank I have in Japan, and still more power than I need for a single activation.\nPlanning Just locations in the official databases of POTA/SOTA can be used for activations. From the map on sotl.as I saw that summit Rondell is nearby, about 20km by bike. In a group of 4 we went off, route planning done with OsmAnd on my Pixel9 running GrapheneOS. The route was over quite some rough trails, at one point we had to lift our bikes over fallen down trees. I was the only one on a bike without electrical support, but this was totally doable - the \u0026lsquo;Rondell\u0026rsquo; is just 522m over sea level.\nArriving at the Rondell Funny enough - even though this is classified as \u0026lsquo;summit\u0026rsquo;, it felt much more like a park! Benches with tables, very nice. Reminded me of K4SWL’s POTA sessions, like this one.\nSetting up At first, we were sitting down at these benches, setting up the equipment. My loop is like this:\ndeciding for a band, i.e. 20m or 30m (the QMX+ does 6m-160m, the Comet antenna without extra coil 10m-40m) wiring all up, logging into the QMX+ transceiver via USB-C cable, and measuring the SWR of the setup adjusting length of the antenna for better SWR, measure again - until SWR is reasonable The odd thing that day: when I had adjusted the setup to SWR of i.e. 1.3 or 1.5, and then started software \u0026lsquo;wsjtx\u0026rsquo;, the QMX+ would show an error indicating a very high SWR. Then measuring the SWR again showed it indeed very high - so some conditions I could not identify had changed it in the meantime. I have a cheat-sheet for tuning the Comet HFJ-350.\nI moved then away from the benches, to the more open area where I could not sit down while operating. The antenna was no longer surrounded by trees, and ontop of the \u0026lsquo;rondell\u0026rsquo; monument itself - this helped. Using an antenna tuner could change this, but would also mean more weight to carry around.\n2 of the others left early, heading towards a gas station in Keula. Dad and I stayed another 30min, completing 5 contacts.\nWay back We caught the others at the gas station, and headed back towards Muehlhausen. Towards the end, I had a flat tyre - my dad headed home, got his car and fetched me from a parking lot.\nEquipment Equipment this time:\npart weight Comet antenna, no extra coil 180g Macbook air M1/Asahi 1280g coax feedline antenna \u0026lt;-\u0026gt; radio 53g 12000mAh power bank 243g QMX+ radio, power cable + adapter 739g usb-c data cable 27g counterpoise cable 82g tripod + adapter 140g Corrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-08-06T15:03:01+09:00","image":"/2026/08/first-SOTA-activation/20260805_114646_870-benches-small.jpg","permalink":"/2026/08/first-SOTA-activation/","title":"First SOTA Activation"},{"content":"In July, I had my first POTA activation. POTA is \u0026lsquo;Parks On The Air\u0026rsquo;, an activity in the field of HAM radio. There are 2 roles:\nThe Activator, someone going into a park, setting up radio equipment and making connections over the air. With at least 10 connections, it\u0026rsquo;s a \u0026lsquo;park activation\u0026rsquo;. The Hunter, folks doing connections to the Activator. Both roles can be combined when \u0026lsquo;park to park connections\u0026rsquo; are done. There are also SOTA (Summits On The Air) and much more. A map with the parks for POTA is here, and one for SOTA is here.\nIn this article, I\u0026rsquo;ll explain the POTA basics, what I find interesting about it, and the basics on the involved gear like antenna. One needs some level of understanding to get a first successful activation - but then ontop many rabbit holes are waiting to optimize.\nWhat\u0026rsquo;s in it? For me, it\u0026rsquo;s these:\nThe incentive to go to the fresh air, and make radio connections. My flat in Tokyo also has bad conditions for radio connections, so going to parks also allows to get better conditions. Plain badge hunting: site pota.app for having statistics, allowing competing with others. Optimizing equipment over time. License requirements I think HAM radio license is required, already the smallest license of Germany or Japan are sufficient. In Germany, the N-class allows operation on 70cm/2m/10m wavelength area, and in Japan the 4th level class can be aqquired after 2 days of training and passing the exam. Japan 4th level allows to use almost all frequencies. Along with a license, one gets a worldwide unique callsign.\nLet\u0026rsquo;s talk about the equipment.\nAntenna When you do radio communication from home, you rather lenghtly plan the best antenna, set it up, tune it, measure. For POTA sessions, you optimize more for easy deployment, and weight. I\u0026rsquo;m using 2 antennas:\nComet HFJ-350: a 1.2m long telescope stick. The technical term is \u0026rsquo;end fed quarter wave\u0026rsquo;: so the part radiating the waves is tuned to be one quarter of the wavelength with which we want to operate. A 5.6m long telescope stick antenna, an \u0026rsquo;end fed half wave\u0026rsquo;. This one is longer, more heavy, not good for taking it to the mountain. As a half wave type, it has better radiation pattern than the other antenna. Both antennas include a coil or a balun, required to match our feedline (a coax cable) to the antenna.\nFor radio connections, we choose a frequency to operate on. The frequency is directly related to the wavelength one is operating on: i.e. frequency 50Mhz means 6m wavelength. Frequency of 30Mhz means 10m wavelength. The wavelengths are also called bands, so \u0026rsquo;the 6m band\u0026rsquo; or \u0026rsquo;the 10m band\u0026rsquo;. Both of my antennas can be used on multiple bands, for example I have used the Comet on the 10m, 15m, 20m and 30m bands.\nOf course: also constructing own antennas is possible and not to hard. For example, a dipole antenna can be constructed from just 2 pieces of wire, each one quarter wavelength long, and a balun. Also for constructing: you can use a \u0026lsquo;recipe\u0026rsquo; off the shelf for an antenna for a cerain band, build and then use it. Or start to experiment with own constructions: radio tech is around since more than 100 years, many docs are around to get started.\nSome antenna related things I did not touch:\nMode of operation, and my Radio transceiver We did not yet talk about our \u0026lsquo;way of talking\u0026rsquo; over the air: POTA is quite flexible there. Depending on our mode of transmission, we can then choose the transceiver.\nVoice: We can bring in a microphone and speaker/headset, and then with a radio transceiver make our connections.\nQuite simple - but for this the connection to the remote station needs to be quite good: the remote signal needs to be much stronger than the noise level we hear, so the signal-to-noise level needs to be good. Morce/CW: If the operator can operate morse code, then they can bring in a morse key/paddle, and the radio transceiver can be smaller than for \u0026lsquo;voice\u0026rsquo;. Also, morse can be operated over bad connections over which \u0026lsquo;voice\u0026rsquo; would not be possible. Digital modes: Remember modems which we connected to the phone and the computer? Digital modes do this over the air. Modes like FT8 can be used, and they can operate under conditions when even Morse code is no longer possible. Downside: we need something to compute/to \u0026lsquo;speak\u0026rsquo; the digital mode. Of these three, I\u0026rsquo;m just doing digimodes right now. My radio transceiver is a QRP Labs QMX+: a device which can operate from 6m-160m on various bands. This means I also need some device to \u0026lsquo;speak\u0026rsquo; the digi mode..\nMy Digital Mode devices research I started out with a Thinkpad L480 in my flat, running Debian Linux and the WSJT software. Research of minimal digimode devices brought me to.\nThe Cardputer M5 is a quite small device, with ESP32 cpu and 512kb RAM, no MMU. It has a USB-C connector, with existing software it can make the QMX+ tranceiver speak \u0026lsquo;FT8\u0026rsquo;. I took this to 2 mountains for testing, and had one successful connection with a remote station - but decided this is really to minimal for me. Can\u0026rsquo;t select where to send, barely get an idea of how occupied the frequencies are.\nNext option: Android mobile phone. I tried 2 softwares. Actually made one or 2 connections. The mobile phone has already a bigger screen than the Cardputer, allows more control. But still does not allow to select the sending frequency.\nThe option I settled with: Apple Air M1 with Asahi Linux. This is a full blown laptop - but extremely small. Lightweight, long lasting battery, screen which can be tuned up and is readable even in sunlight. As for operating system: Linux is simply what I\u0026rsquo;m familiar with, and the open source concept also nicely goes along with the HAM spirit of sharing. Used Apple M1 Air are now sold for 5万円/350$, that\u0026rsquo;s very reasonable.\nMy docs for setting up with disc encryption.\nThe other equipment A feedline: coax cable to connect radio transceiver and antenna. Counterpoise (Ger: Gegengewich): our antenna is a kind of \u0026lsquo;unusually shaped capacitor\u0026rsquo;, these electronical devices with 2 plates which get charged. Our wire or telescope metal are one half, and ideally we should connect a \u0026lsquo;counter part\u0026rsquo; to span up the field, the \u0026lsquo;counterpoise\u0026rsquo;. This is usually one or multiple wires connected to the foot of the antenna. Mine is 3m long, single cable. Battery: my transceiver asks for 12V DC. I got a mobile battery pack offering \u0026lsquo;power delivery\u0026rsquo; over USB-C connector. The transceiver has a barrel shaped connector, so I got a small adapter which is connected to the USB-C cable from the battery, is requesting 12V via Power Delivery, and supplies the transceiver. USB-C cable for transceiver \u0026lt;-\u0026gt; computer connection The physical setup routine We\u0026rsquo;ve gone through the involved pieces, let\u0026rsquo;s look at a setup. I\u0026rsquo;m filling my equipment into a bag, attach it to the bike, and go to a park. We boot the computer, deploy antenna with counterpoise, connect antenna to the transceiver, power to the transceiver, and connect transceiver to the computer.\nNext is to decide which band to operate on. Even if we have a fantastic antenna, the conditions for the propagation of radio waves are changing - like weather forecast. For a point in time, the conditions are also different for the various bands - sites like this one or this one give us hints. So let\u0026rsquo;s say we decide for the 20m band.\nNext we should confirm the antenna is tuned to the band we want to operate on. My transceiver can only send out highfrequency with 5Watt electrical power, that\u0026rsquo;s called \u0026lsquo;QRP\u0026rsquo;, low ressource operation. Even with that: with a badly tuned antenna, that power will not be radiated from the antenna but sent back to the transceiver. My transceiver has just a small display, but allows me to login via serial console (\u0026lsquo;minicom -D /dev/ttyACM0\u0026rsquo; from Linux), and can then go over all bands and tell me which part of the sent out power is coming back. On that base, I confirm is the counterpoise is connected correctly, can adjust the telescope extents to tune the antenna. Following is the initial output, and then the tuning in making the antenna shorter. Cute ASCII-art!\nWhat\u0026rsquo;s the radio doing for this graph? It\u0026rsquo;s sending \u0026lsquo;a bit\u0026rsquo; of high frequency towards the antenna, on various frequencies, that\u0026rsquo;s the X-axis. It\u0026rsquo;s then measuring how much is coming back. The vertical line is the frequency of our 20m band, so for tuning we need to get the value down for exactly that frequency area.\nWhile SWR of 2 is not ideal, I decided to go with it for that day. The QMX+ has also an internal protection mechanism against high SWR. Another option here would be to use an \u0026lsquo;antenna tuner\u0026rsquo;, or a radio transceiver with internal antenna tuner, like various Elecraft radios.\nSoftware Operation Time to use the software, for example WSJTX.\nWhat do we see here? Left side screenshot: we see a frequency area. The software is looking at a whole area of frequencies, all in parallel. As if you could tune a classic radio to many stations at once. From the left to the right, the frequency is increasing. At the top, we have the mose recent data, and that\u0026rsquo;s then moving down. The amplitude at the bottom is also telling us in which area we have right now activity. The red/yellow pieces are signals, which the software does decode for us.\nOn the right side screen shot, we on the left side the messages which we received and decoded. The columns have time stamp, strength of the received signal, frequency, who was sending to whom, and the actual message. At the bottom of the right side we see what I was initially sending out/broadcasting: \u0026ldquo;CQ POTA JL1AYH JL1AYH PM95\u0026rdquo;. So \u0026ldquo;call to everybody, I\u0026rsquo;m doing Parks on the air. Callsign JL1AYH speaking, from the locator PM95\u0026rdquo;. It\u0026rsquo;s a Maidenhead Locator.\nThis was a generic call \u0026rsquo;to all who could hear me\u0026rsquo;, to which I got replies from specific stations (hunters), which we see in the right window. A minimal exchange contains a signal report. I had in parallel announced my frequency/my activation on pota.app, and had actually stations lining up to do a complete connection exchange with me.\nOne further interesting thing about digi modes: they are easy to monitor. After some hours of operation, you can go to an internet site and check where all over the world your signal has been received!\nPictures from the first POTA activation Future optimizations I hope to get a first SOTA activation together in the near future, and do more POTA.\nMy battery is not optimal: I got a 20Ah piece - which turned out as having way to much capacity for what my transceiver consumes, so it\u0026rsquo;s more heavy than it needs to be. I see no good way to sell it though.. Morse code: I\u0026rsquo;m learning since 1.5 years, progressing slowly. Aqquiring that skill, using it for POTA/SOTA would be great. Antenna: if I manage to found a family and live in a house, I can finally setup a proper antenna. For the portable ones, also improvements are possible. For all of the topics in this article, we barely scratched the top of the iceberg. Pieces of what I learned in the last years are crammed into the article.. rabbit holes to dive into are behind all of these topics.\nLet me close with a recommendation to watch one of Thomas/K4SWL\u0026rsquo;s POTA sessions, for example this one, I love the style. Park activations here in Tokyo are a bit different though, more people are around. :)\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-07-11T19:07:01+09:00","image":"/2026/07/starting-POTA-and-optimizing-equipment/20260710_190548_setup_computer_small.jpg","permalink":"/2026/07/starting-POTA-and-optimizing-equipment/","title":"Starting POTA and optimizing equipment"},{"content":"I recently got myself an Apple MacBook Air M1, these are down to 5万円/350$ here in Japan. The CPU efficient, the screen is good - I plan to use the device as portable desktop with Fedora Asahi Remix. Also for HAM activities: to go to parks/summits, setting up an antenna and have the macbook drive my QMX+ radio transceiver to do FT8 transmissions - the macbook is light and small, the screen is good and can be turned up to be still readable in sunlight.\nMy last Asahi system was the MacBook pro M2, it was my workplace system, now returned to the company.\nFor a portable system, we want disk encyption nowadays. In 2023, I was following this guide. Things have changed massively, so this here is a guide as of 2026.\nKudos to\nMarcan and the Asahi team Leif Liddy for asahi-fedora-usb @copyrights@chaos.social for article input Overall procedure These steps require the already setup Asahi installation. We will then\nCreate a USB drive with Asahi Linux Boot from that LUKS Encrypt the Asahi file system Update the Asahi version on the USB drive to match the real installation And then configure grub and rebuild initramfs Step 1: Create the live media and boot from it As of Fedora Asahi Remix 44, this works quite nicely with leifliddy\u0026rsquo;s asahi-fedora-usb.\n# Internet connectivity, if not yet setup: # List wifi networks, then connect nmcli dev wifi nmcli dev wifi connect \u0026lt;your-ssid\u0026gt; password \u0026lt;password\u0026gt; # Install dependencies dnf install arch-install-scripts bubblewrap gdisk \\ pandoc rsync systemd-container git mkosi # checkout the code git clone https://github.com/leifliddy/asahi-fedora-usb.git cd asahi-fedora-usb/ # Now attach the usb media you want to boot from. For me, a USB-C # attached enclosure with an NVMe inside. For some folks, USB-2 # devices did not work for booting, but USB-3 devices. # Confirm where the USB device appears: cat /proc/partitions # For me it\u0026#39;s at /dev/sda. # Build and install a minimal Asahi Fedora image to a USB drive: ./build.sh -d /dev/sda # Until 2026/5/30, this did fail because Fedora 44 comes with # mkosi 26, which the script did not support. Manually changing # mkosi_supported_version to \u0026#34;26\u0026#34; fixes this. Step 2: External Booting and reencoding Now reboot the system.\nAt the UEFI prompt, after \u0026lsquo;u-boot\u0026rsquo; is written on the screen, interrupt in pressing the ESC-key. You get the \u0026ldquo;=\u0026gt;\u0026rdquo; prompt. Enter \u0026ldquo;eficonfig\u0026rdquo;, change boot order so usb is first. Then use \u0026lsquo;run bootcmd\u0026rsquo; on the efi prompt to boot. grub from the external media will run. For me, the Fedora boot procedure was stuck, root partition was not found. Modifying the grub parameters to \u0026lsquo;root=/dev/sda3\u0026rsquo; did not help. What helps: once the boot is stuck, unplug the usb-c connected NVMe drive, then plug it in again. Maybe try a different USB-C port. For me, that did the trick. Then, login with root/fedora Encrypt the Asahi installation: # Using lsblk -f /dev/nvme0n1 identify the partition used for the root filesystem. # This should present as a btrfs file system with the label fedora_asahi. In the # example below /dev/nvme0n1p6 is the Fedora root filesystem that will be encrypted in-place: [root@fedora ~]# lsblk -f /dev/nvme0n1 NAME FSTYPE FSVER LABEL UUID FSAVAIL FSUSE% MOUNTPOINTS nvme0n1 ├─nvme0n1p1 apfs 4ccf344c-1842-4ed2-98f7-d34a509f5a88 ├─nvme0n1p2 apfs dbb4789e-c51d-46bf-8332-90a43b4e4fa7 ├─nvme0n1p3 apfs b98ec259-629b-4aee-9f26-02c5098abcee ├─nvme0n1p4 vfat FAT32 EFI-FEDORA B01E-2641 419.8M 16% /run/.system-efi ├─nvme0n1p5 ext4 1.0 fedora_boot 5b094e58-d15f-4be2-85ff-147859c7b118 ├─nvme0n1p6 btrfs fedora_asahi dd08a2bf-ae63-44e1-881d-fbb8928af4fb └─nvme0n1p7 apfs b465c845-eaef-4bcb-aac9-865c42260844 Now resize the btrfs file system, and onthefly encrypt the existing Asahi installation. Despite the warning, the procedure is expected to keep your data usable/not delete it.\nmount /dev/nvme0n1p6 /mnt btrfs filesystem resize -32M /mnt umount /dev/nvme0n1p6 # LUKS encrypt the root filesystem partition in-place. cryptsetup reencrypt --encrypt --reduce-device-size 32M /dev/nvme0n1p6 Step 3: Ensure Fedora 44 is on the external media From 2026/5/31 on, asahi-fedora-usb is deploying Fedora 44 on the USB media, and this step can be skipped.\nIf you use an older installation on the USB media, you will need to update it to match the version of your internal disk Asahi installation. Luckily, this works with the default \u0026lsquo;Fedora upgrade steps\u0026rsquo;. Ensure you are connected to the internet, then:\ndnf upgrade --refresh dnf system-upgrade download --releasever=44 Then reboot, do the unplug/plug dance again to have the media recognized, and the installation on the usb media gets upgraded to Fedora 44.\nStep 4: Preparing Asahi Now we need to prepare the Asahi installation to deal with the encrypted disk. We are now booted into Fedora 44 from the usb media:\ncryptsetup open /dev/nvme0n1p6 fedora-root cryptsetup status fedora-root mount -o subvol=root /dev/mapper/fedora-root /mnt mount -o subvol=home /dev/mapper/fedora-root /mnt/home mount /dev/nvme0n1p5 /mnt/boot mount /dev/nvme0n1p4 /mnt/boot/efi # Store the LUKS UUID in variable for later use export LUKS_UUID=$(cryptsetup luksUUID /dev/nvme0n1p6 | tee /dev/stderr) mount -t sysfs sysfs /mnt/sys mount -t proc proc /mnt/proc mount -t devtmpfs devtmpfs /mnt/dev chroot /mnt /bin/bash touch /etc/crypttab chmod 0600 /etc/crypttab echo \u0026#34;fedora-root UUID=${LUKS_UUID} none discard\u0026#34; \u0026gt;\u0026gt; /etc/crypttab cat /etc/crypttab # Now append rd.luks.uuid=LUKS_UUID to variable # GRUB_CMDLINE_LINUX_DEFAULT. # Instead of using Perl, you might want to just # echo $LUKS_UUID \u0026gt;\u0026gt;/etc/default/grub # and then use an editor. Pay attention to using two # \u0026gt;\u0026gt; signs, to append to the file. perl -i -pe \u0026#39;s/(GRUB_CMDLINE_LINUX_DEFAULT)=\u0026#34;(.*)\u0026#34;/$1=\u0026#34;$2 rd.luks.uuid=\u0026#39;\u0026#34;${LUKS_UUID}\u0026#34;\u0026#39;\u0026#34;/\u0026#39; /etc/default/grub cat /etc/default/grub # rebuild initramfs grub2-mkconfig -o /boot/grub2/grub.cfg dracut -f # If the kernel versions of your internal and external installation # do not match, you might need something like this for rebuilding # the initramfs: dracut -f /boot/initramfs-6.19.140.asahi.fc40.aarch64.img \\ 6.19.140.asahi.fc40.aarch64 # Then exit the chroot, reboot exit reboot Now, after rebooting the system, we can change the uefi-order back and boot our encrypted Asahi.\nStep 5: Confirm TRIM works At this point, we should ensure proper handling of TRIM at all layers so that the storage is handled in the best way. As we added \u0026lsquo;discard\u0026rsquo; to /etc/crypttab, we were setting up the required pieces. Let\u0026rsquo;s confirm:\nroot@空:~# fstrim / root@空:~# Without TRIM being enabled on all layers, for example without the \u0026lsquo;discard\u0026rsquo; in /etc/crypttab we would get:\nroot@空:~# fstrim / fstrim: /: the discard operation is not supported I\u0026rsquo;ve done a first setup and given KDE a try again, but then I did go back to do a minimal Asahi installation, and installing sway. Details are on the wiki. Now on to installing Japanese keyboard input systems, syncthing and HAM software.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-05-30T11:51:10+09:00","image":"/2026/05/fedora-asahi-remix-with-LUKS-encryption-in-2026/sddm-login-air.jpg","permalink":"/2026/05/fedora-asahi-remix-with-LUKS-encryption-in-2026/","title":"Fedora Asahi Remix With LUKS Encryption in 2026"},{"content":"This is my attempt of describing where Linux power consumption measurement is, the handover of metrics to virtual guests, and the current riddle I see for attributing consumption to processes.\nOverall system consumption So this works reasonably well: Performance Co-Pilot with pmda-denki are work reasonably well for measuring, storing and visualizing the power consumption of whole systems. Reasonably well for various investigations which I\u0026rsquo;ve shared here on the blog.\nThe next challenge is attribution of consumption to single processes. Right now we can say \u0026ldquo;this system is running MariaDB, is constabtly consuming 100Watt power\u0026rdquo;, so then neglecting operating system processes we could say \u0026ldquo;we can attribute these 100Watt to MariaDB\u0026rdquo;. Nice and tidy.\nIf we have a data center with always just a single big application on each system, and no virtual guests, we can also sum up consumption of all of the systems running MariaDB, the systems running PostgreSQL and so on, and have a reasonable overview. Link to pmda-denki and power attribution docs.\nWhy do we want attribution? For a number of reasons, examples:\nYou are on a plane with a charged laptop battery, no option of recharging. You want to watch recordings from the last Chaos Communication congress, and want to compare whether VLC, mplayer or Firefox uses least energy for playing the recordings. You are a company and evaluate 2 softwares which both full fill your requirements: which one uses least energy? As a company, you want to monitor consumption of applications over time. Is a software consuming more after an update? As a developer: you are not only interested in the performance impact of your last commit, but also impact on consumption. pmda-denki and PCP can help with much of this already, but there is much room for improvement.\nProcesses and Virtual Guests In the real world, your webserver is running Nginx, PHP and PostgreSQL - all on the same system. If we want to notice future PHP updates consuming more energy, we need attribution by application, or process.\nWhat if all of my applications run in KVM guests? I also want to do proper attribution of the applications inside the guests. On the hypervisor, if I manage to do an overall attribution of consumption to the guest (\u0026ldquo;the webserver KVM guest uses 50Watt\u0026rdquo;), then I can get that metric (50 Watt) to code inside the guest, and there attribute to single processes.\nFor all of this to work, we are right now facing 2 obstacles.\nThe attribution to single processes is working by some means (see obstacle two below), and we can get a top-like overview of how much the processes are using. Not bad.\nObstacle one: Communication with the Guest Some colleagues and me were pondering about this for a while. As a proof of concept, we can overcome this through the power of PCP, with the following components:\nOn the hypervisor we run the code for attribution to processes, so we get metrics for the current consumption of all processes. Example: \u0026ldquo;the process of KVM-guest1 uses 50 Watt\u0026rdquo;. This gets available in PCP\u0026rsquo;s pmcd, a daemon. Simple code can query it over TCP/IP, and it gives us the metrics: KVM-guest1 uses 50 Watt, KVM-guest2 uses 100 Watt, and so on. In the guest, we can setup networking, fetch the consumption metrics which was computed on the hypervisor, and pick out our own matric: \u0026ldquo;I see, I\u0026rsquo;m a guest and use 50 Watt now\u0026rdquo; With that, the guest can then by itself look at it\u0026rsquo;s processes and attribute the 50 Watt to single applications/processes. That works and is nice, but not really a longterm solution. It needs networking to be up, and code in the guest to query the hypervisors pmcd.\nFirst approach to solve the handover into the guest: RAPL. RAPL is Intels interface to share energy consumption data with operating systems, this is the most heavily used source to query overall systems consumption. pmda-denki and other software can use this data. Code got written to have KVM-qemu offer RAPL metrics to the guest, just like real hardware. The code was discussed on qemu-devel@, but ended up to complex to be taken into distros like RHEL9 for example. Also: it\u0026rsquo;s x86 only. Ideally we would want something also for ARM, RISC-V and so on.\nLast year, I aqquired a Power-Z device: it\u0026rsquo;s a 100$ hardware. It has\na USB-C input, i.e. coming from a power supply a USB-C output, i.e. going to a computer whose consumption we want to measure a USB-C data connector: over this, it offers the current power troughput via an API. The API is available over USB-serial, with just a small additional Linux kernel module, which is part of the upstream kernel. So.. why don\u0026rsquo;t we make the metrics available via this interface?\nQemu-KVM seems to already implement usb-host, so we could get an easy channel from the hypervisor into the guest? Seems like Fedora has the required support, \u0026lsquo;qemu-system-x86_64 -device usb-serial,help\u0026rsquo; looks good. But how to get it into virsh config? That\u0026rsquo;s one TODO.\nObstacle two: The current attribution twist Well, the other current obstacle is an oddity regarding the attribution. My current take on the consumption-attribution to processes is here, and the docs are here.\nLet\u0026rsquo;s start with: the whole attribution is full of compromises. If a single system has all cores (the main power consumer) busy with MariaDB processes, we can attribute the overall systems consumption to MariaDB. The current approach is just considering CPU load, we are using a metric which is available from the PCP framework:\n$ pminfo -f proc.psinfo.utime proc.psinfo.utime inst [1 or \u0026#34;000001 /usr/lib/systemd/systemd\u0026#34;] value 3810 inst [2 or \u0026#34;000002 (kthreadd)\u0026#34;] value 0 inst [3 or \u0026#34;000003 (pool_workqueue_release)\u0026#34;] value 0 [..] inst [9169 or \u0026#34;009169 /home/chris/[..]/steamwebhelper\u0026#34;] value 40 [..] We see here references to the processes: the PID and the command, and then a counter. Looking at this output again after 10sec, we see the counter increased - if the process was running on CPU cores. Under the hood, this metric comes from Linux\u0026rsquo; /proc/PID/stat : it\u0026rsquo;s a counter value in units of milliseconds.\nFor the consumption attribution, we run a daemon (\u0026lsquo;denkid\u0026rsquo;) which is taking data from pmcd, doing computations, and writing output to a textfile which pmcd is then reading and making available as metrics again.\nOk, so with this, we can already get the top-style ranking here: The sorting is still not working in version used for the screenshot, fixes are in PCP-7.1.2 and later.\nNow, let\u0026rsquo;s see how far this brings us under some scenarios!\nTest1: Consumers on hypervisor Running 4 processes on hypervisor (i.e. \u0026lsquo;md5sum /dev/urandom\u0026rsquo;), where each occupies 100% of one cpu-core. No KVM guests running or running but not generating considerable load.\nExpected: denkid is assigning each of the 4 processes 25%, and also \u0026lsquo;Watt per process\u0026rsquo; should show ~25% of the overall consumption State: GOOD Test2: Single process in guest Running a single process in the guest, we expect allmost all consumption to get attributed to the guest, that works too. The load from the guest appears here yellow.\nTest3: Mixed load hypervisor \u0026amp; guest We run 1 process on hypervisor (i.e. \u0026lsquo;md5sum /dev/urandom\u0026rsquo;), which occupies 100% of one cpu-core. 1 virtual KVM guest running, and running one instance of \u0026lsquo;md5sum /dev/urandom\u0026rsquo;.\nExpected: denkid is assigning the hypervisor processes 50%, and also \u0026lsquo;Watt per process\u0026rsquo; should show 50% of the overall consumption. denkid should assign 50% of overall consumption to the qemu-kvm process of the guest. State: NOTOK. qemu-kvm gets counted much more, i.e. twice of the hypervisor md5sum-process. Odd! Are my expectations off? So now we finally have the statement of the missmatch of expectation and outcome. But.. maybe just our expectation is off? Maybe the overhead of emulation is just so big that really the md5sum in the guest is using double resources?\nFirst test trying to find out: comparing \u0026ldquo;how much work\u0026rdquo; is done by md5sum on the hypervisor vs. the guest. In our case, with which data rate it is computing the md5sum. To measure, we install \u0026lsquo;pipeview\u0026rsquo;, it\u0026rsquo;s showing \u0026rsquo;throughput of a pipe\u0026rsquo;. Screenshot shows:\non top: consumption attribution below: troughput via pv on hypervisor bottom: troughput via pv in the guest This debugging is not for free.. we got now as complication that we have 2 processes causing load, not only one. :(\nRegarding \u0026lsquo;work/performance\u0026rsquo; we see the guest does 16% more work, but out attribution suggests that the KVM process did cause 4x the work of the hypervisor process.. hm.\nConclusion: I would not expect much overhead of the virtualization for CPU bound processes. Our guest even performs better, but is accounted much more workshares than expected. I\u0026rsquo;m still not sure if the accounting is incorrect, or my expectation.\nThoughts/analysis Observations:\n\u0026lsquo;pcp htop\u0026rsquo; shows the qemu-kvm process in yellow. Not sure why. The md5sum is green/red. \u0026rsquo;top\u0026rsquo; properly attributes same to qemu-kvm as to the md5sum process. \u0026lsquo;pminfo -f proc.psinfo.utime|egrep \u0026lsquo;qemu-sys|md5sum\u0026rsquo;\u0026rsquo; executed on hypervisor tells us the counters. Executing this with 10sec sleep in between shows that the counter is spinning faster for qemu-kvm, so conclusion: the oddity is not a calculation inside denkid or pmcd, it is already there in the /proc/PID/stat numbers. I confirmed with simple bash code running in loops. In my tests I see that as soon as the host also is under load, the counter for the guest-process seems to spin faster.\nFurther debug ideas:\nneed to confirm what the plain /proc/pid/stat counter expresses. Bounce this with a kernel specialist. Try to get down to easier reproducer? The accounting I use right now seems not ideal. Or is my expectation incorrect? How to express the power attribution differently? We want to attribute power consumption in the most accurate way possible. We see more consumption attributed to the md5sum in the KVM guest. We assume that there is just a small overhead for CPU-loads. We could try to measure where the energy is going: as byproduct it is heating up the system. We can look at temperature data.. could this help?\nWe could also try to clarify how sleep states and cpu clock are impacting the consumption. For the tests so far, I used systems with normal power save settings: this is clocking down unused cores, boosting single core under load, or clocking up multiple cores if there is enough load. And does put cores into C-sleep-states if unloaded. If we force the system to maximum clockspeed and to not go into C-states, then unloaded/loaded system will show same power consumption? Acutally, that\u0026rsquo;s a good experiment to do.\nI also had wondered if this is an x86 oddity, but had also seen this \u0026ldquo;VM with single cpucore-occupying process inside\u0026rdquo; getting attributed double of the same process natively on the hypervisor.\nWell, thanks for following along until here.. the initial version of the article had just the above contents, to order ideas and spark a discussion. And then it resolved. :)\nResolving As an outcome of the Mastodon thread, the issue got solved. After replacing \u0026lsquo;md5sum\u0026rsquo; with a pi-calculator (I used this), it appeared that the accounting was exactly as expected. This screenshot shows 2 processes running on the host, and 2 in the guest, and consumption attribution as expected:\nSo for \u0026ldquo;md5sum /dev/urandom\u0026rdquo;, we saw an additional load on the guest for supplying the /dev/urandom data. The hypervisor spends less cycles, due to /dev/urandom being available differently.\nRevisiting the previous ideas Ok, so with that let\u0026rsquo;s revisit also some of the other thoughts. On an idle Thinkpad T590, I see 2W overall system consumption.\nSingle cpu-eater, overall consumption: Running a single \u0026lsquo;pv /dev/urandom|md5sum\u0026rsquo; on the hypvervisor, I see 245MB/sec troughput, and overall system consumption going up to 10W.\nMostly 4 cores run with 2.5Ghz, the others with 400Mhz.\nRunning the same inside the guest: 265MB/sec, overall 10W, 6 cores at 2.4Ghz and two at 400Mhz. So even with more userland workcycles being done, no increased consumption - I think that\u0026rsquo;s because not single cores are clocked up/down, and we can do the other ontop workload \u0026lsquo;for free\u0026rsquo; without higher consumption.\nFour cpu-eateris overall consumption: Running 4 instances of \u0026lsquo;md5sum /dev/urandom\u0026rsquo; in the hypervisor, 15W overall consumption, all 8 cores at 2.3Ghz.\nRunning 4 instances of \u0026lsquo;md5sum /dev/urandom\u0026rsquo; in the guest: 15W overall system consumption, all 8 cores running at 2.4Ghz.\nPinning cores at high clockspeed, how is consumption changing? Sample size of one so far, Thinkpad T590: Pinning via \u0026lsquo;cpupower frequency-set -d 2.3Ghz\u0026rsquo;, we see this implemented via \u0026lsquo;/proc/cpuinfo\u0026rsquo;, and \u0026lsquo;pminfo hinv.cpu.clock\u0026rsquo;. Idle consumption stays at 2.8W, maybe system lying to us about clockspeed? Intel\u0026rsquo;s powertop is reporting 4W idle power, and also no change after we force the cores into 2.3Ghz. I should revisit this on ARM/Apple silicon. I also did nothing about C-states for this test, so allowed cores to go into deeper C-sleepstates.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-05-20T20:04:33+09:00","image":"/2026/05/the-riddle-of-attributing-power-consumption/pipeview2.png","permalink":"/2026/05/the-riddle-of-attributing-power-consumption/","title":"The Riddle of Attributing Power Consumption to Processes"},{"content":"The Cardputer is a small ESP32 based device with Wifi, Bluetooth, battery and keyboard. Dualcore 240Mhz cpu is quite something, it also has a MMU, but the 512kb of RAM are not enough for NetBSD or Linux.\nI got the device to play with Meshtastic/MeshCore, details here. Yet to be tested is my other use case: for FT8 transmissions, when connected to my QMX+ radio transmitter.\nMy hardware is the \u0026ldquo;M5Stack Cardputer\u0026rdquo; together with the LoRa1262 extension, for ~9000円/55USD. Small disappointment: the antenna is built for 868Mhz: I can use it on the 930Mhz area used for LoRa here in Japan, can do communication to nearby devices which use a proper 930Mhz antenna - but the LoRa1262 extension\u0026rsquo;s antenna won\u0026rsquo;t radiate well enough for far communications.\nFirst things The Cardputer has 8mb flash. Connecting via USB-C to Linux, you can install \u0026ldquo;M5Launcher\u0026rdquo; and from then on that launcher is executed after each start.\nPressing return enters the launcher, and allows to flash software via Wifi from the internet, or from files on the MicroSD card. Without pressing return, the currently flashed software is started.\nHere is a catalog of the software, plus some projects not listed there. Notable software/use cases:\nMeshtastic/MeshCore \u0026lsquo;fat\u0026rsquo; applications which list wlans, do network sniffing, and so on. Reminds of the \u0026ldquo;Flipper Zero\u0026rdquo;, just that the Cardputer has no RFID/ NFC build in. Badapple and Doom, of course :) A few games, emulators for NES etc. MP3 player, webradio Software which allows the Cardputer, connected via USB-C to an other system, to behave like a normal USB-keyboard or mouse The LoRa1262 extension has also GPS etc, so software to read that Antenna connector Trying to connect a 930Mhz antenna, I noticed: the connector on the LoRa1262 extension is not normal SMA: this turns out as RP-SMA, RP meaning \u0026ldquo;reversed polarity\u0026rdquo; of the physical connector. Stores in Akihabara were not able to identify this - but once you know this is RP-SMA, you can order a gender changer/adapter online.\nMeshtastic The catalog has quite some Meshtastic software, I tried:\nMeshtastic For Cardputer-Adv (ozbk): I mostly used this so far, no issues. Meshtastic for Cardputer ADV with Cap Lora (Gt4kill3r) is right now based on a slightly newer upstream Meshtastic software, 2.7.23 Cardputer Mesh Kit Meshtastic Firmware (M5Stack) is based on upstream version 2.7.22 Plai (d4rkmen), upstream, as of 2026/5/6 in version 2.1. I played with all of these - they are interoperable, also Plai is sending/receiving normal Meshtastic, apparently with an own implementation. The first 3 here are based on upstream Meshtastic, no big differences seen apart from changes coming in from upstream.\nPlai is much more shiny and polished, but has a big downside for me: after flashing the Meshtastic based variants, I can open https://client.meshtastic.org/ in Chromium and conveniently configure all settings. Plai is not offering that API, it\u0026rsquo;s really to be used on the Cardputer only.\nEurope settled on 868Mhz LoRa frequencies, Japan with 930Mhz. After handing out one of the 868Mhz antennas to a friend in error, I measured my antennas with a NanoVNA and am now wrapping each antenna with a label: blue is for 868/Europe range.\nMeshCore Already Meshtastic is rare to see here in Tokyo, even less MeshCore, so I have no longterm experiences with the MeshCore implementations.\nFor configuration, many Cardputer MeshCore firmwares come without a profile for Japan, so it\u0026rsquo;s important to manually configure for 920.6Mhz, 62.5, 8, 8, 20mW. Source.\nMeshCore-Cardputer-ADV M5Stack Cap LoRa1262 version!!!! (Stachu0) As of 2026/5/6 in version v1.1.0 MeshClient for CardPuter \u0026amp; ADV (whywilson) has as of 2026/05 no support for the Lora extension I have MeshCore (przemeks), upstream As of 2026/5, version v1.14.1.74 Looking forward to get a 930Mhz antenna for the Cardputer.. and to trying out FT8 with the radio transceiver. There is now a page on the Wiki for the Cardputer.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-05-06T12:20:39+09:00","image":"/2026/05/cardputer-for-lora-mesh-and-more/20260506_1306_cardputer_meshtastic.jpg","permalink":"/2026/05/cardputer-for-lora-mesh-and-more/","title":"The Cardputer M5 for LoRa, Mesh{tastic/core} and more"},{"content":"Remembering Blinry\u0026rsquo;s Fifty Things you can do with SDR, I thought that something like that for HAM Radio should exist - so here is my attempt of a collection. HAM Radio is a huge field, and most radio amateurs I know are interested in one of multiple of these fields. There is overlap between the areas, and room for discussion of what\u0026rsquo;s worth getting called out in this context.\nI\u0026rsquo;m trying to use rather easy to understand terms instead of HAM terms.\nUnderstanding the Physics and Building equipment It started with the discovery of the physics of electromagnetic waves, and transmission without cable. With that knowledge, receivers, senders, antennas and more could be constructed - so these were the first fields of interest for HAMs. With the frequency spectrum getting more used, and also for important services, also regulation started: you could either use equipment which was officially certified by the government, or you needed to pass an exam bevore sending on HAM radio frequencies. The right to built and use equipment on your own is still as of today an important right for radio amateurs. So doing exactly that is interesting for many people, as well as contructing antennas and various transmission lines, and other equipment.\nThe knowledge in all of these areas is constantly growing: the exam is confirming knowledge from various areas like regulations, law, physics and electronics. Based on that, new things get discovered, are from then on taught to the HAMs, and on that base the cycle continues. It\u0026rsquo;s really the feeling of \u0026lsquo;standing on the shoulders of giants and seeing farther\u0026rsquo;.\nThe \u0026lsquo;right to use own equipment\u0026rsquo; is a bit different depending on the country: in Germany, in passing the exam you have shown knowledge to build equipment which can operate following the laws/allowed specs. In Japan, you additionally have to register the specific device you want to use: either a radio transmitter from a big vendor - where the vendor obtains an official device-permission after the government running various tests. Or for your homebrew equipment: also that has to be checked by the government.\nMastering Morse Code Morse code (aka \u0026ldquo;Consinuous Wave\u0026rdquo; ow CW) was the first way of getting messages transmitted via radios. It sounds like just a simple sinus wave, sent out as sequences of short (\u0026ldquo;dit\u0026rdquo;) or longer (\u0026ldquo;dah\u0026rdquo;) signals. We all know the simple plates where then the alphabet and numbers get defined as sequences of dit\u0026rsquo;s and dah\u0026rsquo;s. One can memorize these quite easily, but for real communication one needs to understand them at higher speeds. Learning Morse Code is something between learning a new language, and learning a music instrument - here are details on my endavour.\nAlso for producing own radio transceivers: for receiving/sending morse code, much simpler transceivers can be used.\nYou might know the scene around building and using mechanical computer keyboards. In the same way, we have a huge scene for everything to insert Morse Code into the transceiver. To start with, there is a big field of kinds of input: the most traditional device is the \u0026lsquo;straight key\u0026rsquo;, when pressing it just a simple electrical circuit is closed. It\u0026rsquo;s also no ergonomic, so better devices like paddles were developed. If I ever get my Morse skills usable, I might reward myself in buying one of these.\nRagchew/normal talking Of course, you can also put \u0026ldquo;normal voice\u0026rdquo; onto the air for talking. This is widely done, and known as \u0026ldquo;ragchew\u0026rdquo;.\nGetting connections far away: DXing DXing: When folks have setup radio and antenna, then trying to get as far as possible is another \u0026lsquo;potential goal\u0026rsquo;. You can enter various counting organizations, compote with others. Try to get contact with as many countries as possible - for example contacts with North Korea are rare. Once you have contacts with all countries, try to get contacts to all of them with Morse Code, with voice, and with digital modes.\nTo get successful connections with rare countries or areas, sometimes \u0026ldquo;DX Peditions\u0026rdquo; are formed: a group of HAMs going to that area for one or two weeks. These are often super expensive, involving flights and ship trips. Often many other HAMs are also chipping in to make the trip possible.\nGoing out, into nature POTA, SOTA What if you live just in a tiny flat like me, and can not put up a big antenna? POTA (Parks On The Air) is an activity for going to a park, making a certain number of connections to others, and getting points/badges awarded. Of course, someone sitting at home and interacting with folks at the parks are also part of this. Then, for SOTA (Summits On The Air), you go to mountains and make connections from there! This brings together so many good things:\nespecially for higher frequencies one gets better propagation from summits, so better chances for wider communication You get onto the fresh air. Tuning fow small radio/antenna/batters (or solar cell) gets important. Also, being able to use Morse Code allows you to use a smaller radio, plus these connections give you more points! Thomas (K4SWL) has published some great videos which give an idea what POTA is like.\nThere is also a tradition of sending QSL cards: these are post cards which have the details of a successful communication. In many cases, communications are just \u0026ldquo;confirmed\u0026rdquo; online nowadays, but the tradition of sending QSL cards is also still alive.\nFoxhunt Anonther activity: setting up a device like a beacon which is sending out signals (the \u0026lsquo;fox\u0026rsquo;), and then have others locate the device. This is done with directional antennas, with participants competing on finding the fox first. Page with details, in German: draussenfuchs.de.\nField day For field day, folks gather on one place, setup equipment, talk with each other and make radio connections. Workshops, bbq\u0026rsquo;ing, educational classes for others - all potentially included.\nAntenna building I\u0026rsquo;m fully considering this as an own section. For HAM radio, every country has own rules, for example the frequencies on which HAM\u0026rsquo;s are allowed to send out radio signals. These \u0026ldquo;bands\u0026rdquo; are announced with their frequency or wavelength. And with the amount of power which HAM\u0026rsquo;s are allowed to send out - this also depends on how much the used antenna is \u0026ldquo;focusing\u0026rdquo; the power which the transmitter is putting into the antenna.\nDepending on the frequency used for sending out the signal, also the propagation is very different. Also the antennas. For highest frequencies, parabol dishes or horns get used, for the lowest ones long wires. For lower frequencies, also transmission trough the earth is more important.\nLow Power operation To make \u0026ldquo;as far as possible\u0026rdquo;-connections, many factors should be considered. Most importantly: antenna and \u0026ldquo;with how much power the signal is sent to the antenna\u0026rdquo;. The upper limts are defined by the law regulations and the exact circumstances (i.e. vicinity of other equipment or people).\nOn the low power end, QRP is a flavour of HAM radio where only low power is used, typically below 5 or 20Watt of energy to be output from the transceiver. This style of communication is putting more emphasis on the antenna. I have only operated QRP so far.\nBouncing communication: Satellite, EME The higher the frequency, the more it could be compared to light. For example, for the 430Mhz area, it is said that the signal is propagated \u0026ldquo;line of sight\u0026rdquo;, plus a tiny bit more, following the curve of earth\u0026rsquo;s globe. So sending signals from the top of a mountain is allowing the signal to propagate further as we have a better \u0026ldquo;line of sight\u0026rdquo;.\nThen, depending on many circumstances, signals can also be reflected by the various layers above the earth. Reflection can also happen with the traces left by meteors, or on planes.\nSome HAM\u0026rsquo;s specialize on Earth Moon Earth (EME): they aim at the moon, and work with the reflected signal. This has shifted to a completely new level recently with EVE succeeding.. Earth Venus Earth.\nCommunication can also happen directly with spacecraft like the international space station, or with satellites receiving the signal and \u0026ldquo;replaying\u0026rdquo; it back to the earth.\nData transmission This is around all kinds of data. For example, to run network protocols over radio signals. Or to receive sensor data from a balloon you are launching. If you have the balloon send out via Wspr procotol, you can have someone else receive the signal and obtain the transmitted data.\nHAM\u0026rsquo;s put all kinds of data onto the air:\nSSTV, slow scan television rtty, telewriter Digital Modes Morse Code was the first \u0026ldquo;protocol\u0026rdquo; to be used over radio waves, then voice communication. Later TeleVision. When the connection between 2 stations is good, then all of these can be used. If a connection is bad, then while voice will no longer work, chances are that you can still use Morse Code to communicate: so the \u0026ldquo;signal to noise\u0026rdquo; ration can be lower for Morse Code.\nIn the last years, also digital modes have spread. These cover something which \u0026ldquo;machines are sending/receiving\u0026rdquo;. For example, with the FT8 mode one can transmit even in situations when Morse Code does no longer work. On the other hand, using FT8 also needs equipment like a computer or a smartphone - these get connected to the transmitter and do the sending/receiving/encoding work.\nListening Sounds boring? Think again..\nFor starters, everybody can do it via internet access: for example via servers listed at receiverbook.de. That allows you to connect via internet to one of these systems, and then receive data \u0026ldquo;as of you had an antenna where that system is located\u0026rdquo;.\nRecently, a new number station appeared, with transmissions sent in Farsi, probably aimed at people located in Iran.\nSites like priyom.org have details on these, and with receiverbook.de you can \u0026ldquo;tune into\u0026rdquo; these transmissions.\nWith \u0026ldquo;The Spectrum Monitor\u0026rdquo;, there is a whole monthly magazine dedicated to listening. We have folks for example specializing into research of which frequencies are used by Russian ships, and then tracking movements. Same for ships in general: there are global networks dedicated to track ships positions.\nWe have HAMs who run satellite dishes on rotors, and receive transmissions from satellites. For example cards with weather data.\nMost interesting for me.. As for me, when I was young I started using Citizens Band, for communication with others, and because of the \u0026ldquo;magic of talking with someone over the air\u0026rdquo;. Learning towards the exam was a great trigger to learn more on the physics. I always wanted to learn an instrument, since last year that was replaced with the goal of learning Morse Code. I was already into hinking - now I have the basic equipment together to do POTA and SOTA.\nOne of my other areas of interest is Open Source, and it goes nicely together with HAM radio: it\u0026rsquo;s all about learning, sharing with others, and openness.\nIf you are curious which parts of the hobby are interesting for some people, I can recommend the QSO today podcast: in every episode, one HAM story are shown.\nFinishing this with a great comment about HAM radio which fits really nicely: it\u0026rsquo;s the Golf sport of the engineers. :)\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-04-25T17:29:49+09:00","image":"/2026/04/the-colours-of-ham-radio/20260328_111246_pota_backpack.jpg","permalink":"/2026/04/the-colours-of-ham-radio/","title":"The Colours of Ham Radio"},{"content":"At the middle of last year, I got all the hardware together to do radio transmissions here from my room in the West of Tokyo. It just did not work, and took half a year to debug. This here is a small recap, for me to understand what I could have done better. I will also sprinkle in some tech details - might make it an interesting read for people loosely interested in the topic, and so the specialists can point out which details I did not capture correctly. I\u0026rsquo;ll try to hit just the right level of depth to make it interesting, but not too technical.\nLet\u0026rsquo;s start with the hardware.\nThe antenna The antenna is an \u0026rsquo;exotic variant\u0026rsquo; of a capacitor, the piece of electronics we often think of as 2 metal plates next to each other. The antenna I bought for use from my flat was a telescope antenna, 5.6m long when extended. That metal rod can be thought of as one of the 2 metal plates, and a counterpoise can be the other part.\nWhen you receive a radio program, and use a telescope antenna, then you might have learned that it\u0026rsquo;s not very important if the antenna is full extended - but that\u0026rsquo;s just for receiving. For sending, this gets important: the antenna setup has to be well adjusted to the frequency with which you want to transmit. This antenna is sold as \u0026ldquo;designed for 14M-30MHz operation\u0026rdquo;.\nThe radio transmitter HAM radio operators can build their own transmitters, Here in Japan, you also have then to get it certified by the authorities, to be sure that your transmittions are following the rules. I ordered a small transceiver which could be used on various frequencies between 50Mhz (6m band) and 1.8Mhz (160m band). You find these references to frequency (in Hz) and wavelength (in meter) everywhere, they are directly related: a signal at 1.8Mhz frequency has a wavelength of 160m.\nBy the way, while the QMX+ can send on basically all HAM bands between 6m and 160m, with the lowest level license in Germany, you can only send on 10m. The license also covers 2m and 70cm operation, but the device is not operating on these bands. With the smallest licsense in Japan, you can also send on these frequencies.\nSo now I got a shiny QMX+ transceiver here, and connected it with a 5m coax cable to the antenna. That\u0026rsquo;s it?\nReceiving For receiving, yes. The QMX+ not an analog transceiver, but all the signal handling is digital. You can connect headphones, select a frequency via knob, and can then listen to voice or morse code signals. When connecting a morse key, you can send morse code - but my morse skills do not yet allow that. My intended use for now: connect via USB-C to a Linux computer, and do digital communications over radio wave. One popular mode of communication is called FT8.\nMy flat is at the 7th floor of a steel frame building, with an antenna inside the room the building is acting like a faraday cage and \u0026lsquo;catching it all away\u0026rsquo;. Why are cellphones work nontheless? Because the strength of the signal and distance to the cell tower are tuned for this. Typical HAM radio communication aims at communication further away. So, sticking the antenna outside the balcony, I can receive morse code, and FT8 signals from other people.\nYay!\nHow to verify sending.. That was the state as of autumn last year. Receiving was fine, but not sending. How to try this, most straight forward way? To run the wsjtx software, setup antenna and cable so that I see that I can receive FT8 from other stations, and then to try to send by myself. FT8 allows for example to send messages ala \u0026ldquo;who can hear me? Please reply.\u0026rdquo; or to address specific stations whose signals I can see.\nThe addressing works with the HAM radio callsign we got assigned - so as soon as you see messages from someone, you see their radio name, can conclude on their country and can look them up. Just, nobody seemed to hear when I was sending. The QMX+ seemed to indicate that I was sendingm just noone did reply.\nSearching for the issue: SWR First check was a technical one, checking the Standing Wave Ratio, SWR. A bit simplified, the SWR ratio is a number which can for a setup tell us if the antenna is tuned well. If the SWR ratio is bad, the antenna is not radiating the transmission, and our transceiver can even be damaged. The QMX+ is really cool: it has a few knobs and buttons for direct control, but via USB-C it also allows to open a terminal, and offers menus! Via \u0026lsquo;screen\u0026rsquo; or \u0026lsquo;minicom\u0026rsquo; from Linux. On these menus, the QMX+ is also offering us to \u0026lsquo;swipe\u0026rsquo; through frequency areas, send out at varying frequencies, and measure how much power is reflected. The result is shown as ASCII-art diagram, really nice. For the 6m or 10m areas, I got the setup into a state with a reasonable SWR of 1.5. For this, it was also important to attach a cable to the foot of the antenna, as \u0026lsquo;counterpoise\u0026rsquo; (German: Gegengewicht). The cable was also put onto the balcony, it\u0026rsquo;s then basically acting as \u0026rsquo;the other part of the capacitor\u0026rsquo;.\nSo that part of requirements for sending looked ok. Also, if the SWR is really high, the QMX+ is turning into a \u0026lsquo;I need to protect myself\u0026rsquo;-mode and stops sending. That\u0026rsquo;s clearls shown on the display.\nI later also got a NanoVNA: a device which can get connected to the coax cable instead of the transceiver, and which will on a graphical display show you the SWR for various frequencies.\nGetting third opinions: pskreporter First approach was to get \u0026lsquo;further opinions\u0026rsquo; on whether I\u0026rsquo;m sending something out or not: reports from others. Sites like pskreporter are really nice: on these, many contributors all over the globe report which stations they hear - and then you can look them up by callsign. Software wsjtx can also participate there: so now I could on that side see visually whose FT8 messages I was receiving. Neat! There are further pages like this.\nMore important, the page also confirmed that noone was receiving anything from me. Hm.. but the QMX+ seemed to send out..\nSniffing myself Then I did setup another Thinkpad with Debian, and an RTL-SDR stick: with that I should be able to confirm if I see my own signal. I used software OpenWebRX+ before to decode FT8 data. It\u0026rsquo;s using wsjtx under the hood for the decoding, but that would segfault constantly. Using the wsjtx package from upstream, instead of the package from Debian Trixie repos, did not help either. I found a command line client to decode FT8, but then I discovered that FT8 decoding started to work despite the segfaults.\nI thought I had seen my signal in the waterfall diagram on the monitor-system.. but I did not see my FT8 signal decoded. When looking at the frequency with the GQRX software, I saw that the QMX+ seemed to constantly send out a signal! When I tried to send, it would switch to a different frequency, then go back. This seemed to say \u0026ldquo;it\u0026rsquo;s sending!\u0026rdquo;, but I think now that I\u0026rsquo;m seeing something here just from the QMX+ directly, because the sniffing system is just 1m away. Not really a signal sent out.\nDifferent antenna Well, I anyway wanted to also get another antenna for portable operation: I did not have different results for my tests though. This is a cable antenna, so a cable to be hung down from the balcony, and also another cable as counterpoise attached.\nPower consumption I tried to consider power consumption: I got myself a USB-C to 12V-barrel adapter, with that I could power the QMX+ via USB-C. More important: I could use my Power-Z meter to confirm how much energy was consumed: consumption seemed to not go up, even though the device seemed to tell me it was transmitting! This made me focus more onto the transmitter itself.\nMy next step would now have been to attach a morse key, and send a short morse signal. People all over the world have webradio receivers on the internet for public access, for example receiverbook is a nice overview. Then I would use one nearby, tune to the frequency I have set, and should be able to verify if I see my signal.\nFor that, I was reading the QMX+ manual again, and found the hint I was missing before: when trying to send something, I was seeing 2 square fields, a bit like bars visualizing signal strength, and my interpretation was that it\u0026rsquo;s sending. That was wrong. The computer software is communicating with the QMX+ via USB-C, over this the QMX+ is offering soundcard interfaces: for sending audio to the computer to interpret, and for receiving.\nResolving The issue: that sound level of the computer sending to the QMX+ was set to low per default. Simply pulling that up made the signal getting sent out. Yes, this is also a reminder that manuals should be read properly. But at least I also learned a few things while debugging. :)\nI\u0026rsquo;ve contacted QRP labs Hans Summers, he is considering a change in the user interface to more clearly show when the volume level is to low, for example with a flashing \u0026ldquo;E\u0026rdquo;.\nThe location here is still not good for sending, but then there are nearby parks, and season is great:\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-03-29T18:34:36+09:00","image":"/2026/03/debugging-ham-radio-transmission/20260321_124359_pota_setup_bike.jpg","permalink":"/2026/03/debugging-ham-radio-transmission/","title":"Why is noone hearing it? Basics of sending radio waves."},{"content":"I got this Power-Z measurement device the other day.. what is it, and how does it fare against the other measure methods? My last comparison was in 2022.\nThe contenders I have some sources for measurements here:\nThe Power-Z USB meter, a small device which can be put \u0026lsquo;in the middle\u0026rsquo; of a power connection. It has a USB-C power input, USB-C power output, and a third connector over which I can retrieve the metrics. For the retrieval, a Linux driver exists - directly in the upstream kernel. Switchbot smartplug 100V, running Tasmota firmware. This one hooks into the wlan and allows measurements via API. For the Tasmota, we also are getting multiple \u0026lsquo;consumed power\u0026rsquo; metrics back, not just one. Which one is right? RAPL: these metrics are available on most modern x86 systems. Macbook Pro M2 internal sensors, I retrieved data from these via Asahi Linux on the system. Measure procedure Like in earlier setups, I used 2 systems:\nThe observer: this one runs my Python scripts which are performing the tests. pmlogger is gathering metrics and archiving them into files. Smartplug and Power-Z are queried directly. The system under test With this setup, I did run multiple workloads. First one just \u0026lsquo;idle\u0026rsquo;, the other workloads putting the system under load, keeping up to 8 cpu cores busy. Like in past tests, these were loops which were uncompressing data. This time, we do not care about the power efficiency, but how much the measured metrics divert from each other.\nLet\u0026rsquo;s look at the results, based on measurements with 3 compute devices.\nThinkpad L480 This is an oldish Thinkpad, produced 2018, with 4 cores.\nOn the left, with 0 threads, we see the idle measurement. Clearly, the RAPL-MSR metrics on this old box are unusable, not increasing at all as we use up more power. The Tasmota stays always above the Power-Z.\nSteamDeck First, SteamDeck with dock.\nNow without dock:\nThe dock has a rj45 network and an HDMI connector, we clearly see how these use up extra power. Especially comparing the idle-values on the left, we see that for measurements without dock the RAPL- measurement is much more in line with the external sensors.\nThe smartplug is meant to cover consumptions up into kilo-Watt areas, it\u0026rsquo;s good to see it not much diverging from the Power-Z metrics. SteamDeck RAPL metrics are better than the L480 metrics, but stay way below what the Power-Z measures: for the highest consumption, RAPL reports 73% of what Power-Z measures from the outside.\nMacbook pro M2 The Macbook pro M2 running Asahi Linux/Fedora 43:\nNice! Apple has done a good job with these sensors.. matching up nicely with the Power-Z readings: there is just 2-3% difference!\nThe MacBook allows insights there as it exposes access to 3 relevant sensors:\n(1) is telling us with how much Watt the battery is drained (if it is used) (2) is input power consumption (strangely labeled \u0026ldquo;AC power\u0026rdquo;, but it\u0026rsquo;s PD3.0, DC power. (3) total system consumption From looking at these, it seems like drain from battery vs. supply from external power supply (those metrics are in line with the Power-Z readings) are same.\nAnother interesting sensor of the Macbook pro M2 is the one called \u0026ldquo;Heatpipe power\u0026rdquo; in \u0026rsquo;lmsensors\u0026rsquo;, the underlying sensor key is PHPC: this is an estimate of the heat energy being dissipated by the SoC.\nPD3.0 supply from external is supplying 20V, and internal battery is ~12V. For a given workload, I see the same consumption over external supply vs. internal battery, so seems like efficiency for both voltages is identical.\nSummary Some years ago, with the first measurements, I was still not sure which measure methods I could trust. With the above results, the Power-Z looks really well. Also Apple has done a great job with their sensors!\nTake aways this time:\nSmartmeter and RAPL are better than nothing, but Power-Z and internal Mac sensors are best. The Mac sensors are so nice, you could also use them to measure consumption of external USB devices you attach to the MAC. Although RAPL is off, it\u0026rsquo;s still good enough to compare the energy consumption caused by various workloads on the same system. If you have an x86 system, you could consider to compute a correction factor for RAPL with a precise consumption meter, and then use RAPL+correction to have better values in a setup without external meter. I would like to look at RAPL on newer x86 systems, but the SteamDeck is the newest system here available.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2026-03-23T20:03:40+09:00","image":"/2026/03/comparison-of-computer-power-measurement-methods/20251201_151215_powerz-meter.jpg","permalink":"/2026/03/comparison-of-computer-power-measurement-methods/","title":"Comparing Power Measurement Methods for computers"},{"content":"How good is the performance of mobile phones, and how power efficient are they?\nI have already compared performance/efficiency of architectures with each other. How are mobile phones doing?\nThe device enabling me to measure this: the Power-Z USB meter. It\u0026rsquo;s around 100$, but has a native Linux driver. Together with Performance Co-Pilot and the scripts I used in the past, I looked at these mobile phones:\nHuawei P20 lite (2018), cpu: 2 x Cortex-A53 @ 2.36GHz Pixel 4a (2020), cpu: 2x2.2GHz Kryo 470 Gold + 6×1.8GHz Kryo 470 Silver Pixel 9 (2024), cpu: 1x3.1GHz Cortex-X4 + 3x2.6GHz Cortex-A720 + 4x1.9GHz Cortex-A520 How to get a shell on mobile phones? I installed Termux for this. The version from F-droid turned out to old, I used a release from the github site. After installing, you get a shell running as a user. This worked out nicely on my 2 Pixel phones running GrapheneOS, and the Huawei running Emui, which is Android 9 based.\nMy workload was again uncompression of bzip2 data, like in the past. A detailed description is in this article.\nRunning the workload on Android, I had to take extra care of Android not prioritizing the workload down or even freezing it.\nResults Number of total extractions, Linear and Logarithmic scale:\nWe have many single jobs running in separate threads, so we see nicely how overall job number increases with increasing thread number, as long as the system has free cores to put the threads on.\nConsumption per single extraction:\nPixel 9 and Macbook pro M2 are the winners - I did not have current x86 systems around for comparison.\nCorrections? Questions? -\u0026gt; Fediverse thread.\n","date":"2026-01-11T21:50:10+09:00","image":"/2026/01/performance-and-power-efficiency-of-smartphones/20251201_151215_powerz-meter.jpg","permalink":"/2026/01/performance-and-power-efficiency-of-smartphones/","title":"Performance and Power Efficiency of Smartphones"},{"content":"TL\u0026rsquo;DNR: My steps towards emulating an Amiga 1000 keyboard, using the GPIO (input/output) pins of a Raspberry Pi, and implementing parts of the keyboard protocol in Python.\nBackground Around 1997 I got an Amiga 1000 (\u0026ldquo;A1000\u0026rdquo;), without keyboard or floppies. Released in 1985, the A1000 has no harddisk, and after booting is asking for a 3.5\u0026quot; floppy disk with kickstart firmware.\nOnce that is loaded, another floppy with the workbench (a graphical interface) or floppies with applications or software could be inserted.\nThe disk drive is different from PC 3.5\u0026quot; drives, to obtain these floppy disks I\nconnected an Amiga 500 via serial cable to a Linux box inserted an unused floppy and had the A500 floppy write the images of kickstart and workbench Details are on the wiki. With these floppy disks, the A1000 booted fine into the workbench.\nAlso games like Turrican 2 did start up, and worked with a joystick. The mouse also worked. Just no keyboard..\nHardware The A1000 has a 4 port connector for the keyboard, which turned out to be rj10/rj22. As of 2025, in Germany, cables with fitting connectors are sold as \u0026ldquo;handset cord\u0026rdquo; (\u0026ldquo;Hoerer Kabel\u0026rdquo;), so the cable connecting the handle with microphone/speaker to a classic telephone. At my first attempt, I got rj11 plugs delivered - these are to wide and to not fit the A1000 keyboard outlet.\nSo, the idea to implement the protocol which the A1000 is expecting from the keyboard, and transmit keystrokes. How hard could it be?\nThis is the required hardware:\nrj10/rj22 plug with cable, used to interface with the A1000 Raspberry Pi or a similiar device which can set single output pins. The GPIO from the Raspi is nicely supported, and the Raspi also runs fully fledged Linux, which I\u0026rsquo;m more comfortable with than first having to adapt to a development kit of smaller hardware. 5v/3.3V logic level shifter, as the Raspi is working with 3.3V on its GPIOs, while the Amiga is working with 5V Multimeter to measure voltage is convenient An LED, a 330Ω resistor to confirm power direction solder iron breadboard, a piece of plastic with holes which are connected in a defined pattern. When we stick pins with wires into the holes, the board is creating electrical connections. All of this could also be done with soldering, but the breadboard is much better for quick prototyping connections. some cables with pins, these are for use with the bread- board Sets of breadboards with cables, LEDs and resistors can be ordered from many places.\nRaspi setup I did choose the Raspi to be sure to have enough computing power, and to have a full Linux available. After flashing Raspberry Pi OS, I looked up simple guides and got simple Python code together to confirm these:\nLED blinking, to verify GPIO is working GPIO input: so when then shorting out 2 contacts on the breadboard with a cable, the Python code would notice and report this Hardware setup, Amiga side Starting with the Amiga side, I bought a cable with rj10/rj22 connectors on both ends, cut it into half, separated the 4 single cables inside, and soldered them to 4 short cables with pins at the end. These 4 pins can then be plugged into the breadboard for further experiments.\nDetails on the pin layout are here, we will refer to these as\npin 1: +5V pin 2: CLOCK pin 3: DATA pin 4: GND With the multimeter, at this point we can verify:\nthat there is 5V current between +5V and GND also 5V between CLOCK/GND and between DATA/GND Bringing the hardware together Next, I soldered pin-rows to the level-shifter board and placed it on the breadboard. We need the level-shifter to translate between the 5V signals from the Amiga, and the 3.3V signals from the Raspi. If your hardware can directly deal with 5V (i.e. most Arduino do), then you do not need the level-shifter.\nTo operate properly, the level-shifter needs 5V supplied on HV/GND, the Amiga is providing that for us. The level-shifter also needs 3.3V on LV/GND, we take that from the Raspi.\nThis order should then be used for connecting:\nThe 4 pins from the Amiga should be connected to the level-shifter. Amiga-GND goes to Shifter-GND, Amiga-+5V goes to Shifter-HV, Amiga-CLOCk goes to HV2, Amiga-DATA goes to HV1. We then connect 3.3V to LV and the GND next to that to the Raspi With that, we can use the multimeter and should confirm that we see 3.3V between LV1/GND(on LV-side) and LV2/GND(on LV-side). The Amiga is setting these 2 lines \u0026ldquo;up\u0026rdquo; for us, and with the Raspi we will set these \u0026ldquo;down\u0026rdquo; for the signals.\nConnections from the Raspi:\nRaspi-pin6 to Shifter-GND, that\u0026rsquo;s ground Raspi-pin1 to Shifter-LV, that\u0026rsquo;s providing the 3.3V Raspi-pin7 to Shifter-LV2, that\u0026rsquo;s the GPIO 4 of the Raspi Raspi-pin29 to Shifter-LV1, that\u0026rsquo;s the GPIO 5 of the Raspi Software side For sending the first key press, I\u0026rsquo;m booting the Amiga into workbench, and start the notepad - so most key presses should be visible immediately.\nFor the Raspi side, I used 2 sources:\nHere is an outline of the protocol we need to implement, and a matrix with all the keycodes the keyboard can send kbdbabel.org has this fantastic diagram which made things much more clear for me. An excerpt: In words:\nboth CLOCK and DATA lines are \u0026ldquo;up\u0026rdquo; by default a single transmission starts with the keyboard taking down DATA for 20μs. Then we wait 120μs. then the 8 single bits are transmitted. For each of these, DATA is either taken down (representing a 0) or stays up (representing a 1), and then the CLOCK line is taken down for 19μs. We basically bring DATA into the desired state and then trigger CLOCK to tell the Amiga \u0026ldquo;now look at the DATA line\u0026rdquo;. the last bit represents the key state, a 0 means \u0026ldquo;button down\u0026rdquo; and 1 means \u0026ldquo;button up\u0026rdquo;. If we just send \u0026ldquo;down\u0026rdquo;, then the Amiga will interpret it in repeated key presses. last, there is a final part: my implemntation is here just waiting, but we should seek here for an ACK message from the Amiga. This graphic is implemented in simple-keypress.py, it\u0026rsquo;s simply sending the sequence to signal of pressing down the \u0026lsquo;k\u0026rsquo; key. If you wired up as per above, this should work now.\nCurrent code base, next steps The current code, amikey.py, does this:\nenters a loop waits for a keypress it will lookup that key in a keytable, send the keycode together with \u0026lsquo;key pressed\u0026rsquo; then send the key code with \u0026lsquo;key released\u0026rsquo; enter the loop again The code is for now at https://fluxcoil.net/files/scripts/202512-amikey.\nDetails on the files there:\namikey.py, most usable implementation to send keycodes directly to the Amiga. Once started, you will need to end this code in sending a kill signal from another terminal. amikey-input.py, simpler input loop, you need to press return after entering the key code amikey-wait-for-ack.py code implementing the attempt to read the ACK from the Amiga simple-blink-test.py, simple blinking LED simple-input-test-pulldown.py, simple code to test pull down input simple-input-test-pullup.py, simple code to test pull up input lines simple-keypress.py, the most basic code to send a single keycode Next steps/future extensions:\ninput without a final \u0026lsquo;return\u0026rsquo; should be taken in and transmitted the code should fetch the \u0026lsquo;host ack\u0026rsquo;. The code in amikey-wait-for-ack.py is sending k-down and k-up, and after the second keycode goes into a loop for detecting state change from the other side indicating the ACK, but the code is not detecting that. Maybe I should verify if I see an ACK with an oscilloscope, maybe after sending chars in a loop. I could also open the Amiga 500 and try to sniff the protocol there. Right now, simple \u0026lsquo;sleep\u0026rsquo; are implemented. That\u0026rsquo;s not accounting for the processing time of the other functions, ideally we would look at a wall clock and act based on that? Porting to other languages might make this also more portable to other devices. One really fun implementation I\u0026rsquo;m pondering about: Commodore 64 have a big keyboard, connecting one of these to the Raspi and using the C64 as keyboard for the Amiga would be great!\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-12-28T22:52:41+09:00","image":"/posts/2025/12/emulating-the-amiga-1000-keyboard/20251226_key_input_k.jpg","permalink":"/posts/2025/12/emulating-the-amiga-1000-keyboard/","title":"Emulating the Amiga 1000 Keyboard"},{"content":"TL\u0026rsquo;DNR:\nIntroduction into Meshtastic/Meshcore, hardware recommendations, first steps. If you get 2 devices, you can already play. Depending on where in the world you are, you might see much Meshtastic/Meshcore used, or none at all. Meshtastic seems much more widely spread than Meshcore.\nTerms LoRa (\u0026ldquo;Long Range\u0026rdquo;) is a technology for transmitting data via radio signals. It\u0026rsquo;s designed for much longer distances than Wifi: the record for LoRa is around 300km, and typical distance 10km. Downsides: less throughput than Wifi, and higher latency. LoRa defines various profiles, these allow to fine tune latency/ throughput versus distance of the participating stations. Meshtastic is an OpenSource firmware, designed to run on a variety of devices. Meshtastic implements a mesh network, so even if participating nodes can not reach each other directly via radio, they can exchange data via nodes in the middle. Meshcore is an alternative to Meshtastic, with different pros and cons What do you get from playing with Meshtastic/Meshcore? LoRa devices, together with Meshtastic or Meshcore can communicate with each other. Some of the most common goals of people playing with these:\nTo establish a way of communication alternative to the internet. I\u0026rsquo;ve seen areas in Germany with high density of Meshtastic devices, and active communication in chat channels: people greeting each other in the morning, and so on. So usage as public medium for communication, a bit like Citizens Band (CB) radio. People using LoRa devices to regularly transmit data for home use: this might be temperature data from a sensor which is deployed in a distance not covered by Wifi any more. Or turning a water pump on/off. There are also fun projects with people connecting retro computers like Commodore64 to LoRa hardware, and with that establishing a \u0026ldquo;chat channel\u0026rdquo; where all of their computers can communicate together For range testing, trying to transmit over as big distances as possible. If this sounds interesting to you, you might also consider HAM radio as a hobby. :) Recommended hardware The first guardrails for hardware are set by the used frequencies, which depend on the country where you want to use Meshtastic/Meshcore. I\u0026rsquo;ve purchased devices for Germany and Japan:\nEurope (including Germany) is allowing the 868Mhz band (863-870MHz) for LoRa, I use Heltec LoRa32 V3 devices here. Japan is allowing the 920-925Mhz range, that\u0026rsquo;s also covered by the devices from above. Additionally, 433Mhz area is used in Germany: the HAM radio group in my home town is operating Meshtastic gear in that frequency range, and Heltec LoRa32 V3 also for 433Mhz are available. I bought 4 Heltec devices for 868/923Mhz use, and one for 433Mhz. Having 2 or more of a kind is recommended, so you can test the communication without relying on other nodes.\nThe Heltec are cheap, but also have a reputation of wasting energy - when providing power from batteries that\u0026rsquo;s important. The more expensive but power saving alternative is the RAK 4631 module, including nRF52840 with SX1262-LoRa-Modem. That\u0026rsquo;s a modular system, for use with Meshtastic you need to purchase multiple components and combine them.\nNote: I connected the antennas which were labeled for use on 868Mhz and 923Mhz area to a NanoVNA-v4: turns out they are completely unfit for the 923Mhz area, with an SWR over 5. The NanoVNA might be to expensive for purchasing just for this check, but if you have a HAM club in your area, ask for help for this check!\nInitial Setup w/ Meshtastic I spent 30min trying to do something useful with the default software of the Heltec devices, installed the Arduino software suite, but gave up and flashed Meshtastic onto the devices. The Meshtastic site is great for that, flashing can happen from the Linux command line, or with a Chrome browser.\nCommunication with the Heltec devices is possible via:\nWifi: I did not try that Bluetooth: via the Meshtastic app, on Android or iphone serial via USB-C Bluetooth communication via Meshtastic app is simplest for beginners, navigating the menus is easier than looking up command line options. For command line access on Debian 12 system, this can be used:\nsudo apt update sudo apt install python3-full python3 --version pip3 --version The user should then be added to the group \u0026lsquo;dialout\u0026rsquo;, to get access to the Heltec. Then I connect the Heltec via USB-C to the system, and continue:\ncd ~ python3 -m venv meshtastic export PATH=\u0026#34;~/meshtastic/bin/:$PATH\u0026#34; pip3 install --upgrade esptool pip3 install --upgrade pytap2 pip3 install --upgrade meshtastic pip3 install --upgrade esptool # This does a first communication esptool chip-id # Flash firmware via webflasher and Chrome browser, or # get firmware from https://meshtastic.org/downloads/ ./device-install.sh -f firmware-heltec-v3-2.6.11.60ec05e.bin Now we need to tell the configure the device for our region:\n# for Japan: meshtastic --set lora.region JP # for Europe, 868 range: meshtastic --set lora.region EU_868 meshtastic --get lora meshtastic --get lora.region meshtastic --get 0 If we did all of this also on the other Heltec, we can now connect power to the devices, and if they are within range, they will \u0026ldquo;see\u0026rdquo; each other as peers, we can send messages. For message transmission, one node should be listening via \u0026ldquo;meshtastic \u0026ndash;reply\u0026rdquo; and the other node should send via \u0026ldquo;\u0026ndash;sendtext\u0026rdquo;:\nmeshtastic --nodes meshtastic --sendtext \u0026#39;Hello Mesh!\u0026#39; meshtastic --reply With this, message exchange should work! For connexting to the device with the Android mestastic app via Bluetooth, let\u0026rsquo;s also set the pin:\nmeshtastic --set bluetooth.fixed_pin \u0026lt;some-pin-here\u0026gt; meshtastic --set bluetooth.mode FIXED_PIN It is also recommended to set the position of the LoRa device, and set a proper device name:\nmeshtastic --set-owner \u0026#39;CHtastic1\u0026#39; --set-owner-short CHT1 meshtastic --set position.fixed_position true --setlat 37.8651 --setlon -119.5383 Meshcore Meshcore is an alternative to Meshtastic. Quite active project, quickly getting traction, but right now there are much more Meshtastic nodes than Meshcore nodes. While Meshtastic is forming for mesh network for you, Meshcore is more bare bones: leaving it to you to have the device send out \u0026ldquo;advertisements\u0026rdquo;, to configure routes to other nodes and so on. The idea is that you get more stable routes and operation. Also Meshcore supports the Heltec V3 devices, and has an Android app.\nCommand line client setup:\nsudo apt update sudo apt install python3-full cd ~ python3 -m venv meshcore export PATH=\u0026#34;~/meshcore/bin/:$PATH\u0026#34; echo \u0026#39;export PATH=\u0026#34;~/meshcore/bin/:$PATH\u0026#34;\u0026#39; \u0026gt;\u0026gt;~/.bashrc # Now flash firmware, i.e. with the Chrome browser pip3 install --upgrade pytap2 pip3 install --upgrade meshcore-cli # https://github.com/meshcore-dev/meshcore-cli meshcore-cli -l meshcore-cli -s /dev/ttyUSB2 info First commands:\n$ meshcli -s /dev/ttyUSB0 chat clock clock sync set lat [..] // set your position, Latitude set lon [..] // set your position, Longitude get name set name get radio set radio 869.618,62.5,8,8 # Setting the mode/freq. of neighbouring nodes in Germany. # example: 869.525,250.0,11,5 // this is the default # requires reboot advert floodadv node_discover public hi world list Here is a nice map with Meshcore nodes. I noticed that the nodes around me were all operating with different settings than the Meshcore defaults, so above \u0026lsquo;set radio\u0026rsquo; is configuring the special settings.\nPractical operation, Japan After initial setup and testing transmission between my own devices, I was running one device for some time at my flat in Tokyo, without seeing other nodes. The house is a big metal frame, but also with the antenna outside the balcony no other nodes were seen. I also took the device to a day trip to Mnt.Nabewaliyama: connected via USB-C to the Pixel9 with GrapheneOS, the Pixel9 is providing the power and can control the Helvec device via serial connection. No other nodes were seen though. First other node was when a fellow HAM person brought in his device to a HAM radio meeting.\nThe devices also support an extra setting for HAM radio operators: with that, you have to provide your HAM radio callsign, and communication is happening only unencrypted. The device is then allowed to send with more power.\nPractical operation, Germany In December 2025, I took the devices with me to Munich, and experienced a quite busy Meshtastic network for the first time. I just operated from a hotel room on the 1st floor, and saw many other Meshtastic and Meshcore nodes. Then I did spend some days near Wiesbaden, and again: just using the devices from inside the room, I saw quite some traffic. Meshtastic is then also showing you other nodes which you can not reach directly, but with intermediate nodes, \u0026ldquo;hops\u0026rdquo;. The public channels here were used for casual chat like \u0026ldquo;Good morning. The winter has come now..\u0026rdquo; (in German).\nSetting up mailboxes in such busy areas can be fun, so you could then push messages for other people into such a mail box, and other people could fetch these later - even if they were not online while you were sending the message.\nThen I travelled to my home town in Thuringia, the middle of Germany. Here, run an old Thinkpad x200 at the attic, and connected 3 of the Heltecs via USB. I run a network cable to the Thinkpad, and can from remote conveniently control the Heltec devices - one runs Meshcore, the others Meshtastic. The Thinkpad has also an RTL-SDR device attached, and OpenWebRX+ service allows me to remotely investigate what the SDR-device is receiving. Also when turned off, the Thinkpad is providing power for the Heltec devices. Messages etc. are buffered there, until I access the Heltec devices again via command line.\nAs for Meshtastic: after more than a week, I did not see any other nodes on 433Mhz. On 866Mhz I saw one other node directly (0 hops), and behind that node various other nodes:\nchris@tower:~$ meshtastic -s /dev/ttyUSB1 --nodes Connected to radio ╒═════╤════════════════════════════════╤═══════════╤═══════╤═════════════════════════╤══════════════════════════════════════════════╤═════════════╤════════════╤═════════════╤════════════╤═══════════╤═════════════════╤════════════════╤════════╤════════╤═══════════╤═════════════════════╤═════════════╕ │ N │ User │ ID │ AKA │ Hardware │ Pubkey │ Role │ Latitude │ Longitude │ Altitude │ Battery │ Channel util. │ Tx air util. │ SNR │ Hops │ Channel │ LastHeard │ Since │ ╞═════╪════════════════════════════════╪═══════════╪═══════╪═════════════════════════╪══════════════════════════════════════════════╪═════════════╪════════════╪═════════════╪════════════╪═══════════╪═════════════════╪════════════════╪════════╪════════╪═══════════╪═════════════════════╪═════════════╡ │ 1 │ bemu_t114 │ !908896cb │ bm_t │ HELTEC_MESH_NODE_T114 │ ran0TVJslilQXklYSwGGwFOliZSI0DMxS8Ebiyq0eSg= │ CLIENT_MUTE │ N/A │ N/A │ N/A │ 82% │ N/A │ 0.02% │ -5 dB │ 1 │ 0 │ 2025-12-26 08:17:14 │ 1 min ago │ ├─────┼────────────────────────────────┼───────────┼───────┼─────────────────────────┼──────────────────────────────────────────────┼─────────────┼────────────┼─────────────┼────────────┼───────────┼─────────────────┼────────────────┼────────┼────────┼───────────┼─────────────────────┼─────────────┤ │ 2 │ Station MHL │ !ba655ca8 │ SMHL │ HELTEC_WIRELESS_TRACKER │ zrYq1GzUyr9tFRJuirZ0CJUKGC9rBtgEW7QUBWbiAFk= │ N/A │ 51.2295° │ 10.4530° │ 246m │ Powered │ 2.79% │ 0.32% │ -5 dB │ 1 │ 0 │ 2025-12-26 08:10:44 │ 8 mins ago │ ├─────┼────────────────────────────────┼───────────┼───────┼─────────────────────────┼──────────────────────────────────────────────┼─────────────┼────────────┼─────────────┼────────────┼───────────┼─────────────────┼────────────────┼────────┼────────┼───────────┼─────────────────────┼─────────────┤ │ 3 │ Jens Stadt Repeater │ !b03d8f2c │ JSR │ HELTEC_V3 │ Pxm8Vk3RE0yuM85qPVsIVLV/uHR+56FILMonbTW6VEY= │ N/A │ N/A │ N/A │ N/A │ 95% │ N/A │ 0.21% │ -5 dB │ 0 │ 0 │ 2025-12-26 07:48:13 │ 30 mins ago │ ├─────┼────────────────────────────────┼───────────┼───────┼─────────────────────────┼──────────────────────────────────────────────┼─────────────┼────────────┼─────────────┼────────────┼───────────┼─────────────────┼────────────────┼────────┼────────┼───────────┼─────────────────────┼─────────────┤ │ 4 │ Jens Forstberg │ !05bb1c93 │ JF │ HELTEC_MESH_NODE_T114 │ Ixg+prZxgX30xDFO6fYykJulPv6McfH4ejumhEFq7iA= │ N/A │ N/A │ N/A │ N/A │ Powered │ N/A │ 0.20% │ -6 dB │ 1 │ 0 │ 2025-12-26 06:39:15 │ 1 hour ago │ ├─────┼────────────────────────────────┼───────────┼───────┼─────────────────────────┼──────────────────────────────────────────────┼─────────────┼────────────┼─────────────┼────────────┼───────────┼─────────────────┼────────────────┼────────┼────────┼───────────┼─────────────────────┼─────────────┤ │ 5 │ LSZ4 - Wiegleben 868 (Beam NW) │ !433e4da4 │ LSZ4 │ HELTEC_V3 │ y6Yuw521Ut3nqBZplbFDeVa6cGO9ZPyyNu8i8G0V/gc= │ N/A │ N/A │ N/A │ N/A │ N/A │ N/A │ N/A │ -6 dB │ 1 │ 0 │ 2025-12-26 05:44:17 │ 2 hours ago │ ├─────┼────────────────────────────────┼───────────┼───────┼─────────────────────────┼──────────────────────────────────────────────┼─────────────┼────────────┼─────────────┼────────────┼───────────┼─────────────────┼────────────────┼────────┼────────┼───────────┼─────────────────────┼─────────────┤ │ 6 │ ammera │ !433dab58 │ amrn │ HELTEC_V3 │ 02jzLn70aHkCHlWCRftnX8XJL33vQBomeTSPA24ACiI= │ N/A │ 51.2492° │ 10.4595° │ 219m │ Powered │ N/A │ 0.37% │ -14 dB │ 1 │ 0 │ 2025-12-24 04:40:38 │ 2 days ago │ ├─────┼────────────────────────────────┼───────────┼───────┼─────────────────────────┼──────────────────────────────────────────────┼─────────────┼────────────┼─────────────┼────────────┼───────────┼─────────────────┼────────────────┼────────┼────────┼───────────┼─────────────────────┼─────────────┤ │ 7 │ JensStadtMHL │ !b03d4958 │ DK1Z │ HELTEC_V3 │ 7ub7B5sb+Fh1u7OJct2BHbH0X05RdtlWfwhhCULqC2w= │ N/A │ 51.2110° │ 10.4602° │ N/A │ Powered │ N/A │ 0.30% │ -17 dB │ 1 │ 0 │ 2025-12-22 02:34:24 │ 4 days ago │ ├─────┼────────────────────────────────┼───────────┼───────┼─────────────────────────┼──────────────────────────────────────────────┼─────────────┼────────────┼─────────────┼────────────┼───────────┼─────────────────┼────────────────┼────────┼────────┼───────────┼─────────────────────┼─────────────┤ │ 8 │ CHtastic1 │ !9e780a84 │ CHT1 │ HELTEC_V3 │ Ls3quQNmByASd9WVDq1Gh7G2RWzMVpaN6GuHSi0SJj8= │ N/A │ N/A │ N/A │ N/A │ N/A │ N/A │ N/A │ 7 dB │ 0 │ 0 │ 2025-12-21 12:18:14 │ 4 days ago │ ├─────┼────────────────────────────────┼───────────┼───────┼─────────────────────────┼──────────────────────────────────────────────┼─────────────┼────────────┼─────────────┼────────────┼───────────┼─────────────────┼────────────────┼────────┼────────┼───────────┼─────────────────────┼─────────────┤ │ 9 │ CHtastic0 │ !9e757f80 │ CHT0 │ HELTEC_V3 │ HCMyl2NhO671WLX6MlXi/P2veagLbgGER7ynxSZ3tW4= │ N/A │ 51.2129° │ 10.4499° │ N/A │ Powered │ N/A │ 0.11% │ N/A │ N/A │ 0 │ N/A │ N/A │ ╘═════╧════════════════════════════════╧═══════════╧═══════╧═════════════════════════╧══════════════════════════════════════════════╧═════════════╧════════════╧═════════════╧════════════╧═══════════╧═════════════════╧════════════════╧════════╧════════╧═══════════╧═════════════════════╧═════════════╛ Regarding Meshcore, I also just saw one other remote node.\nThe Future My testing was just done with the tiny default antennas deployed below the roof, so these would be potential next steps:\nget better antennas. The current antennas are also covering all surrounding area equally, with a directional antenna one could focus into one direction. deployment on the roof. I have no pole up there right now, though. I see that the local HAM radio group has put up solid antennas. In Germany, HAM radio operators are operating a network over the air: the HAMnet. MeshCom is the approach to extend that to LoRa devices. The Meshtastic command line client has an option to tunnel IP over Meshtastic: works nicely, but of course comes with quite some latency. I have the idea of bringing back colourful BBSes over LoRa and have played with this, details in article LoRa Meshtastic BBS Testing.\nMeshmap can give you an idea how many Meshtastic devices are in your area, this site has the same for Meshcore.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-12-26T14:31:51+09:00","image":"/2025/12/starting-with-meshtastic-and-first-experiences/20251007_181251_lora_all.jpg","permalink":"/2025/12/starting-with-meshtastic-and-first-experiences/","title":"Starting with Meshtastic/Meshcore, and first experiences"},{"content":" Since spring of this year I\u0026rsquo;m learning to pick up morse code, so to hear sequences of dits and dahs/long and short sounds, and interpret the characters they represent.\nThat\u0026rsquo;s progressing.. but it\u0026rsquo;s just covering the part of interpreting received morse code.\nFor sending, you would want to connect a morse key or paddle to the computer - I\u0026rsquo;m describing here a simple way to do this. Not really rocket science, but capturing it here for others.\nTerms Morse code: a protocol to transmit characters as a sequence of two states, mostly as \u0026ldquo;short\u0026rdquo; or \u0026ldquo;longer\u0026rdquo; sounds. CW, \u0026ldquo;Continuous Wave\u0026rdquo;: another term for Morse code. dits/dahs: terms used for the \u0026ldquo;short\u0026rdquo; and the \u0026ldquo;long\u0026rdquo; sequences used in morse code. Morse code is a sequence is dits (signal for a short amount of time), dahs (signal over a longer amount of time) and blanks where no signal is sent. straight key or morse key: the device used to send out morse code. It\u0026rsquo;s producing a dit when pressed for a short time, and produces a dah when pressed for a longer time, ideally 3x the time of a dit. paddle is a device with 2 buttons. As long as the first button is pressed, \u0026lsquo;dits\u0026rsquo; are produced. As long as the second button is pressed, \u0026lsquo;dahs\u0026rsquo; are produced. Requirements I wanted to do this as simple and as cheap as possible:\na USB mouse: I got one from Hard Off in Japan, for the equivalent of 1 or 2€. The mouse keys have to be working. a straight key: I\u0026rsquo;m using the cheapest one I could find on aliexpress for below 10€. a cable with 3.5mm cinch headphone connectors on both ends, these are the normal plugs used for headphones. A cable like this came with the straight key I receved, it can be used to connect the key to radio transmitters solder iron, solder lead, solder flux/fluid circuit indicator (german: Durchgangsprüfer). I\u0026rsquo;m using a multimeter for this, but any simple loop with for example a power source and a lamp to indicate if the circuit is closed would also work Connecting I\u0026rsquo;m using DJ5CW\u0026rsquo;s page Learn CW Online for learning to listen to morse code, and it also has this section which interprets a pressed mouse button like a pressed morse key. We want this functionality with out morse key.\nAfter opening up the mouse, locate the 2 switches which get pressed as the mouse buttons are pressed. Here these were at a circuit board, on the back side of the board I saw multiple soldered contacts for each of the 2 switches. Using the circuit indicator, I evaluated which of these contacts were not shortened when no button was pressed, and which were shortened when the mouse button was pressed.\nNext, I cut the cable with the 3.5mm connectors into 2 pieces. Three cables appeared, with the circuit indicator I confirmed which 2 of these get shortened while pressing the morse key. I soldered these 2 cables onto the 2 contacts at the mouse circuit board.\nWith that, pressing the morse key is leading to the same result as pressing the left mouse button.\nI had also ordered the cheapest morse paddle. After dissecting it, it works in shorting one line to one of the 2 other lines, as one of the 2 paddles gets pressed - after more testing with the circuit checker I found the fitting third contact at the mouse board and soldered the third cable.\nAs container, I\u0026rsquo;m reusing the shell of the mouse with the scroll wheel removed and an additional opening for the second cable which is connected to the morse key or the paddle.\nI\u0026rsquo;m using this from Linux, but as this is simply an extended way to trigger what the mouse button does, it will work with all operating systems.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-12-25T16:40:13+09:00","image":"/post/2025/12/simple-connection-of-morse-key-to-computer/20251222_mouse_board_stripe.jpg","permalink":"/post/2025/12/simple-connection-of-morse-key-to-computer/","title":"Connecting a Morse Key to the computer"},{"content":"TL\u0026rsquo;DNR: I see warning signs everywhere around the topic. We need more honest discussion of AI, evaluation of pros and cons. For most areas, we should not yet invest our resources into GPUs, model training and inference.\nThese opinions and views are my own.\nResources AI is consuming resources, mostly\nfor production of the hardware for model training and inference, mostly GPUs for model training: for most LLMs, this is using up so much energy that just Google, Meta, IBM etc. can afford this for model inference, so \u0026ldquo;providing the AI infused service\u0026rdquo;, this uses up energy again Compared to the resources consumed in these areas, our brains are very efficient. Why not researching the technology more, and roll it out when we are sure the resources are taken into good use? We already run short on resources, and are fighting the climate crisis.\nI see AI right now in the same efficiency ballpark as \u0026ldquo;proof of work\u0026rdquo; based crypto currencies like bitcoin.\nFinancing Financing is done in loops. The companies buying GPUs from NVidia are in return getting backed by NVidia again. This is not a healthy loop.\nSociety AI has already an impact on society. I have experienced unthinkable behaviour: a person responsible for taking meeting notes of a regular confcall started to send out notes generated by Gemini. When major mistakes in the notes were pointed out, the person did not start to review the notes before sending, or create the notes again like in the past. Instead, a bug report to the vendor of the note-taking software was filed - and again major mistakes occurred.\nWe had a phase of getting used to mobile phones, i.e. not using them extensively when having dinner with others - we need to also find such rules for AI use.\nWe should also consider users looking at LLM\u0026rsquo;s as trusted advisors or friends: an LLM is biased in certain ways via the training, and that will reflect in it\u0026rsquo;s answers. LLM\u0026rsquo;s can even be maniciously manipulated to influence users in certain ways. I\u0026rsquo;m curious if users look as LLM created content as critically as content of random strangers.\nImpact on Knowledge and Science We all have seen the studies on how LLMs get trained on content from other LLMs, and the quality of the results decline. The internet gets flooded with AI generated content, distinguishing it from real content is hard. Some people use search flags which source only content which was published until 2023, so before the first wave of AI content. How to solve this decline in quality better?\nConsumption of only content which someone curated by hand, where it\u0026rsquo;s believed that the content is not from AI?\nLLMs: Communication vs. Thinking I see references like \u0026ldquo;LLMs understand xyz\u0026rdquo; and \u0026ldquo;LLMs are thinking [..]\u0026rdquo;, and I think these terms are misplaced here. LLMs get trained on big amounts of text, and text/language is what we use for communication. Thinking though is something on a higher layer, and language is just used for sharing thoughts.\nIdeas to improve the situation The most promising idea is to stop the current rollout of AI, and reevaluate pros and cons after further development. Proper studies should be done to account all the used resources, and we all should evaluate that against the use cases of AI. For example for medical use like diagnosing cancer this evaluation might return \u0026ldquo;useful\u0026rdquo;, and maybe for other uses not.\nI would also like to see:\nA proper, transparent accounting of the resources which single actions (i.e. a chat query to an LLM) are causing. That should be compared to the cost of queries to classic search engines, which can help with some queries, to a fraction of the resources. That will also allow us to consider whether we should really right now roll out AI to everything, or drive the technology to more efficiency some years and then roll it out. Disclosure: AI usage needs to be made crystal clear. Also, if your companies Human Resources department fears job candidates could use AI to cheat, and asks to not use AI, how about returning this and commit to not use AI on your side? If AI is used, someone needs to take full responsibility.\n\u0026ldquo;A computer can never be held accountable, therefore a computer must never make a management decision.\u0026rdquo; (IBM Training Manual, 1979) applies, and if someone uses AI somewhere, they need to take full responsibility. I was born in Germany, and related to the German history I was taught that \u0026ldquo;If we won\u0026rsquo;t do it, someone else will\u0026rdquo; is not a good excuse.\nAlso in my work life I have witnessed IT projects where colleagues jumped on with \u0026ldquo;this will look great in my CV\u0026rdquo;, but then the project did not run well, and they won\u0026rsquo;t mention it.\nThe biggest warning flag is that AI is hyped without even mentioning the down sides. If you want to promote something, then you should do that in all fairness, calling out also downsides. When that\u0026rsquo;s not happening - like with AI - then it\u0026rsquo;s triggering some of my defense mechanisms which also strike in when I see advertisements.\nImage source: reddit\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-12-17T04:01:00+09:00","image":"/post/2025/12/2025/12/thoughts-on-ai/ibm_accountable.jpg","permalink":"/post/2025/12/2025/12/thoughts-on-ai/","title":"Thoughts on AI"},{"content":"The Steam Deck is a hand held gaming console from Valve. Out of the box, it runs SteamOS, a Linux distribution from Valve. Main application is to run \u0026ldquo;steam\u0026rdquo;, a frontend to download and run games, Valve is operating an online store for games. Most of these were written for Windows, but 90% of these run also on the Steam Deck - also thanks to Valves investments into \u0026ldquo;Proton\u0026rdquo;, a layer for running Windows code on Linux.\nWith the Open Source Summit 2025 upcoming, I prepared the Steam Deck for driving the monitor at our booth at the conference. Capturing a few notes here regarding the setup.\nChoosing OS, installing The Steam Deck is x86-64 based, cpu/gpu from AMD. Thanks to Valve, the device has great Linux support, and drivers etc. went all into the upstream code bases. In the past I had run Fedora on the device: while running nicely for basic tasks, I hit these issues together with KDE/Plasma last year:\nkeyboard: the Steam Deck has a 1280x720 pixel touch screen. Mouse is emulated on a touch sensitive plate on the steam deck, but for keyboard input one has to rely on a visual keyboard provided by the window manager. With KDE, I think I had auto-login enabled, but later for operation calling up the visual keyboard was tricky. multi display setup: the Steam Deck connects via USB-C to a small dock, and from there via HDMI a monitor can be connected. Last year, the booth monitor and the dock just did not get together with a usable configuration. As Red Hat Enterprise Linux (RHEL) 10.1 was just released, I gave that a try - and it works flawlessly for these desktop purposes.\nI did not try auto login, but when logging in the user and entering a password, the visual keyboard got available automatically - great! So the only time I needed to connect a USB-keyboard was for the installation. Install: rhel image on a usb-stick with Ventoy. Ventoy creates partitions on the stick, and multiple images can then be stored on a FAT32 partition. When booting, you get to choose between the images. RHEL 10.1 installed just fine. Only mistake on my side: I intended to install on an NVMe enclosure with a 512GB NVMe, connected via USB to the Steam Deck. Later I got the devices wrong, and installed RHEL onto the internal NVMe. Anyway, nothing which can not be fixed later.\nRemote access How to conveniently operate the Steam Deck, when connected to the boooth monitor? Ideally from remote, a different laptop I would bring in. This would be via network.. but when deploying the Steam Deck, it would also not yet know Wifi credentials, so also getting the network connection is a topic. This is my chain:\nFirst network connection: ethernet. rj45 connected cable, between the dock of the Steam Deck, and the Macbook w/ Asahi Linux, which I brought in. Static IPs configured, so after logging in with a user at the Steam Deck gnome GDM, everything else can be done remotely. After booting also the Macbook, remote login via SSH to the Steam Deck is possible. Then both systems can be connected to the Wifi network via \u0026ldquo;nmcli dev wifi\u0026rdquo; and \u0026ldquo;nmcli dev wifi con (bssid) password (pass)\u0026rdquo;. Then, I connected the booth monitor. On the Steam Deck I had activated remote login, so from the Macbook with \u0026ldquo;remmina\u0026rdquo; as client, I could access the screen of the Steam Deck. After first login, the small internal Steam Deck screen is \u0026ldquo;primary monitor\u0026rdquo;. After switching that in gnome settings, and logging in again, I get to see exactly what the booth monitor is showing. Applications to run Time for actually showing something!\nOn the top right, I placed a terminal and did run \u0026ldquo;bb\u0026rdquo;, an old but great ASCII-art demo. Get an idea here: BB demo video. BB is really old, and by preferred way to run it is via container with Debian, where it\u0026rsquo;s still packaged. So after deploying a container with plain Debian, this is how the container gets built:\n$ mkdir build-trixie $ cat \u0026gt;build-trixie/Containerfile\u0026lt;\u0026lt;EOF FROM docker.io/library/debian:trixie ENV DEBIAN_FRONTEND noninteractive RUN apt update \u0026amp;\u0026amp; apt upgrade -y \u0026amp;\u0026amp; \\ apt install -y sudo systemd procps bb CMD [ \u0026#34;/lib/systemd/systemd\u0026#34; ] EOF $ podman build -t bb build-trixie/ $ podman run --name bb -d \\ --security-opt seccomp=unconfined --hostname bb \\ localhost/bb /lib/systemd/systemd $ podman exec -it bb bash container# /usr/games/bb With parameter \u0026ldquo;-loop\u0026rdquo;, it also goes into a continuous loop.\nThe Steam Deck has capable GPU hardware, and RHEL supports these nicely via Mesa and the kernel. I showed this in the bottom/right area of the screen, in opening https://phoboslab.org/wipeout/ with firefox. This is a WebGL based demo with the data from the Playstation 1 game \u0026ldquo;Wipeout\u0026rdquo;. Runs smoothly with the Deck!\nAt the top left I planned to show slides of the recent Webinar I gave, around topics \u0026ldquo;RHEL 10 news, AI and bootc\u0026rdquo;. The slides are of course stored as pdf file. I did not find an easily available pdf slide viewer, so let\u0026rsquo;s use ImageMagick to convert the slides into single images, and have that shown as slide show:\nwhile :; do for i in {0..33}; do (loupe page_$i.png \u0026amp; ); sleep 15; killall loupe; done; done I had tried all of this already when preparing at the office, and it worked flawlessly at the conference. Great to see RHEL running so nicely also for desktop use on this not-quite-enterprise use case!\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-12-09T07:55:32+09:00","image":"/post/2025/12/rhel10-on-steamdeck-for-booth-operation/20251208_092423_deck_setup_booth.jpg","permalink":"/post/2025/12/rhel10-on-steamdeck-for-booth-operation/","title":"RHEL on the Steam Deck for Booth Operation"},{"content":"Terms Details on some terms I will use:\nLoRa (\u0026ldquo;Long Range\u0026rdquo;) is a technology for transmitting data via radio signals. It\u0026rsquo;s designed for much longer distances than Wifi: the record for LoRa is around 300km, and typical distance 10km. Downsides: less throughput than Wifi, and higher latency. LoRa defines various profiles, these allow to fine tune latency/ throughput versus distance of the participating stations. meshtastic is an OpenSource firmware, designed to run on a variety of devices. Meshtastic implements a mesh network, so even if participating nodes can not reach each other directly via radio, they can exchange data via nodes in the middle. BBS (\u0026ldquo;Bulletin board system\u0026rdquo;) are a kind of software. In the early 1980s, a BBS software did run on a system, and other computers or terminals could dial into this system and interact with the software. The software allows data exchange between multiple users, they can leave messages for each other, or data like pictures. Messages could also be exchanged with other BBSs. With the internet, the \u0026ldquo;always on\u0026rdquo; style appeared. Some BBSs are still run, and can be accesed via dialup or over the internet. The idea of BBS over LoRa Since some time, you can get a small LoRa device with antenna for $30. Such a device can have an own keyboard, and you can exchange messages with similiar devices within your range. This requires that the node sending and the one receiving the message are online at the same time.\nQuickly also simple mailbox software appeared: here you have one central node running constantly. It can store messages which are sent in from other nodes, and hand out the message to the recipient.\nThe idea: how about a revival of the colourful BBS software we had in the past? Basic functionality was also message exchange, but then also file exchange, simple games and more! So we are again limited with throughput here, but we would be using a medium completely independent from the internet!\nMy main question here: is the required software around? Are throughput and latency of LoRa sufficient for this? Spoiler: not sure. We are in the right ballpark, throughput/latency are almost usable, with tuning this might work. For my testing I am using the plain access via LoRa, but have not yet hosted the BBS software myself, and the software setup is a bit complicated.\nThe hardware Meshtastic supports quite some hardware, I was going with Heltec LoRa32 V3 devices. These are based on the ESP32 chip, have no own keyboard but allow connections via USB-C, bluetooth an Wifi. In Germany, mostly the 868MHz band is used for LoRa, but also 433MHz. Here in Japan, the 923MHz area is used. I did ordered:\n2x devices for 868MHz 2x devices for 923MHz 1x device for 433MHz I had not done my research properly: turns out one device type can use both areas, 868+923MHz. Just the 433MHz device is different.\nInitial Setup The vendor recommends to install Arduino software, then test and program the devices - I did not follow that but directly installed meshtastic. I did this on a Fedora 42 as well as a Debian 12 system, steps on Debian, to be executed as user:\nsudo apt update sudo apt install python3 python3 --version pip3 --version The user should then be added to the group \u0026lsquo;dialout\u0026rsquo;, to get access to the Heltec. Then I connect the Heltec via USB-C to the system, and continue:\npip3 install --upgrade esptool # This does a first communication esptool chip-id # get firmware from https://meshtastic.org/downloads/ ./device-install.sh -f firmware-heltec-v3-2.6.11.60ec05e.bin pip3 install --upgrade pytap2 pip3 install --upgrade meshtastic With the next step, we configure the Heltec for our region, for me Japan. Let\u0026rsquo;s then query the current settings:\nmeshtastic --set lora.region JP meshtastic --get lora meshtastic --get lora.region meshtastic --get 0 If we did all of this also on the other Heltec, we can now query for peers, and send messages. For message transmission, one node should be listening via \u0026ldquo;meshtastic \u0026ndash;reply\u0026rdquo; and the Wother node should send via \u0026ldquo;\u0026ndash;sendtext\u0026rdquo;:\nmeshtastic --nodes meshtastic --sendtext \u0026#39;Hello Mesh!\u0026#39; meshtastic --reply With this, message exchange should work! Of course, instread of using USB-C and the command line, the Heltec can also be controlled via Wifi or Bluetooth.\nFor connexting to the device with the Android mestastic app via Bluetooth, let\u0026rsquo;s also set the pin:\nmeshtastic --set bluetooth.fixed_pin \u0026lt;some-pin-here\u0026gt; meshtastic --set bluetooth.mode FIXED_PIN The BBS test So will colourful BBSen work?\nFor the test, I did not host the BBS software myself, but connected to a BBS over the internet. My test uses these layers:\nIP network over the LoRA connection, this gives me an IP for each node, and they can reach each other. Login via SSH, over the IP-over-LoRA connection. Once logged into the remote system, I access the BBS over the internet with telnet, i.e. \u0026rsquo;telnet ip port\u0026rsquo;. Various other combinations are also possible, for example instead of the SSH login the traffic tto the BBS system could be routet over the link.\nThe IP-over-LoRa network setup Let\u0026rsquo;s do a step back and consider the network setup. This is new in meshtastic, and requires to be initiated by the root user. My steps:\npython3 -m venv meshtastic export PATH=/root/meshtastic/bin/:$PATH pip3 install --upgrade esptool pip3 install --upgrade pytap2 pip3 install --upgrade meshtastic meshtastic --tunnel With these steps executed on both nodes, meshtastic outputs the IPs of both nodes. We see in \u0026lsquo;ip link show\u0026rsquo; that an interface mesh0 appears, and we can reach the IPs over mesh0. First test was of course \u0026lsquo;ping remote-ip\u0026rsquo;, and the round trip time turned out to be 20-30 seconds. This would be way to much for convenient access of a classic BBS.\nBut LoRa supports \u0026lsquo;profiles\u0026rsquo;, and after executing\nmeshtastic --set lora.modem_preset SHORT_FAST on both nodes, we get ~2sec round trip time. So after that accessing the remote BBS is somehow barable.\nThe Future There might be optimizations to improve this - but after all we are using here networking designed for IoT to send sensor data over big distances. The fact it works like this is already surprising. Will this catch on? Unlikely.. after all, the frequencies are a shared medium! It\u0026rsquo;s not like a dedicated phone line directly to the BBS.\nBesides meshtestic, also further software is appearing for networking over LoRa, for example lorapipe. Also the full stack could be changed. We actually do not need full networking.. just a serial terminal over LoRa would be enough to make a BBS available.\nReticulum Network Stack looks also promising: it comes apparently with a lean remote shell.\nThe Wiki has now a page on Meshtastic.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-10-14T20:30:41+09:00","image":"/posts/2025/10/lora-meshtastic-bbs-testing/20251007_181251_lora_all.jpg","permalink":"/posts/2025/10/lora-meshtastic-bbs-testing/","title":"Lora Meshtastic BBS Testing"},{"content":"In 2008 I was in Japan for 3 month. After returning to Germany, I thought it might be fun to go to Japan again for holiday, and being able to speak a bit - that started it all.\nMy language learning history I grew up in East Germany, so learning German. In Socialist East, I was part of the last school year of pupils learning Russian as first foreign language, the language of the east block. I think it was not taught in a good way, not with focus on actively speaking and listening. We could put it into practical use though: we had pen pals in the Soviet Union, and dads of other students worked in Russia at the Druzhba pipeline. With the German unification in 1989, most students gave up on Russian learning and started with English. Me too - but just with English as minor study field, and focusing on Math and Physics. The way of teaching English was very good, by scales. The internet was still not there though, there was not much English content which could have sparked our interest outside of school.\nFirst Japanese lesson After my stay in Japan I decided to try learning Japanese, living in a big city like Munich made it easier to find a school class for that. In the first days of January 2009 I went to a class at the citizens university (Volkshochschule), and met a Japanese teacher to whom I\u0026rsquo;m in contact until this very day.\nI will never forget the opening words: \u0026ldquo;So do you think you will be able to speak Japanese after these 5 days?\u0026rdquo; Well, not really.. but the question really made me think about this. Spoiler: even after the 17 year I don\u0026rsquo;t feel like I\u0026rsquo;m speaking it..\nThe teachers style was unusual: we started in listening to him saying full Japanese sentences, along with the German translation. We repeated these until our pronunciation was right. He handed out CD\u0026rsquo;s with many hours of the sentences in Japanese and German, for training. We also talked about the meaning of single words, but I think comprehension appeared just from listening many of these sentences, and \u0026ldquo;letting the brain figure it out\u0026rdquo;. We would not talk about grammar at that time.\nAlong with speaking and listening, we learned to write and read the Hiragana and Katakana alphabets: having learned these characters like ま (\u0026ldquo;ma\u0026rdquo;), む(\u0026ldquo;mu\u0026rdquo;), か (\u0026ldquo;ka\u0026rdquo;) or る(\u0026ldquo;ru\u0026rdquo;) we were able to write down all Japanese sentences we heard. That was also a strict rule: we write Japanese in Hiragana/Katakana or Kanji, not our Arabic letters. That helps the spelling: writing Japanese with Arabic letters leaves much room to guess the correct spelling, but with written down Hiragana/Katakana the spelling is pretty clear.\nI did horrible, btw. My written characters were \u0026ldquo;not clean\u0026rdquo;, and I was bad at memorizing vocab, and bad at correct spelling. We got the CD\u0026rsquo;s for training, but as we also went for beers after the lessons, there was barely time for training.\nGoing to regular lessons I made it a habit of listening to the CD\u0026rsquo;s from the teacher, at any free minute: the Nokia communicator smartphone had a good loud speaker, and already after getting up in the morning, I started to listen to the Japanese sentences. When I did not just have the toothbrush in my mouth, I would repeat them. The next sentence was then the German translation. This resembles how children learn - and they do not even have the translation in a language which they understand.\nI registered for weekly lessons with the same teacher. While commuting by bike to work I would listen to the Japanese CD\u0026rsquo;s with one one ear, and repeat the Japanese. When going to the lessons on Saturday afternoon, I would do the training already while on the train. I would get off at train station \u0026ldquo;Laim\u0026rdquo;, and walk 2 hours trough the city to the citizens school - doing Japanese training all the time. Being subscribed to one lesson of the teacher, we were also free to join to his other lessons over the week - I ended up doing 2 lessons on Saturday and one on Sunday.\nI still think that for spelling, starting with real lessons is best.\nMy online ressources over time Right from start learned vocabulary online, via website https://iknow.jp. There are now many sites following the same idea: they introduce new Japanese vocabulary, ideally also in conjunction with a whole sentence. Once the user has memorized the vocab, the site recognizes that and will after increasing timespans bring up the vocab again. I still like the site very much, catching up with it\u0026rsquo;s vocab training is part of my daily morning routine. Just the Japanese content is limited, after learning that I\u0026rsquo;m just retaining on the site. Fun stats: I\u0026rsquo;ve studied for 1048 hours, so that\u0026rsquo;s 43 days of learning!\nThen I started memrise.com, and also learning there daily for 10-15min - not just retaining but also new vocab. I\u0026rsquo;ve paid lifetime fee on both sites. Both have apps for Android, but I\u0026rsquo;m just using the browser interface.\nMy second important online resource: a page where you write texts in a foreign language, and get it corrected. The page introducing the idea was lang-8.com, it has vanished and https://nyan-8.com/ is an excellent replacement. Example of one of me texts there: link. I recommend the approach for multiple reasons:\nYou write texts in foreign language, and get corrections by natives. You can use that for improving your writing. You also get comments on the contents, this has helped me many times to get impressions and alternate views on matters of daily life in Japan. Writing the text helps me to sort my ideas around a topic. In computer area we know this as \u0026lsquo;rubber duck explaining\u0026rsquo;: explaining a problem or concept to a colleague or a rubber duck helps us to grasp the problem. I\u0026rsquo;ve made great friends via these 2 sites. Having frequent video chats via internet with 2 of them. As a further side affect, this has also brought me into writing blog articles, see blog.fluxcoil.net. I\u0026rsquo;ve also taken over many of my texts from lang-8/nyan-8 into my wiki.\nWhat about writing Japanese? Obviously, everybody is setting their own goals. For me, just writing Japanese on the computer and the mobile phone are important. On these platforms, there are also various tools available. As for reading, I mostly read texts on the computer, and have single words translated into English, via browser plugins.\nRight now, I\u0026rsquo;m just writing Japanese by hand for xmas cards I\u0026rsquo;m sending out, and for training sheets from one of my Japanese teachers. Not writing Kanji by hand is also impacting reading capability though: kids in Japan have a collection of important Kanji, they write each of them many times. This is training muscle memory, and along with your writing you also improve the capability to tell the Kanji apart from each other. I would like to improve my reading without tools like browser plugins, but I was never able to force myself into these long Kanji writing sessions.\nOn the other hand, writing capability is also degrading with Japanese people - everybody is just using the computer and mobile phone these days.\nMotivation over time That\u0026rsquo;s the big topic. Without finding joy, you won\u0026rsquo;t stick with it. For some, obtaining badges in the language learning app is enough. I found the curve of progress important for my motivation: initially, you progress very fast. For myself, I see these mile stones:\nafter a few weeks of learning: you can order a meal or a beer. And memorize a simple self introduction. after months, writing your name in the other language later, conversation about simple topics Then, at some point, you will be in a longer conversation where the language you are learning is the most efficient means of communication. For me, this was 2012, after 3 years of learning, at a homestay with a family in Sapporo. Then, giving a presentation in business context in the foreign language. Later, picking up a hobby not related to language - and doing the learning and exams in Japanese. I passed the Japan HAM radio license, level 4, in 2024. These mile stones can be different for everybody. Also passing of tests can play a role here, for Japanese that\u0026rsquo;s the Japanese Language Proficiency Test (JLPT, 日本語能力試験). It has 5 levels, 5 being the easiest. Many jobs in Japan require level 2 or 1. The goals of such tests do not necessarily align with the goals of the language learner: my goal is to improve speaking/listening. Even when passing the highest JLPT level, one has not spoken a single word though: the test is focusing mostly on quick reading and understanding texts, then answering questions. Also grammar questions, and listening are important.\nWhile I was in Germany, I passed levels 5, 4 and 3. For getting my job in Japan I had to also pass a kind of company own test, to work on a technical support issue in Japanese. Later in Japan I passed level 2, and was failing the level 1 exam. Not planning at trying it again, after I know it\u0026rsquo;s focused on fast reading.\nWhat\u0026rsquo;s also motivating?\nBeing able to read texts in the foreign language, and discovering pieces/texts which I like so much that I say \u0026ldquo;Being able to read this natively is totally worth leaning the language\u0026rdquo;. For me, one example is 雨ニモマケズ from 宮沢賢治, details in English are here Another example: 第三夜 from 夏目漱石. Being able to talk with people in the foreign language. Understanding the language enough to understand the differences to the mother tongue: for example, Japanese has multiple words for \u0026ldquo;to be\u0026rdquo;, depending on whether the object is living (i.e. a pet) or not living, i.e. a table. Simple enough, eh? Until you realize that Japanese sentences about gods are using the word for \u0026ldquo;living\u0026rdquo;. And the word used for robots changed recently, they are now more seen as \u0026ldquo;living\u0026rdquo;! Diving into the foreign culture. Japanese is so different, that you do not only learn about the many ways to show with words that the person you talk with \u0026ldquo;has a higher status\u0026rdquo;, you also learn about the cultural background. And you question how this is done in your own language. You discover movies in the foreign language. And books, but these need to be at the exact right level to already be readable without using to many tools, but als being interesting enough to get you hooked. On the wiki I have references to more approaches. Find something which works for you, keeps you hooked. You won\u0026rsquo;t invest all the time without being interested.\nWhat has language learning done for me? The most obvious part, it enabled me to communicate to some level in Japanese. Along the learning journey, I have met many folks, with many I\u0026rsquo;m staying in contact: teachers, other students. I\u0026rsquo;ve been travelling Japan by myself, as well as with teacher and other students, had dinner with folks I would never have met by myself. I was at a really great home stay in Sapporo, this was my first \u0026ldquo;actually my Japanese is the best means of communications\u0026rdquo;-experience.\nI got to like Japanese culture and Kanji, brought Japanese into my life and together with my other hobbies: text setting with Kanji, in LaTeX, works nicely. The software for Japanese input is an area by itself, and learning about Unicode and Japanese fonts is great.\nWhat are the basics of good language learning? That\u0026rsquo;s a subjective list of course, what works for you might not for for someone else. I myself think this is good language learning:\nMost importantly: it needs to be fun. Look at various ways of learning and consuming content in the other language. When you find a combination which works for you, stay evaluating to the left and to the right. New things get available over time, and also your level of proficiency changes and you will adopt different methods. Think I recognized 2 overall categories of learning strategies: a) to lern vocab and grammar like formula, and then construct sentences like a mathematician. Might work for some people, and be well suited for some languages, but I do not recommend it. It\u0026rsquo;s also like my Russian lessons were done, and English in Japanese schools. b) is rather giving the learner much content, and letting the brain \u0026ldquo;figure it out/recognize the patterns\u0026rdquo;. That\u0026rsquo;s how we pick up our mother tongue, and this works better for me - even though I\u0026rsquo;m all pro mathematics and science in all areas. :) Sticking to plain vocab training and apps will not work out in the long run. Get also a real course, or a tandem partner in your town or over the internet. Only this can help with pronounciation. Becoming proficient in the language, or passing a certain test, can be a goal for you - but it\u0026rsquo;s also totally fine to just like to learn without getting anywhere. I was told by a professional musician that \u0026ldquo;starting now with an instrument will not lead anywhere\u0026rdquo; - but actually me enjoying to make music might be more important than anybody enjoying to listen. Immersion helps a lot, Consume news/drama/movies/content in the language you are learning. It\u0026rsquo;s not a silver bullet though: passively listening to news in Japanese without focusing on them will not help, the brain is just filtering them out. My current strategy I\u0026rsquo;m in Japan since 2016, using Japanese and English at work. My Japanese will never be at the level of my colleagues, but on the other side I feel more comfortable to speak English.\nIn the morning I do vocab training with iknow.jp and memrise Weekly lessons with a Japanese teacher, can really recommend nihongo-lesson.jp. Quite cheap, private lesson with 1 teacher for you as student, so the contents can be tailored exactly to your level.\n2000円/40min lesson. It\u0026rsquo;s apparently so cheap my employer can\u0026rsquo;t believe it and is not paying for it. :) Weekly video chat sessions with a friend who is learning German. Texts on nyan-8 Reading tech news from gigazine.net podcast バイリンガルニュース So how would you spend these 1000 hours? :)\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-10-03T08:18:16+09:00","image":"/posts/2025/10/17-years-of-japanese-learning/temple_poet.jpg","permalink":"/posts/2025/10/17-years-of-japanese-learning/","title":"Coming up: 17th year of Japanese Learning"},{"content":"I presented this topic at TIARA (with emphasis at Linux/the software), and at the Tokyo Linux Users Group, with emphasis at the radio concepts and how they translate to software. The talk was recorded, and the slides are here.\nTL\u0026rsquo;DNR I will describe a setup with Software Defined Radio (SDR) receiver, computer and OpenWebRX+ software. To get an idea what accessing the OpenWebRX+ software is like, try out some stations listed on https://www.receiverbook.de. These are all installations with SDR devices, made available to the public.\nExample what receiving and decoding FT8 signals looks like. FT8 is a digital mode, optimized for transmission also under bad conditions. On the left side, received signals get decoded.\nRecommended Hardware A laptop: I\u0026rsquo;m using a Thinkpad L480. It\u0026rsquo;s overpowered with 16GB of RAM, also 4GB are enough. The laptop runs Debian Linux, OpenWebRX+ and Firefox as webbrowser. Firefox could also be run remotely, from anywhere over the LAN or the internet. SDR device: These get connected to an antenna, and fullfill some functions of classical radios. They make the signal from the antenna available to the computer. Some SDR\u0026rsquo;s can also transmit signals, but our OpenWebRX+ software can not be used for sending. I\u0026rsquo;m using the RTL-SDRv4 device, available of 60-80$ including a simple antenna. Alternatively I use a HackRF clone for 130$. That device can also transmit, and can receive a wider bandwidth than the RTL-SDR. Antenna: I use a simple telescope antenna, bought as set together with the RTL-SDR. The antenna is connected via coax cable to the SDR device, which is connected to the laptop via USB2. The longer the coax cable, the bigger the loss of the signal. In Germany, I deployed the antenna on the roof, and SDR-device/laptop on the attic. From everywhere on the house I could then connect to the laptop via network cable.\nWhat\u0026rsquo;s possible with this setup? Setup OpenWebRX+ on a system, connect an SDR device, then\nrun a web browser (Firefox, Google Chrome etc.) directly on the system OR run the web browser somewhere else and access via network! Then you can listen in on all bands supported by the SDR device, decode signals in many modes: (near)FM, WideFM, SSB, AM. Also many high level decoders available: ADSB (air plane signals), FT8 (also showing contacts on a map), CW (morse code) and much more.\nHow to setup? System setup\nget Debian Bookworm for AMD64 or aarch64 flash it to a USB-stick, install Debian Software\nAdd the OpenWebRX+ repos, details on the OpenWebRX+ site Install OpenWebRX+: apt update; apt install openwebrx If you use RTL-SDR, install the drivers: sudo apt-get install libusb-1.0-0-dev git cmake pkg-config git clone https://github.com/rtlsdrblog/rtl-sdr-blog cd rtl-sdr-blog mkdir build cd build cmake ../ -DINSTALL_UDEV_RULES=ON make sudo make install sudo cp ../rtl-sdr.rules /etc/udev/rules.d/ sudo ldconfig echo \u0026#39;blacklist dvb_usb_rtl28xxu\u0026#39; | sudo tee --append /etc/modprobe.d/blacklist-dvb_usb_rtl28xxu.conf Then start OpenWebRX+: systemctl enable \u0026ndash;now openwebrx Now access http://127.0.0.1:8073 or http://:8073 from a browser This map shows from where we received FT8 signals:\nPictures of the setup:\nI like this setup very much!\nThe slides of the presentation at TIARA are here.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-09-19T22:12:27+09:00","image":"/posts/2025/09/debian-with-rtl-sdr-and-openwebrx-setup/20240814_20m_ft8mrecv-small.jpg","permalink":"/posts/2025/09/debian-with-rtl-sdr-and-openwebrx-setup/","title":"Setting up Debian with RTL-SDR and OpenWebRX+"},{"content":"TL\u0026rsquo;DNR\nIn this article I will look into the details of a special case with fully emulated systems, where an older Linux distro performs better than a new one. I will use RHEL7 and RHEL10 here, but the difference can also be seen with other distros. This is limited to the not very common case of the system running as a fully emulated environment, for example with QEMU. Most environments run virtualized guests instead.\nI was presenting about this topic, the recording is here.\nTerms I will use the terms like this:\nVirtualization in our context here means that when a guest operating system does computations, these get directly offloaded to the hypervisors cpus. One of the benefits is performance, one of the downsides is that the guests cpus need to exactly match the hypervisors cpus, or a subset. Emulation means that the computations are done in software. With this, host/guest architectures can be different. As a downside, this is slower than virtualization. Background Why would one run emulation instead of virtualization? I use an aarch64 based laptop for work. To replicate issues of x86 systems I use full system emulation. This research helped me to understand speed differences between the various x86 cpu/chipsets which can be emulated, and I made an observation:\nThe command for emulating a guest:\n$ qemu-system-x86_64 -cpu Broadwell-v1 -smp cpus=4 -m 4G \\ -drive file=image-v3.qcow2,if=none,id=hd0 \\ -device virtio-blk-pci,drive=hd0 \\ -netdev bridge,br=virbr0,id=n1 -device virtio-net,netdev=n1 \\ -object rng-random,filename=/dev/urandom,id=viornd0 \\ -device virtio-rng-pci,rng=viornd0 \\ -boot order=c Some of the x86 chipsets which QEMU can emulate, 147 right now:\nThe initial observation I was curious if all of these emulated systems perform same, or some are faster, so better suited for my uses. I did run CPU bound workloads in the guest, visualized them. On the following graph, starting from the left, we see a single thread doing \u0026ldquo;file uncompress jobs\u0026rdquo; for 10minutes.\n$ while true; do bzcat file.bz2 \u0026gt;/dev/null COUNTER=$((COUNTER+1)) done Running 2 threads in parallel, the number of extract jobs increases.\nEven more for 4 threads. These are systems with 4 emulated CPUs, so with 8 threads we see the number of extraction jobs declining a bit. On this graph we can see RHEL7 doing better than RHEL10:\nWhy is that? The issue si due to a side affect of the distros getting optimized to newer CPUs. Let\u0026rsquo;s consider RHEL7: released in 2014, it\u0026rsquo;s built for the x86-64 architecture. RHEL10 was released in 2025, and has glibc, one of it\u0026rsquo;s core libraries, built in a way to require CPUs of the x86-64-v3 architecture. In these 11 years, various new CPU generations have been released. Among other changes the new CPUs also got new features, so more efficient ways of doing certain computations. These are not used \u0026ldquo;automatically\u0026rdquo;, but code needs to explicitly utilize the features.\nWhen emulating a full system with QEMU, one gets to choose which cpu features should be emulated - the recommendation is to configure as many features as possible. Our 2 VMs here are both configured with \u0026ldquo;-cpu Broadwell-v1\u0026rdquo;: that\u0026rsquo;s a CPU following the x86-64-v3 spec. RHEL7 is not utilizing the additional features over plain \u0026ldquo;x86-64\u0026rdquo;, but as RHEL10\u0026rsquo;s glibc (and thus whole RHEL10) are optimized for -v3 we really need that.\nThe extensions of your x86 CPU can be verified like this:\nThe decision makes totally sense: systems which implement -v3 and higher are available since many years, so we should utilize these features. Florian Weimer has written on this.\nNow back to our emulated systems: both support -v3, but just on the emulated RHEL10 it\u0026rsquo;s getting used. RHEL7 is getting along with fewer CPU functionality, and for this whole emulated stack that turns out more performance efficient than RHEL10 where we offload things to the CPU which gets also just emulated on the host system. Qemu has a nice overview of which chipsets has which feature set.\nOf course, a niche and rare situation.. but we have here no choice but emulating the whole system. :)\nFascinating to see.. and took me a ticket on qemu-devel, and explanations from colleagues to really understand this.\nWhat do debug tools see? So now with the explanation at hand, our overall expectation is that RHEL7 does more workload in it\u0026rsquo;s userland, and for RHEL10 we are doing more in QEMU emulating these extra registers.\nFor the curious, here is a text file with\nthe details of the single uncompress runs output from \u0026ldquo;strace -c bzcat [..]\u0026rdquo; in the guests, and from \u0026ldquo;strace -c -p pidofqemu\u0026rdquo; watching from the host while data is extracted Not sure if one can see from QEMU which cpu features are triggered, but with perf inside the guest one would be able to see that.\nI tried to look at the assembler code which the workload is executing in the guest, with objdump. This brings up the asm commands for bzcat and also glibc, I compared these between rhel7 and rhel10, but did not really get anywhere. Example:\n$ objdump -d /usr/bin/bzcat /usr/bin/bzcat: file format elf64-x86-64 Disassembly of section .init: 0000000000001000 \u0026lt;.init\u0026gt;: 1000: f3 0f 1e fa endbr64 1004: 48 83 ec 08 sub $0x8,%rsp 1008: 48 8b 05 b9 7f 00 00 mov 0x7fb9(%rip),%rax # 8fc8 \u0026lt;__ctype_b_loc@plt+0x79a8\u0026gt; 100f: 48 85 c0 test %rax,%rax 1012: 74 02 je 1016 \u0026lt;__strcat_chk@plt-0x31a\u0026gt; 1014: ff d0 call *%rax 1016: 48 83 c4 08 add $0x8,%rsp 101a: c3 ret [..] Something else which might be interesting: it should be possible to have QEMU output which asm-code the guest is executing.\nContainers to the rescue As final confirmation of the above theory, we can run the workload in a RHEL7 container inside the RHEL10 guest: this will utilize the containers glibc, so do computations in the way which is more effective in our situation.\nWorks indeed:\n[rhel10]$ podman login registry.redhat.io [rhel10]$ podman pull registry.redhat.io/ubi7/ubi:7.9-1445 Trying to pull registry.redhat.io/ubi7/ubi:7.9-1445... Getting image source signatures [..] [rhel10] podman run -it registry.redhat.io/ubi7/ubi:7.9-1445 bash [root@dae7ac8c8859 /]# [root@dae7ac8c8859 ~]# for i in {1..5}; do time bzcat randfile.bz2 \u0026gt;/dev/null; done real\t3m50.920s user\t3m24.340s sys\t0m12.341s [..] Average over 5 runs: 249sec, quite near to the 223sec from plain RHEL7!\nClosing To repeat: this is an exotic situation to be in, but it\u0026rsquo;s very satisfying to have understood the situation. Interestingly, AlmaLinux10 for x86 is available as a -v3 build (like RHEL10), and also as -v2 for compatability also with older CPUs. I\u0026rsquo;ll compare the download stats in one or 2 years, curious if there is much demand for -v2.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-06-27T20:14:55+09:00","image":"/posts/2025/06/the-riddle-of-rhel7-being-faster-than-rhel10/gnuplot_cpu-performance-emulated-only2.png","permalink":"/posts/2025/06/the-riddle-of-rhel7-being-faster-than-rhel10/","title":"Investigating a RHEL7 vs. RHEL10 speed difference"},{"content":"What\u0026rsquo;s Morse Code? Morse Code, also known as Continuous Wave (CW), is one of the oldest ways of electronic communication. You might also know the term from movies, when in emergency situations someone encodes Morse Code in hitting a pipe with a hammer, and someone else picks this up in listening to the sound transmitted by the pipe.\n~40 characters get encoded into sequences of\nshort dots, also called \u0026ldquo;dits\u0026rdquo; and longer dashes, also called \u0026ldquo;dahs\u0026rdquo; For sending out morse code, one has to translate sequences of words/numbers into the matching sequences of dits and dahs. The length of the dahs is defined as \u0026ldquo;three times the length of a dit\u0026rdquo;. Also the spaces between the characters and between the words are defined, allowing the other side to pick up the transmission.\nThe famous SOS gets encoded as \u0026ldquo;\u0026hellip; - - - \u0026hellip;\u0026rdquo; .\nThis is what CW sounds like.\nWhy learn Morse? 70 years ago, the HAM radio exams all over the world required to also pass a CW test, but that\u0026rsquo;s dropped now mostly. Hundred years ago, CW was the most efficient way to communicate over the air. A simple sound gets either modulated, or not. Short time of modulation is a \u0026ldquo;dit\u0026rdquo;, longer modulation a \u0026ldquo;dah\u0026rdquo;. Transmitters for CW can be much simpler than transmitters for voice modulation, and with high noise levels, CW is more easy to pick up than voice.\nThat said, even if CW can be heard more easy than voice (so \u0026ldquo;is there a signal or not?\u0026rdquo;), learning it is quite hard. Nowadays, digital modes like FT8 where for example a computer is putting a signal onto the air, are even more efficient than CW. So they get the communication through under conditions when CW is no longer possible.\nSo why would one want to learn it? Why do people learn music instruments, or paint pictures? To obtain a skill, to have fun, to potentially entertain others. No matter how bad you are at playing an instrument, it might still spark joy in you - then it\u0026rsquo;s worth doing.\nIn HAM radio, CW is heavily used. There are contest activities like \u0026ldquo;Summits On The Air\u0026rdquo; (SOTA) where HAM\u0026rsquo;s go to mountains, setup an antenna and a transmitter, and make connections to other stations. Points are awarded based on the number of confirmed connections. CW connections get here more points, the CW transmitters are also smaller, easier to handle. Activities like SOTA are a further argument to learn CW.\nSo, let\u0026rsquo;s give it a try.\nCW learning There are 2 parts to CW learning:\nthe picking up, also called \u0026ldquo;to read CW\u0026rdquo; and to send So far, I tried to learn 4 languages in my life, and use 3 of them daily. I used various methods for learning, so the first step to learn CW was to research \u0026ldquo;ways to learn\u0026rdquo;.\nOne would think: let\u0026rsquo;s start like children learning the alphabet or Kanji in school: introduce one after the other, have them memorize, have them read and write. This works also for CW, in the same way you can memorize the alphabet or the Korean Hangul after a short time - or after a longer time Kanji.\nYou can then read or write the letters as dits and dahs, or listen to a very slow Morse Code. The twist: in the real world you need to aim at communication of 20 words per minute, with each word having an average of 5 characters.\nThe Long Island CW Club is picking up this approach to start slow, and then increase the speed. Yet, as the speed increases, the characters will morph into different sounds - and many think that achieving high speeds is hard with having started slowly.\nAn often quoted study on learning CW is from Ludwig Koch. Luckily the study is in German, I read it natively - not claiming my conclusions are all correct, of course. LICW has an English summary. Following the results from the Koch study, one would right from the start learn with 20 words per minute: first one gets introduced to the Morse expressions of 2 characters. Then one gets for 60 seconds sequences of these 2 characters, the goal is to pick them up correctly and note them down. Learn CW online uses the method, I started on the page. You can write down the characters on paper and later fill them into the website text field, or directly write into the website - that\u0026rsquo;s what I do. After listening the Morse Code for ~60 seconds, and writing the sequences you think you picked up, you click \u0026ldquo;check\u0026rdquo; and then see how well you picked up the code. If you feel unsure, you can also listen to the same ~60 seconds sequence multiple times, and make corrections to the characters you noted down. The goal is to pick up 90% or more of the characters correctly after listening to the sequence just a single time. When that\u0026rsquo;s accomplished, a further character is added.\nSounds hard? It is! People with a talent for music seem to do better, pick up things easier. Now after 2.5 months I\u0026rsquo;m doing the lesson with 7 different characters. 40 in total\u0026hellip;\nSide quest: speed typing Starting learning with the Koch method, I did run into a further issue: so far I was reasonably fast with 2-finger-typing. But already when learning the 4th character with the Koch method, I entered the side quest of learning 10 finger typing.\nThat\u0026rsquo;s actually going quite well after 2 weeks of training on typingstudy.\nMore thoughts One interesting point: the sequences on Learn CW online are just random, not actual words. With 7 characters, you can build already some actual words. But then: would it make more sense to learn with words? When reading English or German, after a while you start to recognize all the letters of a word all together, instead of letter-by-letter. Same in Japanese. Also with CW, it is said that one builds up recognition of full words after some time. Needs much exposure to CW, of course. Typical CW communication includes to say the call name of the HAM radio operator, you get fluent to read/write that quickly. But then, reading the call name of the other partner of the communication will look much like random characters, not something like a word. So maybe training with random sequences is not that bad after all.\nFor motivation, the Long Island method might make sense: you can right from start get into online-video sessions with others on your speed level and learn together. Just the \u0026ldquo;speeding up\u0026rdquo; might be harder than directly starting with high speed.\nWill it be easier to learn CW in just listening to it, in the background? I think no. In the morning I\u0026rsquo;m following Japanese news tv online, mostly listening while catching up with the Mastodon time line. One hears every now and then important words, and then I switch to the browser-tab with the news site and actually follow the news. But one needs to really concentrate to understand, I\u0026rsquo;m not convinced that just listening passively is helping.\nFixing habits After 3 months, I got aware of a bad learning behaviour and needed to fix that: usually one listens to ~1min of CW characters, tries to pick them up, and type the characters into a text field. One has 2 options then:\nTo listen again, and make corrections over what was heard and typed in with the first run. Or directly click \u0026ldquo;check result\u0026rdquo; and get rated. I had gotten into the habit of listening 2 or 3 times, and just then \u0026ldquo;check result\u0026rdquo;. After asking in the forum on whether that could also be shown in the stats, I got aware that this is not intended. So since then, in almost all cases I directly check the result.\nLet\u0026rsquo;s see how far I get, if I can at some point go to a mountain and do SOTA with Morse Code. :)\nHints, comments, own experiences welcome! -\u0026gt; Fediverse thread\n","date":"2025-05-11T19:57:58+09:00","image":"/posts/2025/05/learning-morse-code-in-2025/paddle.jpg","permalink":"/posts/2025/05/learning-morse-code-in-2025/","title":"Learning Morse Code in 2025"},{"content":"How it started A friend owns rice paddies in Tochigi-ken, the district north of Tokyo where I live. Around the beginning of May, the rice planting is done. Since some years I attend the rice planting: initially going by train, and now for the third year by bike.\nI can not go by bike right at the day of planting, it\u0026rsquo;s just to far. So I go to Utsunomiya, stay over night, and then at the day of planting go another 30km. My friend can just pin down the day of the event one week in advance, as weather should be good and a neighbor with special equipment needs to have time. As soon as I heard about May 3rd (a Saturday) would be this years rice planting, I started to lookup hotels.\nRice planting always happens around the \u0026lsquo;Golden week\u0026rsquo;, a sequence of national holidays in Japan - when all Japan is out for trips, to hotel rooms are scarse. I looked at my calendar, saw that May 2nd is a work day, and assumed May 1st to be one of the holidays. I got a hotel room from Thursday to Saturday, just to find out later that the assumed holiday for Thursday applies to Germany and not Japan.. so I took off the Thursday and planned for Friday to work from the hotel. Friday turned out to be a rain day anyway.\nPlanning As usual, I planned the trip with brouter.de. This is a nicely done website using data from Openstreetmap and other sources. You assign a start point, an end point, and the site plans a trip to be done by bike which tries to avoid big roads, can try to use trails along rivers and so on.\nbrouter allows many options. Last year, the \u0026rsquo;trip time\u0026rsquo; of my first day (120km) was far off, assumed trip time was way to low. This time I looked up my daily commute to the office, 9.5km, and adjusted the options until the trip time matched up with reality: in setting bikerpower=50 . With this setting, I computed the single trips and exported them as gpx files:\nSetayaga Tokyo -\u0026gt; Utsunomiya hotel, 131km / 10h trip time, broute link Utsunomiya hotel -\u0026gt; friends place, 31km / 3h friends place -\u0026gt; 2nd hotel, 44km / 3.5h 2nd hotel -\u0026gt; Setagaya Tokyo, 117km / 9h Looking up addresses in brouter is not very easy, I looked them up on google maps and then used geographical features (rivers, streets) to find the same place on brouter.\nI copied the maps to the Pixel 9, the Pixel 4a and the iphone from the company. All of these have OSMand installed, which can fetch maps for offline use and can open the routes to \u0026lsquo;follow along\u0026rsquo;.\nLeaving Tokyo I had planned the trips of the last 3 years on brouter.de, always doing a new planning as the hotels I would stay always changed and I wanted to take advantage of potentially improved route selection. All 3 routes had always returning parts. First milestone common to all the trips is \u0026lsquo;Kitchen Toretate\u0026rsquo;, a nice place offering food and banks to sit down, with bikers as main customers. It\u0026rsquo;s nicely located north of Tokyo, like \u0026ldquo;a gate to the outside world\u0026rdquo;.\nThis time I just had the company iphone plan a trip to \u0026lsquo;Kitchen Toretate\u0026rsquo; with google maps: I did not expect to see many nice spots in the city on that part of the trip, and google is just suggesting big and efficient roads - so also quite fast. This part took 80min. From \u0026lsquo;Kitchen Toretate\u0026rsquo; on I used the 2 Android phones until I reached the hotel around 5:30pm. I also started to recharge from the laptop battery, but did not need the recharged phone eventually.\nI had the mobile phone attached to the handle bar. Pixel 9 and Pixel 4a allow unlocking with finger print - convenient. The iphone tried face recognition, failed as I was wearing helmet and sunglasses, and so the phone asked for pin - annoying.\nFor longer sections I turned the screen off to save power. When following the route on the Android phones, I had no online connectivity.\nRural area Now, finally more quiet areas, first rice paddies. The route was often along rivers, trying to stay away from big roads - at the cost of a longer trip, of course.\nI saw the preparations for the \u0026ldquo;big kite festival\u0026rdquo;, so where people try to fly a huge kite. This was especially interesting as two years ago I had witnessed that event: that place was then crowded by people, all if festival mood, watching kites.\nThis image was taken 2 years ago.\nLunch One part of the trip was along highway 4, I had lunch at the same place as last year. At the nearby \u0026ldquo;Bando taro\u0026rdquo; company I witnessed all staff in front, bowing at the arrival of someone, maybe a customer or a boss. \u0026ldquo;Bando taro\u0026rdquo; is in food business, equipping restaurants. The text next to the rocket says 坂東太郎宇宙食開発に挑戦, so \u0026ldquo;Bando Taro takes up the challenge to develop space food\u0026rdquo;.\nThe menu at the restaurant was great, I had the asparagus, along with vegetable. Some time after lunch I passed by this Hokkaido style bakery called ペンギン (penguin), where I got a dessert.\nUtstunomiya Onwards to Utsunomiya, for the first stay at a hotel. Nice green everywhere.\nFriday Friday was all rain, I stayed at the hotel and was working. The wlan was ok. The hotel featured a \u0026ldquo;public bath\u0026rdquo; where one could enter all day from 5am to 2am in the night. I had lunch at a Soba restaurant nearby. A super market was featuring potato chips with Lemon- and Strawberry taste!\nRice planting! Saturday was the rice planting day. I still had to go 31km in the morning, wanted to leave as early as possible, but also wanted to catch breakfast which started at 6:30. So I left the hotel at 7am. The planned route tried to work around big roads, but as there was not much traffic on this Saturday, I often took the road instead. My friend is living in a beautiful valley, farm houses on both sides, rice paddies in the middle. Arriving by bike, I first get a great view of the neighbor valley, then enter the actual valley, and drive up for 5km. Arrival around 9am was just right, the planting had just started.\nUntil around 1970, rice planting was very hard: whole families spreading over the rice paddy and one by one sticking rice plants into the soil. At rice planting time, the soil is ideally just slightly covered with water, In the past, people did not use boots, just bare feet.\nAround 1970, the first rice planting machines appeared, making this much more effecrive. Young rice is grown from winter on, to be in the right state just at planting time. The plants are fed into the machine. The machine is grabbing plant after plant and planting it - amazing technology. The machine is leaving deep traces in the soil, my activity is to leverage that soil out again.\nAround 5 people were working directly on the fields: operating the machine, feeding it with fresh plants. I did just ask my friend for permission to put up his picture. We had break, and later lunch. Lunch was great: hand made sushi. Everybody takes a sheet of seaweed, places cooked rice ontop (here of course from own production), adds anything one likes like cucumber, tuna, bamboo pieces, and eats that.\nLeaving I left around 4pm, and saw a great sunset. Had to attach the light to the bike, navigation also turned to dark mode. This aprt of the trip was 44km, took 3.5h. I spotted HAM radio antennas, but had no time to look for the operator and talk. The hotel room turned out to be a full apartment, many times bigger than my 2 rooms in Tokyo.\nTrip back The trip back to Tokyo was 117km, took 9 hours. Weather was nice, I saw another bikers meet point, before coming back to the Kitchen place, and Tokyo. The thing with \u0026ldquo;クスリのアオキ\u0026rdquo;: the first 3 letters are almost the representation of how I sign my mails in Japanese,　クリス (\u0026ldquo;Chris\u0026rdquo;). On this plate, it reads \u0026ldquo;kusuri\u0026rdquo;, so medicine.\nWhat\u0026rsquo;s now left is to eat the pa-rukan fruit which I got, and to do washing.\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-05-05T08:04:31+09:00","image":"/posts/2025/05/rice-planting-biketrip/33_rice_planting_machine_small.jpg","permalink":"/posts/2025/05/rice-planting-biketrip/","title":"Bike trip to Tochigi and rice planting"},{"content":"With the right wing parties in Germany getting stronger, my Mastodon timeline gets posts from people considering to move to other countries, temporarily or forever. Japan is known for being peaceful, rich of culture, and is often called out as potential destination - so as a German living in Japan since 8 years, I will try to look at the relevant aspects. Maybe this sparks also others living in foreign countries to report?\nThe one thing I can promise here is that you get my honest view on single topics which I find relevant - but I might actually miss facts, and then everybody will weight the single factors differently. For some topics I\u0026rsquo;ll look at the factors, and give points on a scale 1 (worst) to 5 (best) for Germany and Japan.\nHow easy do you get through with English? Not easy. You will in all important places find people who speak English to some degree: in the town hall, when opening a bank account, when buying a phone SIM card, renting a place. But you might have to wait longer, and it will be less convenient than if you know Japanese. Choosing doctors who speak English is possible, but they will be further away. Japanese have a strong urge to receive services in Japanese, so as an IT support person I also had to prove some level of Japanese proficiency to get my Japanese work contract. It\u0026rsquo;s ok to not speak perfectly Japanese, and mobile phones with an internet connection can help very much. Also after coming here I continue to learn Japanese by various means. My non-perfect-Japanese is a slight downside at work, but feeling comfortable with English is an advantage and helps me to connect with our international teams.\nI recommend to learn Japanese, at least to some degree. If you look at it as a chore, it\u0026rsquo;s already over - you need to find beauty in it, and it\u0026rsquo;s also a great way to learn about the culture. I found pieces of Japanese literature which I really like, and which I can read in their native language. This is increadibly satisfyung. I\u0026rsquo;m also still learning, by various means.\nIn Germany: it\u0026rsquo;s much easier to get around just with English. But when looking for a flat, speaking German is a huge advantage.\nJapan: 2/5 Germany: 4/5 Culture differences, peaceful daily life One should not ascribe certain traits to people having a certain citizenship.. but I think it\u0026rsquo;s ok to \u0026ldquo;see some patterns\u0026rdquo;. I ascribe attributes like punctiality, trying to do something rather well than quick, being rather silant than noisy to both Japanese and Germans. So from that aspect, most Japanese people consider the same pieces important as myself.\nLiving in Japan, I feel even safer than in Germany. Parents allow their children quite early to commute by themself with public transport, quite untinkable for example in America.\nIn Japan, the police is not seen as an enemy but as a friend. Many policemen have received training in martual arts, so tend to know how to solve a situation without shooting weapons.\nThe style of thinking in Japan helped a lot in the Corona time: while in Germany some people went onto the streets, screaming \u0026ldquo;I won\u0026rsquo;t wear a mask just because the government tells me\u0026rdquo;, people here simply used more masks than usual, death numbers were very low - despite high population density in Tokyo and other cities.\nJapan: 5/5 Germany: 4/5 How welcome are foreigners? Is it easy to settle in? In Germany, especially right wing groups are right now making it harder to immigrate. Foreigners feel increasingly unwelcome. Germany has a strong record of accepting immigrants or foreigners living in Germany temporarily: this peaked 2015. Without immigrants, the German population would shrink.\nJapan is quite different: it had long times of self isolation. Immigration numbers are much lower than for Germany. Right now Japans polulation is 150% of Germanys polulation, but it\u0026rsquo;s shrinking without the stream of immigrants. Overaging, and keeping the pension system up is even more an issue than in Germany. That being said, I never felt unwelcome here. Some Japanese have not much contact with foreigners and feel uneasy when interacting, what to expect - but I did not sense \u0026ldquo;you are unwelcome\u0026rdquo;.\nJapan has one complication though when settling in/renting a flat: you need the signature of a \u0026ldquo;guarantor\u0026rdquo;. You also put down \u0026ldquo;security money\u0026rdquo; to restore the place in case you move out and leave havoc, but additionally the guarantor can be made responsible for your actions. In some cases companies can sign up as \u0026ldquo;guarantor\u0026rdquo;, for a fee.\nVisa: Germans can stay up to 3 month with a visitor visa. For longer, you need to apply: one can get a visa \u0026ldquo;for work\u0026rdquo;, to \u0026ldquo;do arts\u0026rdquo;, to \u0026ldquo;teach a religion\u0026rdquo; or for other reasons.\nJapan: 3/5 Germany: 3/5 Ethical differences Traits like \u0026ldquo;it\u0026rsquo;s good to be on time\u0026rdquo; are equally important for Germans and Japanese, that\u0026rsquo;s already good. Stealing things or murder are forbidden by law, such things do also match. Looking into the details, you can spot differences though. Death penalty does exist in Japan: the criminal might be living in prison for years and will just one morning learn that he will be executed on that very day, one or 2 hours before the execution.\nFurther difference: pets are popular both in Germany and Japan. In Japan, abandoned dogs and cats are killed at shelters, more than 40.000 in 2014.\nSomething great in Japan: how to deal with things in a careful manner. The second hand market of clothes/electrical applicances/bikes etc. is big, and things are often in great shape. I love to look through thrift stores like \u0026ldquo;hard off\u0026rdquo;, for retro computer hardware and gaming consoles. Unfortunatelly I live on limited space.\nAnother great thing in Japan: trust by default. When paying taxes in Germany, you fill out forms (in paper or digitally on a website) and then have to provide papers to prove what entered into the forms. In Japan, tax for my normal wages are deduced automatically. I need to declare stock though which I receive from the company. For that, I can compute the value of stock I received per year myself, and enter that value into the forms, then I\u0026rsquo;m getting taxed on that. All proof for that stays with me, I\u0026rsquo;m trusted by default to have declared this right - and I\u0026rsquo;m asked to keep the papers for 5 years in case there are any questions.\nJapan: 3/5 Germany: 4/5 Prices, wages, work, pension Prices of commodities like food are now lower in Japan, as we have seen less inflation in the last years. Rent in Tokyo is quite high though, I pay 108800円/700eur per month for ~40qm without water/gas/energy, and I\u0026rsquo;m not really living in the city. In my IT area, I earn 82% of what I would earn in Germany right now.\nWork environment in Japan: Japan is known for long work hours, that\u0026rsquo;s true for many traditional Japanese companies. I\u0026rsquo;m working at the branch of an American company and can get along with the conditions. The tax system is comparable to Germany: they are automatically deduced from your monthly wages. I need to do an extra tax declaraion in Japan, once per year, as my company is giving me company stock.\nIn general, I think I pay a lower percentage of my wages as taxes than in Germany - but the services by the government are also a bit different. Most prominently, education of children in Japan tends to be more expensive in Japan. Many children are sent to \u0026ldquo;after school\u0026rdquo;. So for me, earning for dozens of years the lower Japanese wages, and then after retirement living with that in the more expensive Germany, would be a bad plan. For now I\u0026rsquo;m ignoring that line of thinking, as this unbalance has just come up in the last 3 years.\nI\u0026rsquo;m paying money into the Japanese pension system and hope to get some out when I retire. The idea is the same as in Germany.\nJapan: 4/5 Germany: 4/5 Health The health insurance system is quite comparable to Germany: either from the employer or from the government you enter health insurance. It\u0026rsquo;s just \u0026ldquo;one class\u0026rdquo;, not two classes as in Germany. The employer is deciding the health insurance company. For medicine or medical services I pay 30%, the rest is covered.\nIn Japan, you can drink water from the tap - and you should, it\u0026rsquo;s well checked like in Germany. Like many Germans, also many Japanese prefer bottled water. I use the tap, my day starts with green tea from the tap.\nEmployers give employees one free day per year, and pay for a health checkup. In general, the rate of obese people is low. I think that influences the own standard of \u0026ldquo;what is considered normal\u0026rdquo;, so sets incentives to stay slim/live healthy also for me. As per my understanding, obesity in Europe is bigger, and even bigger in America.\nMore details in article The health systems in Japan and Germany.\nJapan: 5/5 Germany: 3/5 Climate, danger environment wise Japan is shaped a bit like a boomerang, and covering more climate zones than Germany. I\u0026rsquo;m living in Tokyo, it\u0026rsquo;s in average warmer than Germany. July to September/October are scorching hot, unbarable without aircon. One can also move to Hokkaido in the north (colder), or to one of the many places in the south (warmer - but more dry air).\nJapan is facing much higher danger regarding natural deasasters: light earth quakes are almost daily. Big ones are appearing roughly in 30 year cycles, and a big once for Kantou, where I live, is overdue. Japan is also experiencing Tsunamis (huge waves), and storm. There is a rain season, traditionally one week of rain. Many buildings are made from wood, there have been deadly fires in the past. I should mention that the geological features also have good sides: Onsen, where hot water comes from the earth and resulted in bathing culture. More in article Earthquakes, floods and Japan: human factor and technology.\nScientists predict Japan to suffer more with the intensifying climate crisis than Germany.\nIn Germany, I was separating the year into 4 seasions. Here in Japan, I\u0026rsquo;m for myself counting much more: winter, spring starting, cherry blooming, rain season, summer, taifun season, autumn.\nJapan: 2/5 Germany: 3/5 Environment protection and climate crisis awareness Things are dark in Germany, and even darker in Japan: people are not giving the climate crisis the attention it requires. In Japan, even less recycling is done, even less awareness that one should prevent trash in the first place, save energy and so on.\nTo be fair: it also seems \u0026ldquo;harder\u0026rdquo; for Japan to get green than for Germany. Less areas for solar cells, these would need to be mounted more expensively to cope with the rock and danger of earth quakes. Wind mills on the ocean: the ground of the nearby ocean regions is very deep, so wind mills would need to swim, making them more expensive. Translating wave energy into electric energy seems to just work in 2 places of the world.\nJapan: 2/5 Germany: 2/5 Traffic, Transport and communications Public transport and traffic in the big cities like Tokyo and Osaka are excellent. Everything reachable with the train, efficient payment with cards - while staying anonymous. High speed trains through whole Japan, on separate tracks, highly reliable and on time. Great mobile networks, affordable. Great internet connectivity via fiber in most places. Of course, on the country there are also places with less frequent buses.\nFrom most places, one can also reach one or more airports conveniently.\nWith a German drivers license, you can drive in Japan for 3 month. If you stay longer, you anyway need a visa, and you can get the German license converted into a Japanese one.\nWithin 1 hour my train I can reach a train station from where I can go hiking for some hours. Just jogging is not ideal: there are parks for jogging, but no longer ways for jogging without traffic lights. Also just asphalt, no soft ways like you would find in the woods.\nTokyo is not ideal for cycling, but it\u0026rsquo;s not bad either. Most days I\u0026rsquo;m commuting by bike to the company, in 45min one way.\nIn Japan, Internet access is not censored, not like i.e. Korea where you easily hit zensored pages, or in Germany where you also might hit DNS zensored pages at times (i.e. imposed by some ISPs).\nA downside in Japan: transfers of money between banks can be hard. Some banks do not talk to the others, so I have friends who got new bank accounts just so their employer could pay them, or that they could send over monthly rent per bank transfer. That\u0026rsquo;s something which is unthinkable in the European Union with SEPA transfers.\nJapan: 4/5 Germany: 3/5 Entertainment, finding communities In the big cities like Tokyo, everything is happening. There are meetups of all kinds of folks: around certain technology like Linux or HAM radio, around languages like \u0026ldquo;German\u0026rdquo; and so on. There are maker spaces, 2600-meetings. Culture wise, all one could imagine is offered: rock concerts, classical concerts, cinema and so on.\nOnsen (bathes) are also a fantastic: I enjoy these very much, visiting them once per week.\nWhen you live in bigger cities, language is usually not an issue when connecting with communities: there are English speaking HAM radio groups, Linux users, makers, nerd nite. German speaking groups where I attended presentations on topics like Japan history, \u0026ldquo;influence of 2nd world war on Japan Manga\u0026rdquo;, \u0026ldquo;Getting green in Germany and Japan\u0026rdquo; and more - with amazing technical depth. I meet amazing specialists and discuss differences between Japan and Germany. When coming to Japan, I tried to focus on Japanese speaking communities to better learn Japanese. Now I more focus just on the topic of the groups, ignoring the language.\nJapan: 5/5 Germany: 4/5 Food ..is superb, for my taste. Great lunch is available for 1100円(7eur), no upper limits, of course. I\u0026rsquo;m not cooking at all, always eating at restaurants. Or getting sushi etc. at super markets. For example, have a look at A week of autumn food in Japan.\nJapan: 5/5 Germany: 3/5 Politics, bureaucracy Politics in Germany as well as in Japan are quite conservitive. Right wing tendencies in Germany got much stronger lately, but in Japan these have always been there.\nAs for feeling included in Japanese policits: I\u0026rsquo;m not included at all, as a foreigner I can not vote, not even though I\u0026rsquo;m registered as \u0026ldquo;permanent resident\u0026rdquo;.\nAn example of Japanese politics working for the people: the political decisions to influence money inflation. In America the inflation is steered to be very high. In Europe it\u0026rsquo;s also high, although not on American levels. In Japan, it\u0026rsquo;s very low - which means that the savings of people are rather stable, they are not loosing their value. This decision is despite the Japanese Yen loosing value would help the state itself to devalue its debt. Another downside is overtourism, people streaming in for very cheap holidays.\nBureaucracy is bad in Germany, and even worse in Japan. Much paperwork, not much digitilization.\nJapan: 1/5 Germany: 2/5 Child Education I\u0026rsquo;ll share thoughts on this one - but no ranking as I do not have children.\nAs for money: child education is more expensive in Japan, more is directly payed by the parents. In Germany, taxes are higher, but more of education is payed via the government/tax system.\nBoth countries have high standards for i.e. mathematic education. In Germany, 2nd and 3rd languages are much more important, i.e. to speak English. In Japan, I attended school lessons where students did read English texts which were really hard to understand. In Japan, much more time is spent on learning the ~2000 Kanji required for daily life. It\u0026rsquo;s my impression that there is less focus on own thinking, and more on memorizing things - well, to learn Kanji then memorizing is also the more helpful skill. Other fellow Germans with family and children running through the Japanese education were really positive though on the system: they felt their children well prepared also for international carreers.\nSummary I have written about many aspects in detail, on the blog and here.\nAlready choosing the relevant factors is of course subjective, and also the single ratings. Of course, there are further factors not fitting into above categories. Until my 12th year I was growing up in Socialist East Germany, knowing I would never be able to go to America or West Germany. Having the option of living abroad is great!\nWhich factors did I forget? Super curious also how you rate the country you are living in!\nCorrections? Questions? -\u0026gt; Fediverse thread\n","date":"2025-02-22T08:46:14+09:00","image":"/posts/2025/02/living-abroad-in-japan/20210209_221733610_filing_5year_visa_extent_spirit3_small.jpg","permalink":"/posts/2025/02/living-abroad-in-japan/","title":"Living in Japan: How does it compare to Germany?"},{"content":"Virtualization? Emulation? What are these, after all?\nVirtualization in our context here means that when a guest operating system does computations, these get directly offloaded to the hypervisors cpus. One of the benefits is performance, one of the downsides is that the guests cpus need to exactly match the hypervisors cpus, or a subset. Emulation means that the computations are done in software. With this, host/guest architectures can be different. As a downside, this is slower than virtualization. Using emulation, we can for example start a x86 based system, and then emulate completely different architectures like ARM or sparc64. How much performance does that cost us, how much power does this consume?\nThe setup I\u0026rsquo;m experimenting since some time with setups to measure the electical consumption of systems, the setup is described in more detail here and here. The summary: I\u0026rsquo;m using a Thinkpad T590 (x86 architecture) with Fedora to measure power consumption of the cpu/memory/gpu components. The Intel cpu exposes these via RAPL mechanism. The Performance Co-Pilot (PCP) framework is constantly measuring and archiving these metrics for me, so I can use this archive to \u0026rsquo;travel back in time\u0026rsquo; and look at the metrics at any point in the past.\nWith the measurements running, I do these:\nrun the virtual guest to me measured, for example a guest with emulated arm64 architecture then I run Ansible scripts from the hypervisor, these copy further scripts into the guest, into a RAM disk then workloads are run in the guest, each workload for 300 seconds after each workload I capture how often the workload was running in the guest (for example how often a file got unzipped in these 300sec), and I capture the power which was consumed on the hypervisor. The hypervisor is for these measurements not running other workloads, the bigges workload is the emulation of the guest.\nAs workloads, I try to cover the main resources which are relevant:\ncompute power: I measure how often a file can be uncompressed network throughput: iperf2 and iperf3 are used memory: for this, I write block files with zeros into a RAM disk disk I/O: - let\u0026rsquo;s read and write blocks of data in the guest, with and without sync For all of there, I\u0026rsquo;m not only curious on the actual performance in the guest, but also how much electrical power was used. As for operating system in the guest: Linux is not really covering many architectures, that\u0026rsquo;s the strong point of NetBSD.\nNotes on emulating the single architectures ARM64 is quite solid: emulating 16GB of memory is no issue. 32 cpus is also possible, but then the 8 cores of my hypervisor are 100% busy. I settled with emulating cpu cortex-a53, after cpu=cortex-a57 and cpu=max turned out slightly slower. The NetBSD RISC-V port is quite new, instead of doing an installation of the stable/released version 10.x I had to use the 10.99.12 development version. With that, I did not see signs of instability. My measure scripts use sudo, iperf3 and bash - for stable NetBSD on aarch64, sparc64 and so on I could fetch and install these as precompiled packages but here I compiled these myself, with pkgsrc. sparc64 with the sun4u architecture is just getting emulated by qemu with one cpu and up to 2GB of RAM. NetBSD has also since some versions problems to access ATA bus (for the emulated disc) and the network card. A 100Mhz UltraSPARC-IIi cpu gets emulated. Both NetBSD and OpenBSD have issues to run on this emulation, interrupts scheduling related to ATA disk and network card are involved. sparc, I used \u0026lsquo;sun sparc 20\u0026rsquo; emulation, configs with 4 cpus and 2GB RAM can be configured. When running \u0026lsquo;iperf\u0026rsquo; or doing block I/O to ramdisk, the guest system did run into stalls for me. To replicate, execute \u0026lsquo;mount_mfs -s 1200m -u chris swap /dev/shm\u0026rsquo; and then \u0026lsquo;dd if=/dev/zero of=/dev/shm/1gbfile bs=1m count=1024\u0026rsquo;. The cpu bound workloads did run without issues, or I/O to disk. Another obersvation: even with more than one cpu emulated, the performance does not increase with multiple threads started. AMD64/x86_64 is of course matching my hypervisor for these tests, so besides emulating (\u0026quot;-cpu Icelake-server-noTSX\u0026quot;) I did for reference also run NetBSD with KVM, so virtualized. I did also run it virtualized with KVM and Linux as guest, but then with a much more complex qemu config - not started by hand with \u0026ldquo;qemu-system-x86_64\u0026rdquo; but with a config suggested by libvirt. i386 turned out also quite stable, I emulated and benchmarked 2 configs: \u0026ldquo;-cpu qemu32-v1\u0026rdquo; and \u0026ldquo;-cpu pentium2\u0026rdquo; to see how much they are different. Comparing CPU power over the architectures Qemu is emulating the various architectures differently \u0026ldquo;well\u0026rdquo;, restricted to numbers of cpus and so on. Also NetBSD does not support all architectures equally well, i.e. for some it is limited to a single cpu. I looked at cpu power like this:\nspawn a number of threads with bash: 1, 2, 4 or 8 inside each of the threads run a loop which extracts a file (\u0026lsquo;bzcat file\u0026rsquo;), increases a counter of completed extractions, and jump back to extracting the file again. So there run multiple \u0026rsquo;extraction loops\u0026rsquo; independent of each other. If these loops match the number of cpu cores, then each loop gets scheduled to run on it\u0026rsquo;s own core We run these loops for example for 5min, then we kill the loops and collect the number of completed extractions Let\u0026rsquo;s have a look:\nFrom the left to the right, the number of \u0026lsquo;parallel jobs\u0026rsquo; was increased. On the left, for \u0026ldquo;1\u0026rdquo;, one loop which concstanly extracted the same file was running for 5min. qemu can emulate multiple cpus for most architectures, so when 4 loops with extract jobs run in parallel, the maximum of work is getting done. As we increase the number of parallel threads further, the threads do no longer have one cpu.\nThe red line at the bottom is from the sparc64 emulation, as mentioned before it did not run properly. Also just a single cpu could be used. The blue line in the middle is \u0026lsquo;sparc\u0026rsquo;: performing better, but even when configured with 4 cpus just one seems to get load. Then the area of well emulated architectures: i386 on the lower end, and aarch64 at the top end.\nDifferences also in this field are quite big: with 4 parallel threads, i386-pentium2 completes 107 jobs, while aarch64 completes 144 jobs.\nThat\u0026rsquo;s 34% more!\nNetwork and Disk I/O How about the other resources?\nQuite some variety here, the x86 based architectures have the best network throughput. This was tested with iperf3, and for \u0026lsquo;sparc\u0026rsquo; with \u0026lsquo;iperf2\u0026rsquo;.\nFor I/O to disks, the reads for the emulated amd64 are exceptionally fast, I tested that value multiple times. Also big variety among the others: RISC-V is slowest for reads, that\u0026rsquo;s why compiling there was so painfully slow for me. Energy consumption over all of the architectures is more or less same while performing these tests. The energy is here measured with the cpu RAPL-MSR metrics, via pmda-denki.\nEnergy efficiency If we would emulate systems to do raw computing, how efficient are they? In other words, to extract a bzip2 file in the emulated guest, how much electrical power is spent?\nFor this, I was looking at RAPL-MSR metrics again. Various metrics are available, for example \u0026ldquo;overall system consumption\u0026rdquo;, so for example consumption of nvme and attached USB-devices are also counted in that metric. I was here using the metric for the cpu/memory/gpu subsystems. These consume power even when nothing gets computed. The more cpus we can use for our computations, the better gets our \u0026ldquo;power per extract-operation\u0026rdquo; value.\nHere we see nicely how sparc is using just a single cpu, and starting more jobs is just making the efficiency worse. The other architectures have 4 cpus, so the consumption per single extract operation goes down until 4 threads, and is then increasing.\nSummary If you are not after performance, emulation for most architectures is usable. Most architectures are stable enough to just try operating systems like Solaris. NetBSD is the most flexible candidate to cycle through various architectures and everywhere find the same environment for benchmarking. \u0026lsquo;aarch64\u0026rsquo; emulation turned out quite performant, if your hypervisor has many cores then you can afford to also emulate many guest cpus. RISC-V worked surprisingly well. Getting the whole setups and tests to run took quite some time, but was fun.\nI just looked at qemu here, but there are many more options, some are listed here. There is so much more to try in this area: use BSD as hypervisor. Verify if for example aarch64 as hypervisor shows the same performance. Try other BSD and Linux as guests. I\u0026rsquo;m also curious about performance of emulators for consoles like SuperFamicon, or Amiga, and compare how efficiently they emulate their target while producing graphical output.\nI will soon do a performance comparison of emulation and virtualization. What would you like to see?\nNotes/additional Sascha Opletal shared that with the following options, the sparc emulation can activate multiple threads for sparc/32. So for 4 cpus this will enable 4 thread, and allow the hypervisor to schedule these to multiple host CPUs, for better performance:\nqemu-system-sparc -M SS-20 -smp 4 --accel tcg,thread=multi [..] The guest has of course also to make use of multiple CPUs. If the default NetBSD kernel for your version is not doing that, install (or build) a GENERIC.MP kernel.\nCorrections? Questions? -\u0026gt; Fediverse thread\nLinks the job control scripts PCP pmda-denki handbook Raw data of the tests, images etc. are here ","date":"2025-02-07T21:05:24+09:00","image":"/posts/2025/02/emulation-performance-and-consumption/gnuplot_cpu-performance-emulation_small.jpg","permalink":"/posts/2025/02/emulation-performance-and-consumption/","title":"Emulating architectures with qemu: which performs best?"},{"content":"I\u0026rsquo;ve been living in Germany and Japan, let\u0026rsquo;s look at some aspects of the health systems.\nRundown of a typical doctor visit I\u0026rsquo;m entering the doctors office at 10:30 on a weekday. The naming is already interesting: in Japan, the normal places where doctors investigate and treat patients are called clinic (クリニック), which sounds like the German \u0026ldquo;Klinik\u0026rdquo; - but in Germany these are big complexes to treat patients, so in English \u0026ldquo;hospitals\u0026rdquo;.\nThis doctor is for generic health issues, also Japan has of course all the specializations of doctors into teeth, eyes, and so on. Most doctors receive patients over the week, some additionally on Saturday or Sunday. This is to give more flexibility to Japanese who simply do not find the time to see a doctor under the week. The worktimes in Japan are not as focused on Monday-Friday as in Germany: stores are also open on weekends, construction works on roads are often done in the night to not interrupt traffic.\nSo I enter the doctors place, and 4 people are already waiting. Some doctors accept no appointments and their patients simply appear - most of the patients of this doctor come without appointment. In contracst, I had also visited an ear doctor, with just people with appointment waiting - so I had to wait 2.5h until examination.\nI go to the reception desk, nurses greet me and ask for my \u0026ldquo;customer card\u0026rdquo; and \u0026ldquo;health insurance card\u0026rdquo;. Employees at medical institutions in general wear face masks, and also patients are asked to do so - just now I remember and put on a face mask. These \u0026ldquo;customer cards\u0026rdquo; are handed out by each doctor or hospital. They have the same size as a credit card, and are used to identify me - basically just assigning a number to me. The cards are made from plastic on paper, and can also be machine readable via bar-code or magnetic stripe. The pursue is full of these, as each doctor or hospital is handing out such a card to you.\nI am sitting down, and other patients get called into the doctors examination room. Some of them arrived after me - and when these are called into the doctors room also a \u0026ldquo;by appointment\u0026rdquo; is added. With that, us patients without appointment clearly understand that the order is arranged to accomodate here someone with an appointment. Clearly better for the doctor to state this, instead of having patients on the internet complain \u0026ldquo;that others, who arrived after me, were examined before me\u0026rdquo;.\nAfter 30min I get called to the doctor. He measures blood pressure, we have a 5min chat where I report on the results of blood pressure measurements I do every morning, then he prescribes me again 2 months of blood lowering medicine. I\u0026rsquo;m also asking about a \u0026lsquo;clogged nose\u0026rsquo; which I experience recently at night. I have ruled out that it\u0026rsquo;s a plain cold by myself already. He prescribes a nasal spray, to be inserted into the nose before sleeping.\nThan I\u0026rsquo;m sitting down in the waiting area again, after 5min I\u0026rsquo;m called to the reception, I pay 1000yen (~6eur) for the consultation. The actual price is 3 times of that, but the other part is payed by health insurance. Then I leave.\nReceiving the medicine The doctor just gave me the prescription - like in Germany, most medicine requires prescription. I enter a pharmacy, in Japanese 薬局. I hand over the prescription from the doctor, the health insurance card, and the medicine book (おくすり手帳). The medicine book contains a list of all medicine I receive - so even if I was visiting the whole country, receiving medicine from various places, this book would still allow to have an overview of what was received at which date, to check for conflicting medicines and so on. Such a handbook is just fantastic - not storing the data in a central place, and keeping the data in the patients hands!\n![At the pharmacy](1/at the pharmacy.jpg) ![reminder from dentist](1/handwritten reminder from the dentist to get a checkup appointment.jpg) ![Burned calories at stairs](1/hospital stairs with burned calorie counter.jpg) ![Instructions to use spray](1/Instructions to use a nasal spray.jpg) ![paymeny machines at clinic](1/payment machines at a bigger clinic.jpg) ![checkups include lunch](1/Some checkups include a lunch.jpg) ![the medicine book](1/the medicine book.jpg)\nI sit down, and see how employees take the medicine out of cupboards, count the medicine, and use small stamps (hako) to mark on the paper that they verified the kind and amount of medicine. I also had instances of medicine not available, where staff asked me to return later again, or did calls to pharmacies in the neighborhood, to find out who had the medicine. The pharmacy will also ask you if you are ok to use generica, so medicine produced with the same ingredients but by a different company.\nAfter 10min I receive the medicine, along with the instructions on how to take it. I\u0026rsquo;m again paying around 1000yen, also here the other 2/3rds payed by the health insurance.\nThe health insurance system in Japan 8 years ago, when I arrived in Japan, I had to register as resident (住民) at the local townhall. As it was the default, I entered then the \u0026lsquo;public health insurance\u0026rsquo;. Some weeks later, when I started to work, I changed into the company provided health insurance. So like in Germany, there are 2 basic kinds of health insurance:\npublic via the company The one related to the company is providing slightly different services and is usually preferred. Like in Germany, the health insurance pays a share of medical costs like doctor visit or medicine.\nOne really nasty thing about Germany is that there are 2 kinds of health insurances: \u0026lsquo;private health insurance\u0026rsquo; and \u0026lsquo;compulsory health insurance\u0026rsquo; (\u0026ldquo;Gesetzliche Krankenversicherung\u0026rdquo;).\nThe kinds of services included in both are different, but the biggest difference: the \u0026lsquo;private insurances\u0026rsquo; are preferrable for doctors - so in some cases patients in need of diagnosis ask for appointment and the type of health insurance decides then between having to wait 2 days (private insurance) and multiple months (compulsory insurance).\nDigitilization of the Japanese health system People living in Japan get a unique ID assigned, the \u0026lsquo;my number\u0026rsquo;. I also got one, the ID is used for example when paying taxes, and recently also under some circumstances as replacement of the health insurance card.\nI like the \u0026ldquo;medicine handbook\u0026rdquo; as a method to keep an overview of medications, while keeping the data in the users hands. For exachanging data of one patient between multiple doctors and hospitals, Germany is right now trying to install the electronic patiens file (\u0026ldquo;elektronische Patientenakte\u0026rdquo;), but they struggle to get this setup in ways which protect the patients data. In Japan, you mostly get the data in your hands, so you can then take it to the doctor. So when I had headache and an MRT-scan of my head was done, I received that scan data as compact disk (CD), to take it with me to the doctor.\nQuality of the service The quality of the medical services in Japan is considered high, about the same level as in Germany. There is one piece though: the language barrier leads to Japanese in average not speak English well, and it is rumoured that\u0026rsquo;s also delaying medical news/developments/studies to doctors in Japan.\nInterestingly, the German language is playing a big role in Japans health sector. Many terms come from German, sometimes with different meaning. For example, the patients file in Japan is called \u0026ldquo;Karte\u0026rdquo;.\nFurther health impacting details Yearly health check: Japanese employers have to dedicate one day for employees per year for a health checkup. The company also pays for the check. It includes examination of blood and stool samples, xraying the chest, ultrasound examination, search for cancer in the stomach and much more. Generic health awareness: Being surrounded by people who are in general not overweight is also setting the standards for yourself. Japan has fewer overweight people than Europe, and Europe has less than America. Tax: When payments for medicine and doctors reaches a certain amount per year, Japanese can use the bills to lower the taxable income when doing tax declaration. Stress test: Government regulations dictate that yearly stress questionaires are done, to prevent people from overwork. Employees also have to fill in sheets with their daily work time, another counter measure to prevent overwork. Corrections? Questions? -\u0026gt; Mastodon thread\n","date":"2025-01-19T10:20:47+09:00","image":"/posts/2025/01/health-systems-in-japan-and-germany/medicine-book.jpg","permalink":"/posts/2025/01/health-systems-in-japan-and-germany/","title":"The health systems in Japan and Germany"},{"content":"Capturing some projects of this holiday season. Partly to get a feeling of having accomplished something, and partly as reminder to continue the unfinished projects.\nGerman HAM radio exam Right at the start I was sitting down in the German HAM radio exam in Munich. There are 3 levels of HAM radio certificates in Germany. You can apply for the highest one, work through various exam sheets, and if you only pass one of the lower levels then you can walk home with that lower level certificate. I had started quite late to look into the material, with website 50ohm. I felt quite comfortable in the material of the lowest level, and had then started to look at the material for the higher ranks. The exam was multiple choice, I was directly informed that I passed the lowest of the 3 levels, and can now within 2 years try the higher level exams again without having to do the law exam a second time.\nThe HAM radio exam in Japan was also multiple choise based, but took many weeks until announcement of the results.\nDetails are here.\nSoftware defined radio experiments I worked for one day from the Munich office. My worktime aligned to Japanese business hours ended around lunchtime, in the afternoon I looked at the radio communication of 3 LoRa devices which a colleague had kindly brought in. I could trigger messages between the 3 devices, looked up the frequency areas of LoRa, and spotted the signals. More details here..\nIn my house, I installed a simple antenna with 2 telescope sticks (1m) on the attic, and connected SDR devices: RTL-SDR and a HackRF clone. These were connected to a Thinkpad running Debian Bookworm and OpenWebRX+. From all over the house I could with a browser acess the OpenWebRX+ interface, and have a look at the signals in the area.\nHere is an example of an OpenWebRX instance kindly offered for public access.\nI bought 25m of dual-wire copper cable, unrolled both wires, soldered them together to a 50m cable. Then I did span 35m cable from the roof of the 3 story house to the garden, to be used as an antenna. I connected the end to the dual-telescope-antenna, instead of one of the 2 telescope sticks. This was not a complete \u0026rsquo;end fed antenna\u0026rsquo;, as I was not intending to send out with it. As \u0026rsquo;end fed antenna\u0026rsquo;, more hardware would be needed. The results with the antenna were not great, I also have not done a direct comparison of reception with this wire antenna against the 2-telescope-sticks antenna.\nASCII-art experiments I was this fantastic piece of ASCII-art some time ago and tried to do something similar for my website.\nFiglet can do simple art, but not what I was looking for. This flow looks good:\nGenerate a 3D image of the text: I tried Blender and Inkscape, but then settled with gimp. Choose font: I tried many. First find a good candidate with gnome-font-viewer, then generate an image with the letters \u0026lsquo;fluxcoil\u0026rsquo; with 3D-effect. For the 3D effect, gimp allows to choose from where the viewer seemingly is looking at the letters.. but in the original on int10h, the viewer is even in the middle and the viewpoint to the sides is changing. One would need a real renderer like Bender or povray, I did not want to dive into these. jp2a then converts the image to ASCII-art The result is now used on fluxcoil.net. I also tried to use this for a new avatar image on the mastodon profile, but while the coloured ASCI-art of faces looks nice on the terminal, it is to dark when used as small avatar picture.\nLooks a bit better when opened in full size.\nRobotron EC1834 I did research last year on the system. The system turns on, disk is spinning, it sounds like DOS is loaded. The original monitor does not put out an image. I got this converter to convert the CGA/MGA signal from the system to HDMI. The converter is based on a Raspi, I fetched the newest firmware and tried to use it. I see something vaguely resembling a prompt from the EC1834, but no proper picture. The next step would probably be to\neither exchange parts with a known working EC1834 (but I do not know other owners) or try to understand if the graphics card is delivering a proper signal, or if it\u0026rsquo;s broken. I researched oscilloscopes which can be attached to computers via USB, but actually oscilloscopes with buildin screen are down to ~3000yen (23eur), I got one of these. Just did not find the time so far to use it to look at the signal of the EC1834. It also does not have the usual ISA-slots, so just getting a random ISA-graphics card and test is not an option.\nAmiga shenanigans Towards this holiday season in Germany, I was bidding on some Amiga 1200 on ebay. The most interesting one, offered with accelerator card and other pieces, found a new owner for 660EUR. I had left bidding at around 550EUR. Ah well, maybe in the future.\nRegarding my Amiga 500: I\u0026rsquo;m using the original power supply, it\u0026rsquo;s making high frequency noises. Jan Beta made a video where he uses a “MEAN WELL RT-50B” to power an Amiga 500 - just that shipment of these takes time, it would have arrived after my holiday.\nI also have an Amiga 1000, but without keyboard. It\u0026rsquo;s not completely unusable like this: the system boots fine and for example Turrican is playable via joystick. There are no clear hints on which connector is used for the keyboard: I ordered rj10 connectors, and found that these are to big for the opening on the A1000. I\u0026rsquo;ll try rj22 connectors next time. A serial protocol is spoken, so then connecting to a Raspi and actually speaking the protocol will be another challenge - but I saw C-code from someone doing this. Another approach would be to get an adapter to PS2 on ebay, and then PS2 to USB.\nZigbee sensors and smartplugs One year ago I had connected a Zigbee-usb-device to a Raspi4 in the house, and had connected 5 Zigbee temperature sensors which I deployed at various places in the house. Software zigbee2mqtt on the Raspi is used for collecting the temperature readings, which I fed into Performance Co-Pilot and then visualized with Grafana. That worked nicely for a few weeks, I did not check any more - any by now the sensors were all dead.\nZigbee is supporting mesh-style communication: so not all devices are required to directly reach the central usb-device on the Raspi, it\u0026rsquo;s already enough if they reach at least one neighbor. Simple temperature sensors work with a battery and can not act as Zigbee router, but newly bought smart plugs can: I deployed 4 of these now in the house. With that, I can remotely switch on/off, and measure power consumption of connected devices. I have an electrical heater in a bath, to prevent the water pipe from freezing - curious how much energy that heater will consume this winter.\nThese plugs do not include temperature sensors, unfortunately. I replaced batteries in 2 of the temperature sensors, but did not get them reliably into service again - more research needed.\nLinux power consumption per process In the last weeks I had written pull requests for pmda-denki to use MSR-registers: with this we can without additional hardware read the current power consumption of most Intel based systems.\nThis holiday season I used these metrics to attribute overall system consumption to single processes, based on how many compute cycles of the CPUs they are using up. Works nicely! The metric can be made available to Performance Co-Pilot and be graphed with Grafana. As you see, the process of game ParticleFleet is getting 17W attributed:\nThe code is here.\nRetro Linux install My first Red Hat flavour Linux was Red Hat Linux 5.2 Lehmanns edition. Lehmanns was a big \u0026ldquo;books and more\u0026rdquo; online store, before Amazon took over the biggest share of the market.\nThat distro was released in 1998, so 26 years ago. Comes wiith kernel 2.0.36, KDE1, XFree86. My first attempts were with KVM/Qemu: just offering the hosts cpu is enough to boot the cdrom. The kernel just copes with 512MB of RAM. Installation succeeds, but I did only get xfree86 to run with a very low resolution. Many different x-servers are offered, also settings for monitors: after hours of a cycle of modifying the X-config, starting with \u0026lsquo;startx\u0026rsquo;, reading the logfile, I gave up.\ndosbox-X is supposed to be able to boot from cdrom, but was not able to boot the install cdrom. Bochs looked also good, installation succeeds, but I was not able to login on the installed system: after inputting the username, the question for the password does just vanish, and the tty resets.\nPictures are here.\nFuture I will keep an eye on second hand Amiga 1200.. towards next time I\u0026rsquo;m in Germany. Amiga was not big in Japan. Another unsolved item: I wanted to research more on Mastodon servers. I feld the canonical Mastodon to be to heavy for me as single user, needing more maintenance than I want to do. gotosocial is nice, but the frontend is then separated: having a frontend which allows posts in markup would be nice. HAMnet availability in Japan, and what I would need in Germany to join HAMnet. Will investigate to install SuSE 6.2, the first SuSE Linux I used. Released around same time as Red Hat 5.2. Comments? Questions? -\u0026gt; Mastodon thread\n","date":"2024-12-29T03:49:25+09:00","image":"/posts/2024/12/tech-sidequests/20241224_172234445.cdb.jpg","permalink":"/posts/2024/12/tech-sidequests/","title":"Holiday season side quests"},{"content":"What\u0026rsquo;s this about? Since a while I was wondering how to setup a chain of LEDs, where each LEDs colour and brightness could be assigned by me. In Germany I got 2 LED chains which turned out to just allow pre-programmed colours and blink patterns. I\u0026rsquo;m looking for full flexibility, so I could also for example deploy them behind the monitor and have them shine in the dominant colours of the current movie scene.\nIn my social bubble, many are playing with these - but it seemed to me like the presentations were always on types of LEDs etc. instead of covering the basics. So this is what works for me, and which I can now extend from, in future setups.\nThe components I\u0026rsquo;m using now Controller: I planned initially to just control the LEDs plainly via USB, but now settled with a small controller box which contains an ESP32, join a wifi network and can be controled over network. These cost ~20EUR, I got this one. That controller would also be good to connect additional LED stripes. Power: my first idea was to use USB from a PC, but that won\u0026rsquo;t deliver enough power for more than a handful LEDs, maybe 10, depending on type/consumption. For now, I use a 12V/2A power supply which in the past powered a router. LEDs: A huge variety is available, common names are ws2811, ws2812, sk6812, tm1814, ws2813, ws2815. For starters, I got this, which is described as \u0026ldquo;WS2815, 3M 60 IP30, Black PCB\u0026rdquo;, meaning: WS2815 - chip 3M - 3 meter 60 - this kind of stripe has 60 LEDs per meter, so I have 180 LEDs Software: WLED is the software I use on the ESP32 based controller. Search terms for controllers like this: \u0026ldquo;wled controller esp32\u0026rdquo;. My controller was already flashed with the WLED software, last stable release. When looking up the hardware, some pieces have to match:\nthe voltage of the power supply has to be covered by the controller, and to fit the voltage required by the LEDs the maximum current of the power supply has to fit to the consumption of the LEDs. I consider my 12V/2A power supply slightly underpowered, but surprisingly the WLED software allows to configure the power supply specs, and then stays with brightness/effects below the limits! Setup The hardware arrived after 2 weeks. Not even soltering required. The LED chain had cables with connectors on both ends: one can be connected to the controller, the other could be used to chain other LED-chains. There was also a spare connector with loose ends: I simply connected these cables with the \u0026ldquo;controller\u0026rdquo;, and the LED-chain was connected. The power supply had a 3.5mm barrel plug, which I connected to the controller.\nAs the ESP32 contoller got power, it started offering a wifi network. On Linux, \u0026rsquo;nmcli dev wifi\u0026rsquo; showed the network, and with\n# nmcli dev wifi con WLED-AP password wled1234 # dhclient \u0026lt;wifi-interface\u0026gt; one joins the network. I got IP 4.3.2.2 assigned, and the ESP webserver was available at http://4.3.2.1 . From that interface, I configured the ESP to join the existing iot-wifi-network here. All of this could also easily be done from a mobile phone. The WLED pages are designed for mobile devices, but also have a \u0026ldquo;PC mode\u0026rdquo;, an interface optimized for desktops.\nI also got a matrix built from LEDs shipped, looking into that next.\nPower Running the 180 LEDs on 20% brightness, the power supply takes 0.1A here on the 100V network in Japan, so 9 or 10W are consumed. The power supply gets a bit warm. On full LED brightness, consumption goes up to 24W.\nMatrix The matrix is a 32 columns / 8 rows device, sold for 16 euro here. Product name: \u0026ldquo;DC12V WS2815 8x32\u0026rdquo;, so 12V powered, LED type is WS2815. Works really nicely, and WLED can also control it. The LEDs are switched as if they were located in a long stripe, but WLED then aligns them into a matrix - with the user configuring the 32x8 setting. WLED also knows many effects for these 2D-setups.\nLinks NeoPixel, a widely used library for controlling LEDs hyperhdr is a project to control stripes for ambient light behind monitors. Here is a description of their hardware. more minimal controller, project Comments? Questions? -\u0026gt; Mastodon thread\n","date":"2024-10-20T19:44:49+09:00","image":"/posts/2024/10/simple-setup-of-controllable-leds/20241020_082011854_leds_1024x.jpg","permalink":"/posts/2024/10/simple-setup-of-controllable-leds/","title":"A simple LED setup, controllable via wlan with WLED"},{"content":"What the ISP offers I\u0026rsquo;m using NTT flets as physical provider (fiber), and Biglobe for the packets ontop. Regarding IPv6, Biglobe allows on a webfrontent to request v6, which I did. Then they offer 2 options:\nIPv6オプション (\u0026ldquo;option\u0026rdquo;) IPv6オプションライト(\u0026ldquo;option light\u0026rdquo;) I was first on the former, then switched to the latter. Their site on this. With both of these options, I get IPv4 via PPPoE tunnel, and IPv6 plainly onto my ethernet interface facing towards the ONU/modem.\nBasic network setup of PPPoE, for IPv4 connectivity With that, we can swap both NVMe and boot from the new system. No network configured right now, so access via serial console, details here. The previous distro on my Star64 router was Ubuntu based and I used NetworkManager (NM), which seemed preferred by the distro, and I\u0026rsquo;m also somewhat familiar with NM. It made setting up a Wifi AccessPoint very simple. As a downside: I never got the PPPoE session properly integrated with NM, so I always controlled the PPPoE interface in executing \u0026ldquo;pon dsl-provider\u0026rdquo; to activate and \u0026ldquo;poff\u0026rdquo; to bring down the interface.\nThe Debian way for Bookworm and later seems to be systemd-networkd right now, so I\u0026rsquo;m going with that. My ethernet devices appear as \u0026ldquo;end0\u0026rdquo; and \u0026ldquo;end1\u0026rdquo;, and wifi as \u0026ldquo;wlx14f5f9517bcc\u0026rdquo;.\n### Remove the old interfaces file out of the way mv /etc/network/interfaces /etc/network/interfaces.save systemctl enable systemd-networkd cd /etc/systemd/network vi end1.network cat end1.network [Match] Name=end1 [Network] Description=Link to the LAN for my clients [Address] Address=fc00::1/64 [Address] Address=192.168.1.1/24 networkctl reload This the LAN facing interface, just setting a static v4 and v6 address. After rebooting, this becomes active, \u0026ldquo;networkctl\u0026rdquo; and \u0026ldquo;ip addr s\u0026rdquo; confirm.\nPPPoE and NAT setup The pppoeconf tool does the setup for us - it\u0026rsquo;s working outside of systemd-networkd and setting up pieces in /etc/network/interfaces, so pppoe is started after reboots.\npppoeconf # At this point we have internet access, and can set our clock: ntpdate ntp1.ptb.de Setting up IPv4 NAT for the LAN clients, and Wifi clients on 192.168.5.0/24 network:\necho \u0026#39;net.ipv4.ip_forward = 1\u0026#39; \u0026gt;/etc/sysctl.d/20-cust.conf echo \u0026#39;net.ipv6.conf.all.forwarding = 1\u0026#39; \u0026gt;\u0026gt;/etc/sysctl.d/20-cust.conf sysctl -p /etc/sysctl.d/20-cust.conf iptables -t nat -o ppp0 -A POSTROUTING -s 192.168.1.0/24 -j MASQUERADE iptables -t nat -o ppp0 -A POSTROUTING -s 192.168.5.0/24 -j MASQUERADE # Adding the last 2 commands to /etc/rc.local to survive reboots vi /etc/rc.local Manual triggering of the PPPoE session with \u0026ldquo;pon dsl-provider\u0026rdquo; and \u0026ldquo;poff\u0026rdquo; works then. The user/password for the pppoeconf tool are the ones used for the data ISP, here Biglobe. pppoe is then getting listed by networkctl, but is not fully managed by systemd-networkd:\n$ networkctl IDX LINK TYPE OPERATIONAL SETUP 1 lo loopback carrier unmanaged 2 end0 ether routable configured 3 end1 ether routable configured 4 wlx14f5f9517bcc wlan routable configured 5 ppp0 ppp routable unmanaged 5 links listed. IPv6 for router and LAN Now it\u0026rsquo;s getting ugly.. I struggle with this forever. All ISP\u0026rsquo;s seem to do this differently. One needs to find out what the ISP is offering, how to get v6 for the router, and then how to best make it available for the LAN.\nIn my case: first step is requesting v6 with the ISP, via a web frontend from the ISP. Once the ISP has enabled IPv6 from their side, the kernel recognizes announcements and configures interface end0 - even if the PPPoE session is down:\n[root@star64 60c ~]$ ip addr s [..] 2: end0: \u0026lt;BROADCAST,MULTICAST,UP,LOWER_UP\u0026gt; mtu 1500 qdisc mq state UP group default qlen 1000 link/ether 6c:cf:39:00:75:5f brd ff:ff:ff:ff:ff:ff inet6 2404:7a80:3820:3700:6ecf:39ff:fe00:755f/64 scope global dynamic mngtmpaddr noprefixroute valid_lft 2591978sec preferred_lft 604778sec inet6 fe80::6ecf:39ff:fe00:755f/64 scope link valid_lft forever preferred_lft forever From that point on, I can do IPv6 traffic from the router/star64.\nI then had a journey through various tools, trying to understand what the ISP is offering, and how to best make IPv6 available to the systems on the LAN. The inconvenient truth is that NAT with IPv6 is the only thing I got working. These are the options:\nOption 1: Subnet delegation, via systemd-networkd Best option would be the ISP to provide a delegation of a /56 or bigger, which could then be used from the LAN. The router would then plainly route through. I sniffed traffic on the end0 interface right after activating the interface, seems like I get. One would use \u0026ldquo;radvd\u0026rdquo; to advertise the network to the LAN clients, activate forwarding on the router. I had no luck with requesting subnet delegation.\nWith\n[Network] DHCP=yes [DHCPv6] PrefixDelegationHint=::/56 in my end0.network interface config file. The ISP facing interfact ends then up like this:\n2: end0: \u0026lt;BROADCAST,MULTICAST,UP,LOWER_UP\u0026gt; mtu 1500 qdisc mq state UP group default qlen 1000 link/ether 6c:cf:39:00:75:5f brd ff:ff:ff:ff:ff:ff inet6 2404:7a80:3820:3700:6ecf:39ff:fe00:755f/64 scope global dynamic mngtmpaddr noprefixroute valid_lft 2591944sec preferred_lft 604744sec inet6 fe80::6ecf:39ff:fe00:755f/64 scope link valid_lft forever preferred_lft forever ..which looks like I got a /64 network. But then, adding\n[Network] IPv6SendRA=yes DHCPv6PrefixDelegation=yes to the interface facing to the LAN, I do not get networks announced on the systems connected on the LAN side.\nOption 2: Subnet delegation, via wide-dhcpv6-client This is another option to request a subnet from the ISP, instead of having systemd-networkd request the subnet. Package wide-dhcpv6-client is available in Debian repos. In removing file /etc/systemd/network/end0.network, I can take interface end0 out of systemd-networkd\u0026rsquo;s management, and try wide-dhcpv6-client:\n[root@star64 58c ~]$ cat /etc/wide-dhcpv6/dhcp6c.conf interface end0 { send ia-pd 1; send ia-na 1; }; id-assoc na 1 { }; id-assoc pd 1 { # You should also try /56 and /64 here, but for me no # difference, unfortunatelly. prefix ::/48 infinity; prefix-interface end0 { sla-id 1; sla-len 8; }; prefix-interface end1 { sla-id 2; sla-len 8; }; }; Yet, after starting /etc/init.d/wide-dhcpv6-client, my interface just gets a link-local address. Not even the SLAAC configured 2404[..] address. I can run wide-dhcpv6 in debug mode, and it seems like no PD is offered by the ISP:\n/usr/sbin/dhcp6c -Pdefault -c /etc/wide-dhcpv6/dhcp6c.conf -D -f end0 [..] Oct/06/2024 15:21:25: dhcp6_get_options: get DHCP option status code, len 16 Oct/06/2024 15:21:25: status code: no prefixes Oct/06/2024 15:21:25: client6_recvadvert: server ID: 00:03:00:01:00:[..], pref=-1 Oct/06/2024 15:21:25: client6_recvadvert: advertise contains NoPrefixAvail status Dumping the traffic and investigating with tshark, I see my request for delegation, and the denial from the other side:\n$ tshark -r 2_dump_wide -V| [..] Identity Association for Prefix Delegation Option: Identity Association for Prefix Delegation (25) Length: 41 IAID: 00000001 T1: 0 T2: 0 IA Prefix Option: IA Prefix (26) Length: 25 Preferred lifetime: infinity Valid lifetime: infinity Prefix length: 48 Prefix address: :: [..] Status code Option: Status code (13) Length: 16 Status Code: NoPrefixAvail (6) Status Message: NOPREFIX-AVAIL Also re questing /56 or /64 subnets makes no difference.\nOption 3: Subnet delegation, via network scripts After installing package isc-dhcp-client, I can also use\niface end0 inet6 auto dhcp 1 request_prefix 1 in /etc/network/interfaces to get a prefix. After a \u0026ldquo;ifup end0\u0026rdquo;, I get the 2404[..] address, but no network.\nOption 4: DHCPv6 relay via wide-dhcpv6-relay The idea is here to have\nLAN clients issue DHCPv6 requests have the router relay these to the ISP facing network and have the reply also relayed to the LAN client, so it can configure it\u0026rsquo;s interface then also ensure routing/proxying for normal v6 traffic from the LAN client After installing package wide-dhcpv6-relay on the router, it seems like relaying is done, just that the IPv6 interface on the LAN client is not getting an address:\n# on router [root@star64 59c network]$ dhcp6relay -D -f -r end0 end1 Oct/08/2024 07:48:15: main: dhcp6relay started Oct/08/2024 07:49:35: relay6_recv: from fe80::9ad:54d1:c1ef:d93%end1, size 85 Oct/08/2024 07:49:35: relay6_recv: received solicit from fe80::9ad:54d1:c1ef:d93%end1 Oct/08/2024 07:49:35: copy_option: set relay message (len 85) Oct/08/2024 07:49:35: copy_option: set interface ID (len 4) Oct/08/2024 07:49:35: relay_to_server: relay a message to a server ff05::1:3 Oct/08/2024 07:49:36: relay6_recv: from fe80::9ad:54d1:c1ef:d93%end1, size 85 Oct/08/2024 07:49:36: relay6_recv: received solicit from fe80::9ad:54d1:c1ef:d93%end1 Oct/08/2024 07:49:36: copy_option: set relay message (len 85) Oct/08/2024 07:49:36: copy_option: set interface ID (len 4) Oct/08/2024 07:49:36: relay_to_server: relay a message to a server ff05::1:3 # on LAN client 2: enp0s31f6: \u0026lt;BROADCAST,MULTICAST,UP,LOWER_UP\u0026gt; mtu 1500 qdisc fq_codel state UP group default qlen 1000 link/ether 98:fa:9b:91:06:37 brd ff:ff:ff:ff:ff:ff inet 192.168.1.2/24 brd 192.168.1.255 scope global noprefixroute enp0s31f6 valid_lft forever preferred_lft forever inet6 fe80::9ad:54d1:c1ef:d93/64 scope link noprefixroute valid_lft forever preferred_lft forever Option 5: DHCPv6 relay via dibbler-relay Also with this, the interface on the LAN system is not getting an address:\n# tail -f /var/log/dibbler/dibbler-relay.log 14:41 Relay Notice Received SOLICIT on end1/3,trans-id=0xea650, 6 opts: 14 3 39 6 1 8 14:41 Relay Debug Interface-id option added before relayed message. 14:41 Relay Notice Relaying encapsulated SOLICIT message on the end0/2 interface to multicast (ff05::1:3) address, port 547. 14:41 Relay Debug Accepting messages. 14:42 Relay Debug Received 85 bytes on the end1/3 interface (socket=0, addr=fe80::9ad:54d1:c1ef:d93, port=547). 14:42 Relay Notice Received SOLICIT on end1/3,trans-id=0xea650, 6 opts: 14 3 39 6 1 8 14:42 Relay Debug Interface-id option added before relayed message. 14:42 Relay Notice Relaying encapsulated SOLICIT message on the end0/2 interface to multicast (ff05::1:3) address, port 547. 14:42 Relay Debug Accepting messages. Option 6: NDP proxy Next best option is the router acting as IPv6 NDP proxy: this would proxy IPv6 config requests from the LAN to the ISP, and also answers. I installed ndppd and played around, without success.\nOption 7: IPv6 NAT/masquerading My last resort: a static IPv6 IP on the end1 interface which is facing the LAN, and static IPv6 IP on the LAN system. The router is then getting configured for NAT:\n# ip6tables -t nat -A POSTROUTING -s fc00::2 -j MASQUERADE This state is not satisfying.. we should setup an exhaustive list of debug steps for IPv6, to find out what an ISP is offering and then investigate the options.\n","date":"2024-10-06T16:44:22+09:00","image":"/posts/2024/10/ipv6-for-the-lan-blues/20240929_star64_debian.jpg","permalink":"/posts/2024/10/ipv6-for-the-lan-blues/","title":"Bringing IPv6 into the LAN Blues"},{"content":"Note: I\u0026rsquo;ve taken out the IPv6 pieces here and moved them to a dedicated blog article.\nSince the beginning of 2024, I run my own router. I\u0026rsquo;m using NTT flets GBit fiber as physical provider, and Biglobe as ISP ontop - other ISP\u0026rsquo;s do the IPv6 things differently. I started with 2 OpenWRT routers, after both of them went defective, I run the Star64 RISC-V board as router, with this functionality:\nTalking PPPoE with the ISP, and provide IPv4 NATing for the LAN systems, currently the Thinkpad T590 Wireless access AP for my mobile phone and other devices like smartplugs and Thinkpads Nginx/PHP hosting Librespeed, allowing to quickly verify bandwidth and latency to LAN and Wifi clients IPv6 access to the LAN The Star64 here is running Linux from an NVMe card, plugged into a PCIe card, plugged into the PCIe slot of the Star64. This became possible after a firmware update. Until now I was running a distro based on \u0026ldquo;Ubuntu 23.04 (Lunar Lobster)\u0026rdquo; - which is no longer receiving updates. Time to upgrade to Debian.\nInstalling Debian My old setup was on an NVMe, with a single GPT partition, with ext4 file system. I have a USB3 connected NVMe enclosure and another NVMe, which I could mount on the running system and setup the new distro. Then did shutdown the star64, swap both NVMe\u0026rsquo;s, and boot from the new system. For debugging, that was done several times. I also needed the serial console access to the Star64 for debugging when no network is available.\nWhy Debian as new distro? Because it\u0026rsquo;s just uncomplicated for updating to new major or minor versions. First issue: to find mirrors with RISC-V. Ideally I wanted to install Debian 12/Bookworm, as I do not need latest features, wanted rather stability and packages not much in flux/frequently updated. I experimented with debootstrap to get a basic Debian system together: available on the old Ubuntu, also as package on Fedora 40. The Fedora package worked well, and I found RISC-V available as Debian sid (unstable) and Trixie, the Debian testing variant right now. Creating partition, file system:\n# parted /dev/sda mklabel gpt # partprobe # parted /dev/sda mkpart primary ext4 Start? 1 End? -1 # mkfs.ext4 -L trixie /dev/sda1 # mkdir -p /mnt/tmp # mount /dev/sda1 /mnt/tmp I run debootstrap on Fedora, then tar up the file system, and untar it to the new NVMe:\n# debootstrap --arch=riscv64 sid debian-sid/ http://mirror.23m.com/debian/ # debootstrap --arch=riscv64 testing debian-testing/ http://mirror.23m.com/debian/ debootstrap --arch=riscv64 trixie debian-trixie/ http://mirror.23m.com/debian/ tar cfP debian-trixie-riscv.tar debian-trixie scp debian-trixie-riscv.tar star: # On the Star64: cd /mnt/tmp \u0026amp;\u0026amp; tar xfP /home/chris/debian-trixie.tar mv debian-trixie/* . \u0026amp;\u0026amp; rmdir debian-trixie/ mount proc /mnt/tmp/proc -t proc mount sysfs /mnt/tmp/sys -t sysfs cp /etc/hosts /etc/fstab /mnt/tmp/etc/ cp -r /boot /mnt/tmp/ cp -r /usr/lib/modules /mnt/tmp/usr/lib/ sync blkid chroot /mnt/tmp /bin/bash echo \u0026#39;nameserver 8.8.8.8\u0026#39; \u0026gt;/etc/resolv.conf echo star64.local \u0026gt;/etc/hostname # configure LABEL=trixie of rootfs for boot and in fstab. # For extlinux.conf, I also removed \u0026#34;net.ifnames=0\u0026#34; which was # used by the old distro. vi /etc/fstab vi /boot/extlinux/extlinux.conf apt update \u0026amp;\u0026amp; apt dist-upgrade apt install ntpdate openssh-server sudo lsof strace pppoeconf \\ iptables binutils rsyslog tcpdump useradd -m chris usermod -s /usr/bin/bash chris passwd passwd chris # I prefer a nondefault SSH port vi /etc/ssh/sshd_config echo \u0026#39;export EDITOR=vi\u0026#39; \u0026gt;\u0026gt;/etc/bash.bashrc echo \u0026#34;alias su=\u0026#39;sudo su -\u0026#39;\u0026#34; \u0026gt;\u0026gt;/etc/bash.bashrc Basic network setup With that, we can swap both NVMe and boot from the new system. No network configured right now, so access via serial console, details here. On the Ubuntu based distro I used NetworkManager, some pros and cons:\npro: the setup of the Wifi AP is extremely easy con: I did not get pppoe properly setup with NetworkManager The Debian suggestion seems to be systemd-networkd right now, so going with that. My ethernet devices appear as \u0026ldquo;end0\u0026rdquo; and \u0026ldquo;end1\u0026rdquo;, and wifi as \u0026ldquo;wlx14f5f9517bcc\u0026rdquo;.\n# Remove the old interfaces file out of the way mv /etc/network/interfaces /etc/network/interfaces.save systemctl enable systemd-networkd cd /etc/systemd/network vi end1.network cat end1.network [Match] Name=end1 [Network] Description=Link to the LAN for my clients [Address] Address=fc00::1/64 [Address] Address=192.168.1.1/24 This the LAN facing interface, just setting a static v4 and v6 address. After rebooting, this becomes active, \u0026ldquo;networkctl\u0026rdquo; and \u0026ldquo;ip addr s\u0026rdquo; confirm.\nPPPoE and NAT setup The pppoeconf tool does the setup for us - it\u0026rsquo;s working outside of systemd-networkd and setting up pieces in /etc/network/interfaces, so pppoe is started after reboots.\npppoeconf # At this point we have internet access, and can set our clock: ntpdate ntp1.ptb.de Setting up IPv4 NAT for the LAN clients, and Wifi clients on 192.168.5.0/24 network:\necho \u0026#39;net.ipv4.ip_forward = 1\u0026#39; \u0026gt;/etc/sysctl.d/20-cust.conf echo \u0026#39;net.ipv6.conf.all.forwarding = 1\u0026#39; \u0026gt;\u0026gt;/etc/sysctl.d/20-cust.conf sysctl -p /etc/sysctl.d/20-cust.conf iptables -t nat -o ppp0 -A POSTROUTING -s 192.168.1.0/24 -j MASQUERADE iptables -t nat -o ppp0 -A POSTROUTING -s 192.168.5.0/24 -j MASQUERADE # Adding the last 2 commands to /etc/rc.local to survive reboots vi /etc/rc.local Manual triggering of the PPPoE session with \u0026ldquo;pon dsl-provider\u0026rdquo; and \u0026ldquo;poff\u0026rdquo; works then. The user/password for the pppoeconf tool are the ones used for the data ISP, here Biglobe. pppoe is then getting listed by networkctl, but is not fully managed by systemd-networkd:\n$ networkctl IDX LINK TYPE OPERATIONAL SETUP 1 lo loopback carrier unmanaged 2 end0 ether routable configured 3 end1 ether routable configured 4 wlx14f5f9517bcc wlan routable configured 5 ppp0 ppp routable unmanaged 5 links listed. Wifi AP With NetworkManagers, this is as easy as just defining a new connection with \u0026ldquo;nmcli\u0026rdquo; and parameters. On Debian, I use now a dedicated dhcpd and hostapd.\napt install isc-dhcp-server hostapd [root@star64 60c network]$ cat /etc/dhcp/dhcpd.conf ddns-update-style none; subnet 192.168.5.0 netmask 255.255.255.0 { interface wlx14f5f9517bcc; # your interface name here range 192.168.5.10 192.168.5.20; # desired ip range option routers 192.168.5.1; option subnet-mask 255.255.255.0; option broadcast-address 192.168.5.255; option domain-name-servers 8.8.8.8; } [root@star64 60c network]$ cat /etc/hostapd/hostapd.conf interface=wlx14f5f9517bcc channel=6 ieee80211n=1 hw_mode=g ssid=fluxnet-iot-ap wpa=2 wpa_passphrase=somepassphrase wpa_key_mgmt=WPA-PSK rsn_pairwise=CCMP auth_algs=1 [root@star64 60c network]$ echo \u0026#39;INTERFACESv4=\u0026#34;wlx14f5f9517bcc\u0026#34;\u0026#39; \u0026gt;\u0026gt;\\ /etc/default/isc-dhcp-server [root@star64 60c network]$ systemctl enable --now hostapd isc-dhcp-server Nebula After compiling an own kernel for the Star64, I also have now the tun kernel module which was missing so far, and can run the Nebula client on the system.\n","date":"2024-09-29T16:18:32+09:00","image":"/posts/2024/09/debianizing-star64-and-ipv6-for-lan/20240929_star64_debian.jpg","permalink":"/posts/2024/09/debianizing-star64-and-ipv6-for-lan/","title":"Debianizing the Star64 board"},{"content":"So.. I got a new monitor, the INNOCN 32M2V.\nWhich resultion can I drive it with, from my Thinkpad T590?\nInvesting the Thinkpad T590 capabilities My daily driver at home is a Thinkpad T590, with Intel i7-8665U CPU.\nFirst limitation is the GPU, the Thinkpad has a Intel UHD Graphics 620, which is imbedded in the processor. Intel lists for this cpu a maximum of 4096x2304@24Hz for HDMI output, and 4096x2304@60Hz via DisplayPort. The T590 has a builtin HDMI1.4b port, that allows 4096x2160@24Hz, 3840x2160@30Hz, 2560x1440@60Hz and 1920x1080@120Hz. Via HDMI connection to the monitor, I get these listed as available in the output of \u0026lsquo;wlr-randr\u0026rsquo;. Sadly, the T590 does not have a DisplayPort. DisplayPort1.4 would be capable of: 3440x1440@240Hz/144Hz/360Hz, 4096x2160240Hz with DSC, 3840x2160@120Hz / 10 Bit Chroma / HDR The Thinkpad has 2 USB-C ports, with different capabilities:\nA) can handle network + usb-devices when connected to a Thinpad docking station, but supports lower graphic modes B) can not handle network + usb-devices, but allows higher graphic modes. With lsusb, we see the modes of the USB-C ports:\n[chris@космос ~]$ lsusb -t /: Bus 001.Port 001: Dev 001, Class=root_hub, Driver=xhci_hcd/12p, 480M |__ Port 003: Dev 002, If 0, Class=Hub, Driver=hub/5p, 480M |__ Port 002: Dev 004, If 0, Class=Human Interface Device, Driver=usbhid, 12M |__ Port 005: Dev 006, If 0, Class=Human Interface Device, Driver=usbhid, 12M |__ Port 005: Dev 006, If 1, Class=Human Interface Device, Driver=usbhid, 12M |__ Port 005: Dev 006, If 2, Class=Human Interface Device, Driver=usbhid, 12M |__ Port 008: Dev 003, If 0, Class=Video, Driver=uvcvideo, 480M |__ Port 008: Dev 003, If 1, Class=Video, Driver=uvcvideo, 480M |__ Port 010: Dev 005, If 0, Class=Wireless, Driver=btusb, 12M |__ Port 010: Dev 005, If 1, Class=Wireless, Driver=btusb, 12M /: Bus 002.Port 001: Dev 001, Class=root_hub, Driver=xhci_hcd/6p, 10000M |__ Port 002: Dev 002, If 0, Class=Video, Driver=uvcvideo, 5000M \u0026lt;- port A |__ Port 002: Dev 002, If 1, Class=Video, Driver=uvcvideo, 5000M \u0026lt;- port B |__ Port 002: Dev 002, If 2, Class=Audio, Driver=snd-usb-audio, 5000M |__ Port 002: Dev 002, If 3, Class=Audio, Driver=snd-usb-audio, 5000M |__ Port 002: Dev 002, If 4, Class=Vendor Specific Class, Driver=[none], 5000M |__ Port 002: Dev 002, If 5, Class=Human Interface Device, Driver=usbhid, 5000M 5GBit/sec is listed for both ports, so USB C 3.1 Gen 1. \u0026rsquo;lsusb -t\u0026rsquo; is not showing the traffic which goes over the ports for video.\nInvestigating the monitor capabilities The next important data is what the INNOCN 32M2V monitor itself can handle. This model has 4k resolution, a 32\u0026quot; screen, frame rates up to 144Hz. Has HDMI2.1, DisplayPort1.4 and USB-C connectors, speaker buildin. The model was released in April 2024, brightness sensor, HDR picture mode. 10.7kg weight.\nThe monitor does offer 3840x2160@120Hz. Both HDMI2.1 and DisplayPort1.4 support that. No details on supported modes via USB-C found.\nUsing Lenovo Gen2 dock The \u0026ldquo;Lenovo 40AS ThinkPad USB-C Dock Gen2\u0026rdquo; has itself a DisplayPort 1.4, and HDMI. I can use USB-C from the Thinkpad to the dock, and from there DisplayPort or HDMI to the monitor, allowing the same modes: 3840x2160@30Hz, 2560x1440@60Hz and 1920x1080@120Hz.\nUsing Lenovo Thunderbolt 4 dock This dock is quite new, my employer had a spare one, available for me. This dock also has DisplayPort and HDMI available as output to the monitor, and allows Thunderbolt 4 to the Thinkpad, which gives me new modes when using the USB-C port B) of the Thinkpad: 3840x2160@60Hz, 2560x1440@120Hz and 1920x1080@120Hz. Also, this dock comes with a heavy 300W power supply. Trying to connect a smaller 60W power supply, the dock generates own output to the screen, asking you to connect the right/compatible power supply.\nSo, trying out these modes practically, brings up these graphs:\n\u0026ndash;\nI\u0026rsquo;ll settle with 2560x1440@120Hz for now. Of course, if the Thinkpad had DisplayPort1.4 and could drive the monitor directly, that would be best - but I won\u0026rsquo;t buy a new computer just for that.\nIt\u0026rsquo;s 2024, shouldn\u0026rsquo;t such things be easier to find out?\nComments? Mastodon thread\n","date":"2024-09-02T21:03:17+09:00","image":"/posts/2024/09/monitor-resolution-research/20240810_232248775_monitor_HDR_test2.jpg","permalink":"/posts/2024/09/monitor-resolution-research/","title":"Searching the best resolution for a monitor"},{"content":"Choosing a monitor is hard! When you buy shoes, you try them out.. but you just know for real after wearing them for some weeks. We might see monitors at stores, or try to get advice from others - but everybody has different use styles, so it\u0026rsquo;s hard to decide for a model.\nWhat I use the monitor for:\nworking with multiple terminals, but not more than 3 or 4 on one screen. I heavily switch between virtual screens. I use webbrowsers, when these are in Japanese, I often have to increase the font size more rarely: watching a movie or a drama, or playing The previous monitor In 2016, when I came to Japan, I got a 27\u0026quot; monitor at Bic Camera, one of the huge electronic stores in Tokyo. This was an I-O DATA LCD-MF276XDB, introduced 2014, with ADS panel. Maximum resolution 1920x1080, refresh rate 60Hz, builtin speakers. I use the monitor heavily, but was looking for something new. At work, I got a Lenovo one, liked the 2560x1440 resolution, and more vivid colours.\nOne thing was holding me back from getting a new monitor: I do not want to have 2 monitors around, this flat is small. In Japan, we have Hardoff stores to sell electronics, and online places like Mercari, but I did not feel to good about these options.\nGetting the old monitor to good hands This was quite a stunt! A friend living near Osaka was looking for a monitor and was willing to take it - but how to get it there, from Tokyo? Before we started to research shipping of such a rather big item, 2 friends from Germany moved into the picture. They were on a tour through Japan: Hokkaido, Hakodate, Tokyo, Osaka/Kyoto. They graciously accepted to take the monitor along with them to Osaka - creating a nice reason to meet up with the friend in Osaka, spend some time and also hand over the monitor.\nComparing models I had walked around at Bic Camera some weeks ago. Their selection of business monitors is big, and they even opened up now extra spaces for gaming focused computers/monitors. But for taking time and comparing features, doing internet research seems more promising.\nSo I spent 3 evenings on researching options.. quite a rabbit hole! For a start, I read articles like this one, which say one should not start with resolutions lower than 4k, and 120Hz refresh rates.\nRather than specific models, I started to decide on the single features. I could then do a more specific search on amazon.co.jp to see which models are available.\nRelevant features, and reasoning:\nSpeakers: I need them builtin, as my desk is not big, and already fiber modem, own router (Star64), Thinkpad and their power supplies are stacked up Overall size: I tried to get again a 27\u0026quot; model again Resolution: Tricky. Wanted 2560x1440 at least, as from my experiences at the company I knew I could use it directly without scaling. Native 4k would be to small for me - all applications and browser fonts would be way to small by defailt. But I\u0026rsquo;m not buying a monitor every year, not getting a 4k these days would not make sense. Refresh rate: 60Hz is the default since the world switched from CRT\u0026rsquo;s to flat screen monitors. But nowadays, also 120Hz and 144Hz models are around: advertised for gaming. Some also say they are better for browsing. With these, I came down to a selection of 5 monitors, looked at further factors, and down to 2, and took the INNOCN 32M2V: 4k resolution, 32\u0026quot; screen, up to 144Hz, 100,250円 (~625EUR). Has HDMI2.1, DisplayPort1.4 and USB-C connectors, speaker buildin, model released in April 2024, brightness sensor, HDR picture mode.\nHeavier than others, slight downide: 10.7kg.\nDriving the monitor While I can use now more convenient modes than with the old monitor, I can not drive the monitor with all it offers.\nSee for details.\nFor normal usage, I have settled with 2560x1440@120Hz for now. When I refresh my hardware in the future, I will reevaluate the modes.\nUsing the monitor My first impression was not good, the light sensor was on and took the brightness down way to much: to 6% out of 100, increasing that helped. I\u0026rsquo;m running now with 20%, the monitor uses 32W then. Running 4k and activating HDR mode takes brightness to 60%, and the monitor consumes 42W then.\nSeems like a quickstart document is available, but no full manual. Cheat sheet for monitor settings:\nprofessional settings DCI-P3 Mode: mode mostly used in digital photography C.T settings: Colour temperature picture settings DCR: Dynamic Contrast Ratio HDR: High Dynamic Range PIP/PBP mode: allows picture in picture; multiple sources to be shown at the same time other auto power: power on computer via usb-c after power loss? eyeshield remind: ? light sensor: adjust brightness to environment usb upgrade: used when upgrading monitor firmware via usb-storage Was it worth the upgrade? Have the single points played out? I\u0026rsquo;m happy with the upgrade, even though I can not use the highest modes the monitor offers right now. TODO: hook up the Macbook pro M2, with Asahi Linux and MacosX, see which modes are offered. Likely usb-c is again the bottle neck, so probably same modes will be available.\nQuestions? Comments? -\u0026gt; Mastodon thread\n","date":"2024-08-10T16:12:20+09:00","image":"/posts/2024/08/choosing-a-monitor-in-2024/package.jpg","permalink":"/posts/2024/08/choosing-a-monitor-in-2024/","title":"Monitor refresh in 2024"},{"content":"While last years summer in Germany became the Retro summer, this time I took the RTL-SDR software defined radio with me for playing. RTL-SDR is a usb2 stick which you can plug into a computer, attach an antenna, and then have a look what\u0026rsquo;s \u0026ldquo;in the air\u0026rdquo;.\nSoftware I\u0026rsquo;m using the antennas which I got with the RTL-SDR device: 2 teleskope sticks with length of up to 1m each. Main issue: for good reception, I place the antenna directly on the attic. But I want to be on the lower floors for sitting down and using the SDR. So I setup an older Thinkpad x200 with Debian Bookworm, and did setup that with the RTL-SDR device and the antenna on the attic.\nMy wifi does not cover the attic. I played with a set of power-over-ethernet devices, but then layed down a cat5 cable with RJ45 connector to the Thinkpad.\nTo use the SDR remotely, I used this software:\nOpenWebRX: this software is directly built to be used remotely, from a browser. Using ~300kbyte/sec when accessed with a single client. OpenWebRX+ is a fork with more functions, but I did not get it to work with my RTL-SDR\u0026rsquo;s rtl_433: command line utility for scanning for sensor data. Output directly to the terminal. SDR++ is a great software. I used \u0026lsquo;rtl_tcp\u0026rsquo; on the system with the RTL-SDR device connected, and then had SDR++ connect to that. Works ok - but used so much wifi bandwidth that other clients on the wifi network were severely impacted, i.e. network streaming getting unusable Which software for what? Which signals did you find? In Japan I\u0026rsquo;m in a mansion with a metal frame, having bad reception. Here it\u0026rsquo;s better, especially once I mounted the antenna onto the roof.\nOpenWebRX is really nice. receiverbook.de is a directory for freely accessible OpenWebRX instances - so people setting their up and offering it for public access over the internet, kind of a looking glass on what\u0026rsquo;s \u0026ldquo;in the air\u0026rdquo; at various places. Biggest downside for me: OpenWebRX does not allow to freely set the frequency to tune in, but the operator has to create \u0026ldquo;profiles\u0026rdquo; which just span a quite narrow band. Clients can select one of these profiles and then just select a frequency within the limits of that \u0026ldquo;profile\u0026rdquo;. OpenWebRX+ is supposed to allow to freely change the frequency, but at least for me it seemed like OpenWebRX+ did not get any signals from my RTL-SDR device.\nrtl_433 is also nice, it can decode data from sensors. You have to lookup frequencies which sensors potentially could use, and then see all kinds of sensors, for example:\ntyre air pressure sensors meat temperature sensors plant moisture sensor (69C sounds to hot..) remote controls, reporting which buttons were pressed weather stations (one constantly reporting \u0026ldquo;-102.4 C\u0026rdquo; temperature.. broken?) remote control for gas heater SDR++ is great, but with this high bandwidth usage, all involved systems should be on wired network.\n![CQ Ham Radio mag](sdr/0 CQ Ham Radio magazine with FT8 addon.jpg) ![Activity on freenet range](sdr/Activity on freenet range.jpg) ![ADS-B decoding](sdr/ADS-B planes decoding.jpg) ![ADS-B map](sdr/ADS-B planes map.jpg) ![Antenna on roof](sdr/Antenna on the roof.jpg) ![Decoding FT8](sdr/Decoding FT8 signals.jpg) ![Decoding sensor data](sdr/Decoding sensor data.jpg) ![SDR++ interface](sdr/The SDR++ interface.jpg)\nAntenna, receiving signals This house here has a wood frame, but of course putting up the antenna as high as possible helps with reception. Eventually I mounted it ontop of the roof, requiring the entrance to the attic stay open - so just a temporary setup. What I found/identified:\nUKW radio of course, that can be received even under bad conditions German citizens band (CB) around 27Mhz: I did not hear anything in that area, quite surprising. Freetalk area around 149Mhz: there was a bit of activity. The area can be used without HAM license, like CB. ADS-B readings: airplanes reporting flight number and location FT8: a digital mode, very interesting field. There are frequency areas in many bands which are often used for FT8 now. Around 28.074Mhz (10m band), I received FT8 messages from all over Europe, also Saudi Arabia, Russia. OpenWebRX can directly decode FT8, I used that. On the trip to Germany, I read the Japanese \u0026ldquo;CQ Ham Radio July 2024\u0026rdquo; magazine, coming with an addon magazine with FT8 details.\nI played with an own antenna: 2 copper wires of 5m lenght, 1.5mm diameter. This did worse than the factory built antenna, so I used that again.\nKiwiSDR Looking for OpenWebRX alternatives, I found Kiwisdr. They have a web interface with the same roots as OpenWebRX, but unfortunatelly they only run on their own hardware: these are small boxes based on an SDR and a BeagleBone board, and a GPS receiver. You connect it to your antenna, rj45 network cable, and can use it. Costs ~400Eur. The Kiwi frontend has features like direct frequency selection. Most cool feature: you can make your KiwiSDR available for others over the internet. When you want to find out from where a signal is sent, you can use other public KiwiSDR installations for that: with GPS their time is tuned with high precision, and having multiple Kiwi look at the time difference in reception of the same signal helps to locate the source!\nI found this great presentation of the features of the Kiwi web frontent. A publicly reachable Kiwi installation: Click the link, zoom fully out (press middle mouse button, move mouse/trackpointer fully down) to see what\u0026rsquo;s going on at all of the bands This is an introductory video, helping to differentiate various modulations of signals. Future When back in Japan, I plan to look more at sensors there, maybe I manage to read the metrics of my power meter. I got a second RTL-SDR which I will keep in Germany, maybe constantly connected, ready to be used over the internet. Setting up a pole to properly mount an antenna would require works on the roof again, I want to prevent that for now. Will look around for other compact antennas though.\nAs for my location in Japan: with the metal frame mansion, my options to receive signals from far away are quite limited. Another HAM operator (I assume he is one) in my building has stuck 2 telescope sticks out of his balcony, and has a cable/dipole antenna between the tips of the 2 sticks.\nNice idea!\nPorting features from the Kiwi webfrontend over to OpenWebRX would make a nice project.\nblinry\u0026rsquo;s Fifty Things you can do with a Software Defined Radio More RTL-SDR notes Questions? Comments? -\u0026gt; Mastodon thread\n","date":"2024-07-28T07:23:55+09:00","image":"/posts/2024/07/playing-with-rtl-sdr/sdrpp.jpg","permalink":"/posts/2024/07/playing-with-rtl-sdr/","title":"Playing with Software Defined Radio (SDR)"},{"content":"What\u0026rsquo;s this about? Colleagues asked: \u0026ldquo;Hey Chris, so what\u0026rsquo;s actually in that mobile HAM radio device, what do you get for 550EUR?\u0026rdquo; - I\u0026rsquo;ll sum that up here. I\u0026rsquo;m coming back to the hobby after 20 years, so also discovering what\u0026rsquo;s possible now. 20 years ago, I used CB-radio. After getting HAM radio licensed here in Japan, I got a decent HAM radio transmitter for hand held use, a Kenwood TH-D75. The article is not specific for the TH-D75 though, the features are also in other modern devices.\nLet\u0026rsquo;s have a look!\nPlain listening Everybody knows \u0026lsquo;world receivers\u0026rsquo;: these are a kind of radio, covering many bands of the frequency spectrum. When using these, you first select a \u0026lsquo;band\u0026rsquo; (the rough category of the frequency spectrum), and then fine tune the frequency you want to receive with a knob or control dial. The TH-D75 can do the same, but extended: instead of single bands, you can choose the frequency freely between 0.1Mhz and 524Mhz - that\u0026rsquo;s much more than world-radios.\nThe voice, music or data are \u0026rsquo;encoded\u0026rsquo; into the radio wave, the way of encoding is called modulation. For successful transmission of voice or data, sender and transceiver have to be configured with the same modulation. The world receiver knows AM (Amplitude Modulation) or FM (Frequency Modulation). The TH-D75 allows additional simple modulations: SSB (single-sideband), CW (\u0026lsquo;morse code\u0026rsquo;).\nA further difference: world receivers are listening \u0026ldquo;in one place\u0026rdquo;, so on one frequency. The TH-D75 can operate \u0026ldquo;in two places\u0026rdquo;: so you can for example constantly listen on one frequency (for example as you know someone will reach out to you there), while at the same time scanning other frequencies. Or, you could listen to UKW radio of your home area, while also monitoring a second frequency as you expect someone to reach out to you on that frequency. Basically 2 radios in one!\nSending The TH-D75 can also send, so you can communicate with others. The antenna and electronics of the device are optimized for certain select frequencies. 3 models of the TH-D75 exist: the models for Japan and Europe are 2-band-transceivers, they can send on the 70cm and 2m wavelength bands. The American model has an additional band. I graphed the areas of the three models here:\nThese differences are born in how these areas assigned frequencies, and has some funny implications: as these radio transceivers can send signal, they have the potential to interfere with all kinds of communications i.e. of police, firefighters and so on. Thus, the devices need to pass technical certifications for the country where they get used (and additionally one needs a HAM radio license). The frequency differences lead to for example the American model not being licensed in Japan, so you can also not easily buy the device in Japan when on holiday, and then use it in Europe. With that, also differences in pricing are possible, and exactly that can be seen for the TH-D75: I payed in Japan 89760円(~530EUR), while the European model is sold for 879EUR.\nThe difference in price makes sense though: Japan has seen much less inflation in the last years than Europe or even America, also wages are lower here than in Europe. JVC/Kenwood, the company producing the device is based in Japan.\nHow easy are receiving/sending/scan? Plain receiving and sending in AM/FM/SSB is a basic functionality, how hard is it to use? Without reading the manual, I managed to switch between the AM/FM/SSB modes, and change the frequency. Frequency change can be done simply with one of the 2 dials on the device - but I would always stay within a quite narrow frequency area. So one can not simply \u0026lsquo;dial\u0026rsquo; trough all frequencies between 0 and 500Mhz. That huge range is divided into \u0026lsquo;bands\u0026rsquo;, and already switching these needed the manual. Also \u0026lsquo;scans\u0026rsquo; are important: having the TH-D75 change probe frequencies, until a signal (i.e. a radio station) is found.\nThe third important function is \u0026lsquo;squelch\u0026rsquo;, so \u0026ldquo;which signal level needs to be found on the frequency to have the scan recognize it as \u0026lsquo;signal\u0026rsquo;\u0026rdquo;. Also for this I needed the manual.\nDigital modes AM, FM, SSB are mostly used to transport voice, are a long time around, can also since more than 100 years ago be implemented with normal electronic circuits. Mode \u0026ldquo;CW\u0026rdquo; is a bit special: it\u0026rsquo;s also known as \u0026ldquo;morse code\u0026rdquo;, so requires more skills for sending/listening, but can still be used under conditions when the other modes like SSB are no longer possible.\nModern transceivers also cover digital modes, the TH-D75 can transmit with D-STAR. That\u0026rsquo;s a digital mode, so for example voice gets first digitalized and then modulated onto radio waves. Such transmissions can be used under conditions when i.e. SSB is no longer possible. This also allows usage of \u0026ldquo;digital repeaters\u0026rdquo;: if 2 stations can not reach each other directly, you can setup a \u0026ldquo;repeater\u0026rdquo; in the middle, i.e. on a mountain. As long as both stations independently reach the repeater, indirect communication is then possible.\nI just received my HAM radio handle for use in Japan, and have now started registration for D-STAR.\nAPRS The TH-D75 speaks APRS (Automatic Packet Reporting System), a system for exchanging location information. As the TH-D75 has also GPS, you can use APRS to constantly send out your location. Moving with a bike, someone can remotely track your location on a map. You can select a frequency for D-STAR communication, and then via APRS send out your current location + details of the frequency where you are reachable via D-STAR.\nRepeaters can also be used with APRS, so you can not only exchange APRS data with stations within direct reachability, but extend that with repeaters. The transmissions are not limited to one single repeater, forming a mesh network. Also the TH-D75 itself can be used as repeater.\nAPRS is widely used: you can setup a device to just receive, and then catch APRS broadcasts from nearby airplanes, informing about their position and flight number.\nNetwork/Packet radio The TH-D75 can also be connected to a computer via USB-C, and be used for network communication. The KISS protocol gets used to communicate with the TH-D75, typical speeds are 1200 or 9600bps (baud per second). 20 years ago, when I used \u0026lsquo;packet radio\u0026rsquo; in Germany on the license free citizens band, 1200 or 2400bps were used.\nNot yet sure if I have nearby stations - the speed would be good to run BBS :)\nFurther.. The TH-D75 also has Bluetooth, supporting HSP profile (i.e. to connect headsets) or SPP (Serial Port Profile) to exchange data with PCs.\nThe SD-card slot is used\nfor audio recording logs of GPS coordinates lists of repeaters, and more. Direct output of received signals via USB-C is also possible, for processing/analyzing them with software.\nQuestions? Comments? -\u0026gt; Mastodon thread\nRelated:\nBlinry\u0026rsquo;s Fifty Things you can do with a Software Defined Radio Details on HAM radio training on Japan ","date":"2024-06-12T07:23:55+09:00","image":"/posts/2024/06/what-can-you-do-with-a-ham-radio-transceiver/20240524_115437490.kenwood_night2.jpg","permalink":"/posts/2024/06/what-can-you-do-with-a-ham-radio-transceiver/","title":"What can you do with a modern HAM radio Transceiver?"},{"content":"Let me capture my favourite hiking spots around Tokyo. This can give others hints, and I have a convenient page to refer to when choosing a hiking tour together. When you are in the area and in the mood, let me know! I\u0026rsquo;m out in the green almost every weekend.\nAs for planning: I have more details for the trips on the wiki. You can simply cut\u0026rsquo;n\u0026rsquo;paste the start station into google maps and have it plain the required train/bus trip. With a Suica or Pasmo card, you can pay at the trains, buses, and also the convenience stores to buy food or snacks.\nThe generic hiking hints are on the wiki. Some of the trips are also on Yamaquest, which has then also hints on whether you should take for example hand gloves with you in the month you go on the trip, and more.\nThe Mount Jinbasan/Takaosan trip This is my favourite one, report of what a typical trip looks like is here. Best points of this trip:\nNicely reachable: Fujino/藤野 station, the typical start point, is reachable from central Tokyo in under 90min. The last station is Takaosanguchi/高尾山口, with trains departing frequently towards Tokyo center. No bus ride, drop out of Fujino station, and start to walk. The trip spans 4 mountain peaks after each other, and you are not returning the way you came from - I love that The last peak is Takaosan/高尾山, which is the \u0026ldquo;default mountain\u0026rdquo; in Tokyos vicinity. Often quite busy, with families around. You can also do the trip in revered order: starting from Takaosanguchi/高尾山口. With that, you can even use chairlift or ropeway at the start - you do not need good condition to reach this first peak. If you have power left, you can go for the second peak, it\u0026rsquo;s 45-60min away. The trip from Fujino station to Takaosanguchi is 19km, takes 5-7 hours. Another feature: Takaosanguchi station, at the end of the trip, has also a nice bath. The last part between Takaosan and Takaosanguchi station has also a shrine and a temple.\nAll of the peaks also have at least one restaurant, but I mostly just take onigiri and sweets with me for the trip.\nDetails of the trip on the wiki Pictures of my first walk on the route Pics from all over the year, Takaosan 鍋割山/Nabewariyama This is my second recommendation: the Nabewariyama peak, in the Tanzawa area, in the southwest of Tokyo. Enter \u0026lsquo;大倉（バス）\u0026rsquo; as destination in google maps, have it plan the trip until there: it will include a train trip until Shibusawa-station/渋沢駅, and then 20min by bus.\nYou then walk 2.5-3.5 until the peak. One of the specialities: at the foot of the mountain, you can volunteer to take water bottles with you. At the summit, Yamaudon is sold, quite tasty. The water is used for this. The Udon is just sold until ~13:00, so plan accordingly. For this trip, you walk back the same way as you were climbing up.\nDetails on the wiki Nabewariyama in autumn The team preparing the Yamaudon has also a website and is on twitter.\n御嶽山/(Mnt. Mitake)/大岳山(Odake) trip Another nice trip, and this time without repitition. Mitake station/御嶽駅 is a bit far from central Tokyo. For the first part, a cable car can be used.\nYou can walk to the Musashi-Mitake Shrine (武蔵御嶽神社), then you can go to 大岳山(Odakesan, 1266.5m), then via 鋸山(Nokogirisan) you can go to 奥多摩駅(Okutama station)\nI do not like the Onsen near Okutama station, but when going back by train, you pass through Kabe-satation, which has a nice bath.\nPictures from that trip, and details on the wiki. The trip is also on yamaquest.\n塔ノ岳/とうのだけ/Tounodake trip Big recommendation also for this one: multiple peaks, great views. Have a trip planned to \u0026ldquo;ヤビツ峠\u0026rdquo;, the starting point. The last part will involve a 45min ride by bus. One speciality for this one: there is a rather steep part, where you have to grab a metal chains to safely go down. Totally doable, but you should be aware of this. At the end of the trip, you are at \u0026ldquo;大倉 (バス)\u0026rdquo;, so a different place than from where you started. That\u0026rsquo;s the direction where the steep place with the chains is easy to use, as you are just moving down using the chains.\nOne time I went for this trip in winter. The bus takes normally 45min and did not even drive until the \u0026ldquo;ヤビツ峠\u0026rdquo; that day. I met ice and spikes for the shoes would really have been right that day. The pictures with the snow worked out great :)\nDetails on the wiki Pictures from the trip, on a hot and humid day 大山/Ooyama trip This is a really nice one, in the Tanzawa area, south/west of Tokyo. Starting from Isehara-station(伊勢原), you can either walk or take a cablecar. There are 2 routes to choose from, labeled \u0026ldquo;men way\u0026rdquo; and \u0026ldquo;women way\u0026rdquo;.\nDetails on the wiki\nHonnitasan/本仁田山 trip A nice trip, 4 hours, in the Okutama area. I\u0026rsquo;ve done it just a single time.\nDetails on the wiki.\nTakigosan/滝子山 trip 5-7h trip, where you are returning the same way you climbed up. Just did this one a single time. Details.\nSekirouzan/石老山 This is a quite quick one, just 3 hours. You start at the Sagamiko lake. An onsen is also nearby. I did this trip just a single time. Details on the wiki\nFujisan/富士山 Not really Tokyo area :)\nDetails are here (Ascent) and here.\nQuestions? Comments? -\u0026gt; Mastodon thread.\n","date":"2024-05-28T19:49:36+09:00","image":"/posts/2024/05/favourite-tokyo-area-hiking-places/20230107_022545726_tounodake_small.jpg","permalink":"/posts/2024/05/favourite-tokyo-area-hiking-places/","title":"My Favourite Hiking areas around Tokyo"},{"content":"When I lived in Munich, the Alps were near - but I never got into the habit of hiking. Munich had plenty of green spots already in the city, I had a woods directly behind the house, great for jogging.\nTokyo is very different. Full trains, skyscrapers and lots of concrete push people into hobbies like camping, or hiking.\nTokyo has a great public transport, it\u0026rsquo;s easy to get to nearby destinations.\nSometimes I just go 70min by train, and can then directly start walking. For the farther locations I need 2.5 hours of train, plus 30-40min by bus.\nThe Mount Jinbasan/Takaosan trip Let\u0026rsquo;s look at a trip to my favourite location: at one year I went there more than 20 times. On a Saturday I\u0026rsquo;m leaving the flat at 6:50 by bike. After 10min I reach train station Chitose-Karasuyama. Storing the bike in a parking house, in the evening I pay 150yen for the day. At a nearby convenience store I fetch 3 Onigiri (rice balls) and some sweets. At 7:22 I enter the train, often there are other hikers, pupils going to lessons. You also see cyclists, with the bike stowed away into a big bag.\nAfter one change of trains I arrive at Fujino-station, it\u0026rsquo;s 8:33. I\u0026rsquo;m starting to walk, heading north. When visiting new hiking trails, I usually use the OSMand app on Android. OSMand using OpenStreetmap data, is more detailed than google maps, and can be used offline. After 25min I arrive at the foot of the mountain (\u0026ldquo;陣馬登山口\u0026rdquo;), a bus could also be used until that place. From there, 2 ways are running up to Mount Jinba. Farmers are putting out products: mandarine, tomato, sour plum, kiwi fruit, green tea. The price is written on small price tags, most portions cost 100yen. For payment, you insert the money in a small box. After 20min, on a day with clear view, we can see Mnt. Fuji nicely for the first time. We enter a forest.\nGreeting other hikers On most days, a few others are also walking into our direction, or walk the same way from the opposite direction. People greet each other: while still early with Ohayougozaimas/おはようございます. Later with Konnichiha/こんにちは. Not greeting would feel awkward, especially if the other party is greeting first - you have to return this. People are in general in a good mood, very different from commuters you meet on a weekday in the city. Here in the mountains, you also should be grateful for others around: if there is an accident, people will have to mutually provide help. I have also seen that: an insured person could not go down any more by herself, a team of \u0026lsquo;mountain helpers\u0026rsquo; brought a hull (shaped a bit like a coffin) up the mountain, stored the person inside, and used ropes to get the person down. These activities cost quite some manpower and put even more people at risk - for that reason, in the Corona time, people were also asked to not go hiking.\nArriving at Mount Jinba Around 10:15, we arrive at Mount Jinba. Speciality: a big white horse. Usually also other hikers are there, taking a break, some have brought cookers and prepare a meal. 2 restaurants offer cold and hot meals, drinks. Also more obscure things like Yuzuwine. Yuzu is a kind of lemon. Somewhere a Diesel power generater is running.\nWe leave the summit and walk towards 景信山/Kagenobusan, the next peak. This trail is used more heavily than the previous one. I think that trail is on a mountain ridge, but it\u0026rsquo;s not like one would look downwards into valleys on both sides all the time. After one hour we arrive at the Kagenobu peak, also here one or two shops are open. Tempura with vegetables, ramen, drinks, and souvenirs are offered. It\u0026rsquo;s usually as busy as the peak of Mnt. Jinba. Sometimes there are events, like preparation of Mochi (sweet rice cake), which involves hammering down rice over a longer time.\nWe leave towards the next peak, takes 60min until 小仏城山/Kobotokeshirosama. From here then 45min until 高尾山/ Mnt Takao. That one is well known: it\u0026rsquo;s the default mountain to go to from Tokyo. That means that our last part of the way until Takaosanguchi trainstation is quite busy. Also families with baby strollers, and bigger groups are around. We can choose between 6 different trails to get to the train station. The most famous one is also featuring one Shrine and one Temple. I usually stop at any of these, throw money in a box, bow. Sometimes monks are reading sutra. Souvenirs can be bought, and omamoli/talisman for protection. Some visitors hand over a small notebook to temple employees, to get a stamp from the temple, together with a personalized Buddhist phrase. Like a stamp rally, some people enjoy going to various temples and collect these stamps/signatures. On this last part, one can also use chair lifts or ropeway instead of walking.\nAs we arrive at the train station, we have further options:\nhaving a full meal in restaurants there going to one of the 2 or 3 museums. Most famously, the \u0026rsquo;trick art museum\u0026rsquo; visiting the bath right next to the train station. Nothing required, they rent out towels. The train to Chitose-Karasuyama station takes 45-60min, I fetch the bike and drive back home. The trip from Fujino-station to Takao-station is 19km, and can be done in 5 hours. Walking the route in the opposite direction is also nice: with the ropeway/chairlift on the first part, basically everybody can enjoy at least to go to Takaosan. If the condition allows it, one can then extend the trip to the next peak, or walk until Jinbasan. Just then when going to Fujino-station, trains back to Tokyo are more rarely running than from Takaoguchi-station.\nWhich kinds of folks are around? You meet many people. Young and older people just walking by them self, or in small groups. Also whole families. Mountain runners are not only walking but running the whole way. Especially between on the last part, there are sometimes longer groups composed of many women/man couples who walk next to each other - these are often organized by businesses, and aimed at finding friends or a partner for life. After some minutes, people rotate, so everybody gets to speak with everybody.\nI also have met Germans on these trails. Mostly I\u0026rsquo;m there alone, but sometimes with friends or colleagues. Some years ago, I walked for 90min with 2 Americans who worked at a nearby American base. America had just moved out of Afghanistan at that time, that also made an interesting topic to talk about.\nWhy I enjoy Hiking in Japan First factor: contact with nature. The Japanese term 森林浴/Shinlinyoku describes the act of moving in the woods, enjoying nature, \u0026ldquo;diving into the woods\u0026rdquo;.\nIt\u0026rsquo;s also not getting boring for me: when talking with a friend or coworker, you get the chance for longer chats. When alone, I\u0026rsquo;m listening to podcasts with one ear. Moving for some hours is also good for health.\nViews from the mountains are also nice.\nSpecial experiences? The more you go, the higher are also chances for unusual experiences. When going in the winter, I was one time slipping and falling down, damaging a limb, I felt that for some weeks. Professionals use in these situations \u0026ldquo;aisen\u0026rdquo;, metal frames with spikes which can be attached to the shoes.\nAnother day in late spring, on one part of the trail, many caterpillars were winding down from the trees above. They used thin threads like also spiders produce them. They were small enough to not immediately be seen. The caterpillars were spinning down over ~100 meter along the way, this was quite scary. The path was at that area made from wood, and walking around the event was not possible.\nI also had periods of rain, and after days of rain much mud on the trails. Once snow was falling, this was very nice - especially as it was also falling when afterwards going into the bath, and soaking into the hot outside pools.\nMore:\nPictures from my hike on the Jinbasan/Takaosan route Details on the route on the wiki Comments? -\u0026gt; Mastodon thread.\n","date":"2024-05-16T21:16:58+09:00","image":"/posts/2024/05/hiking-trip-in-tokyo-area/fujisan.jpg","permalink":"/posts/2024/05/hiking-trip-in-tokyo-area/","title":"A typical Hiking trip in the Tokyo area"},{"content":"We get workplace computers refreshed in 3 year cycles. My old system was a Thinkpad T590 with 32GB memory, and 8th gen Intel CPU. When it was up for refresh, I did choose a Macbook, and got a \u0026ldquo;Apple Macbook M2\u0026rdquo;, 2023 model, with 16GB RAM.\nI work as Linux engineer, doing support/development/communication. My goal is to use Linux directly on the AppleSilicon hardware. Now after some months, it\u0026rsquo;s time to recap what I settled with, and how well I can do my work with the system.\nTL\u0026rsquo;DNR All I need for my work is available, the Macbook with Asahi is now my daily work driver. In parallel, I could keep my previous system, a Thinkpad T590, so I keep now the Macbook at work and use the Thinkpad when working from home. Most data is anyway shared in company git repos, and the mail folders.\nDisk encryption Most important, this is 2024 and a laptop storage needs to be encrypted. The Asahi installer does not yet support encryption natively.\nNote: I\u0026rsquo;ve captured a howto for encypting an Asahi installation as of May 2026 in this article.\nThe Asahi Fedora remix install procedure guides to shrink down the MacOS area on the storage, and install Linux in a new area. While the whole setup is a bit complicated, Apple seems to explicitly allow non-MacOS systems on the hardware, while still allowing features like a chain of cryptographically verifying the booted MacOSX kernel, so what SecureBoot does on x86.\nThe MacOS installation is also the way to receive firmware updates: MacOS gets patched, and if required Linux can get updated firmware which needs to be loaded at runtime from the MacOS installation. Of course, while we have now fwupd on x86 for easy firmware updates, that can not be used on AppleSilicon.\nI followed the normal procedure to install Asahi Fedora remix, then creates a USB-live-media with Asahi, one boots from that and encrypts the disk. Details are on the wiki.\nThe Asahi Fedora remix distro consists mostly of Fedora aarch64 packages, plus some extra packages from Asahi, like the kernel.\nWhich of the Mac hardware features are supported? This is a recap of the features I need for my work, and how well they are supported. The Asahi feature page lists the detailed support states, I have a feed in my rss-feed reader for direct notification on any changes.\nImportant features, along with availability status:\nBuiltin webcam works without issues Sound: Bluetooth, also buildin speakers, and 3.5mm headphone jack all work. I mostly use my bluetooth connected Sony headset. Input engine Fcitx5 works, I use it for Japanese input The Macbook keyboard and touchpad are usable, but I\u0026rsquo;m not happy with the keyboard hardware wise. At home I use the System 76 launch keyboard, but it\u0026rsquo;s to pricey to just buy a second one for use at work. external monitor: HDMI output was implemented December 2023, for now without audio over HDMI. Before that, I investigated some workarounds via ethernet which use ffmpeg for encoding/decoding and then shows the\npicture on a monitor connected to a different Linux system, i.e. a Thinkpad or Raspi. External output via USB-C does not yet work. Controlling brightness of the buildin screen works via utility \u0026rsquo;light\u0026rsquo; (it\u0026rsquo;s in a package with the same name). Builtin microphone does not yet work, but I use the Sony bluetooth headset anyway For network, wireless works. I also use GBit ethernet via a docking station which is connected to the Macbook via USB-C How about software? The Asahi project traditionally used/recommended KDE Plasma, but basically all Wayland based windowmanagers should be fine. Xorg might work, but the Asahi team focuses on Wayland.\nI\u0026rsquo;m using the Sway window manager, works nicely on the Mac. Details are on the wiki.\nCompiling things like Performance Co-Pilot is blazingly fast, thanks to the 10 cores of the system - real cores, no hyperthreading.\nFor video chat: Google meet works nicely. Traditionally I used Firefox mostly, and in some cases like for Google meet I used Goodle chrome. Google does not provide a build of chrome for Linux on aarch64, I\u0026rsquo;m using chromium instead. Microsoft Teams also works, and Slack - using Firefox for that. Slack huddles do not simply work with Firefox, I did not research that deeper, not sure if they offer a Slack package for aarch64 Linux.\nSkype is almost unusable for me, I tried the implementation via https://web.skype.com . It complains that it\u0026rsquo;s not running on browser \u0026ldquo;Microsoft edge\u0026rdquo;. When started on Firefox, it Skype has much lag - sure if due to bad codec choice, or wlan throughput issues - but I did not notice wlan issues otherwise.\nIn Firefox, on proper keyboards, I use crtl+pageup and ctrl+pagedown to switch through browser tabs. These keys do not exist - on the Macbook. Replacement is \u0026lt;ctrl\u0026gt;+\u0026lt;tab\u0026gt; and \u0026lt;ctrl\u0026gt;+\u0026lt;tab\u0026gt;+\u0026lt;shift\u0026gt;, that\u0026rsquo;s a bit inconvenient.\nFurther software, which runs just smoothly:\nLibreoffice Inkscape Krita, using this for i.e. cropping and simple modification tasks. In the past, Gimp did lockup on my Wayland based systems. Latex/Xelatex Imagemagick, evince for pdf viewing, Python scripts for interaction with Jira cups for printing, PCP/Grafana for performance monitoring, including pmda-denki First things after installation Most hints for \u0026ldquo;first things after installation\u0026rdquo; are identical to Sway installations on other hardware like Thinkpads, they are on the wiki.\nOn the Mac, I have setup keyboard hotkeys for:\nturning the buildin screen on/off for tuning screen brightness up/down for locking the screen, I use \u0026lsquo;xlock -m blank\u0026rsquo; in a terminal, haven\u0026rsquo;t yet configured that as hotkey Then you should activate TRIM:\nvi /etc/crypttab # add \u0026#39;discard\u0026#39;, like for example: cat /etc/crypttab fedora-root UUID=8faa8bde-522d-4c59-9827-958afa662374 none discard # then rebuild initrd (maybe not required) dracut -f # then reboot reboot With that, you can execute \u0026lsquo;fstrim /\u0026rsquo;.\nVirtualization KVM virtualization for aarch64 Linux: I use this to replicate customer issues. Older kernels, i.e. RHEL7 or RHEL8, can not be virtualized. Qemu emulation: I use this when virtualization is not possible. x86 can be emulated, performance is ok! RHEL installation via kickstart into a virtualized guest takes 3 minutes, emulated x86 guest takes 21 minutes. For all my purposes it\u0026rsquo;s fast enough - just setup of IdM (FreeIPA) would be noticably slow. One issue I noticed: I got a Macbook M2 with 16GB of RAM. My window manager \u0026lsquo;Sway\u0026rsquo; has a low memory footprint, but Firefox uses quite some memory. When then running a KVM guest with 4GB of RAM, I\u0026rsquo;m running into 14 or 15GB used memory. Also running Firefox and chromium at the same time uses quite some memory. For running many KVM guests in parallel, I have now to go back to the T590 with 32GB of RAM, or to spare systems on the intranet via Beaker reservation.\nThe wiki has more details on performance comparison, and how I install Fedora and RHEL as KVM (virtualization) or Qemu (emulation) guests.\nLinux@apple silicon, my conclusion for now All the important pieces at least for my work are together for use at work: encrypted disk, openvpn access, wlan/ethernet, access to mail, video conferences.\nThe Macbook hardware is quite nice, the screen has nice colours, compared to my Thinkpad T590. I\u0026rsquo;m not a fan of the keyboard and the touchpad, might get an external keyboard in the future.\nPerformance Co-Pilot compiles in under 6min. Runtime on battery is very nice, and the Macbook/Asahi does exceptionally well regarding power efficiency in the comparison I\u0026rsquo;m doing right now, article with the results is under review for getting published.\n","date":"2024-03-24T21:51:08+09:00","image":"/posts/2024/03/asahi-fedora-remix-as-worksystem/20240325_000609981_macbook_asahi_onigiri.jpg","permalink":"/posts/2024/03/asahi-fedora-remix-as-worksystem/","title":"Asahi Fedora Remix as daily work driver"},{"content":"HAM radio? HAM radio or \u0026lsquo;Amateur radio\u0026rsquo; is the use of radio frequencies for noncomercial private purposes. After the German unification, I had been doing CB-radio in Germany. CB is \u0026lsquo;Citizens Band\u0026rsquo;, these are a few frequency channels which can be used without any license. I had gone as far as setup a metal antenna on the roof of the house, had spent nights with friends on a mountain, trying to do radio communication with people in other countries.\nCB is restricted for example with the available frequencies, and the power of the radio waves. HAM radio amateurs have to prove knowledge in passing an exam, and can then use more frequencies. If everybody was allowed to randomly use radio technology, then all kinds of services would probably be impacted: communication of police, ambulance, planes and so on. Multiple HAM amateur levels exist.\nThis technology was always interesting for me. One could learn for the exam with books, or trainings. My home town in East Germany was to small for trainings, and later when I lived in Munich I was focusing on connecting with computer nerds, and later Japanese learners.\nNow, living in Japan, I registered for a weekend full of training - that was last weekend. I registered online for training+exam, payed 2.5万円(~150eur).\nStart on Saturday The training was in Sugamo near Ikebukuro/Tokyo - in cycling distance for me, as I live in the West of Tokyo. Weather was good, after 75min of cycling I arrived in time for start at 9am. We were ~35 trainees, including a maybe 10 years old girl and her mother. Maybe the father running a HAM radio station and they wanted to join him? We received material for training. Saturday was dedicated to 法規/Hoiki, so laws and legislations. A single instructor was guiding through the full day, all hours. We got a text book which had all of the contents for both days. Following the explanations was not to hard - but then we got example questions, in the style of what was to be expected in the exam. Reading and understanding Japanese texts.. this is hard for me. As you see on the first picture, many 60min lessons with 10min breaks. Lunch break was 50min, I went to a nearby place, could have been better.\nWe fininished aroung 17:15, when I arrived at home it was already dark. There was homework though: we got sheets of paper with tech knowledge like Ohm\u0026rsquo;s law. \u0026ldquo;We are really running through it on Sunday, read this in the evening so you remember next day\u0026hellip;\u0026rdquo;\nSunday: Radio Tech Rain was forecast, so I arrived by public transport. Everybody knows the Nato alphabet, \u0026ldquo;A as Alpha, B like Bravo\u0026rdquo; and so on, that was also taught. Well, you can\u0026rsquo;t really expect teaching, but hinting on things the trainees had ideally studies themself before. I had prior knowledge from by CB times - but had to fight with the Japanese terms for things I knew in German and English. I work in IT, most terms are taken from English, transformed into a certain way to fit the Japanese tongue, and written in the Katakana alphabet. That makes it easy to connect the English and Japanese spelling - but the radio and electronics are older and got native Japanese terms. You could say \u0026lsquo;frequency\u0026rsquo;, but 周波数 (shouhasu) is used in Japanese. So the teacher started to run through the tech content, referring to vocab I was hearing for the first time.. to make it in time, the speed had to be high - also questions were not possible.\nFun fact, the Nato alphabet has in Japan then a brother, the Katakana alphabet: \u0026ldquo;ア like 朝のア\u0026rdquo; and so on.\nThe famous Yagi antenna was invented by the Japanese Mr. Yagi/八木. We got sheets with knowledge to prepare for the exam. At the previous evening I had looked up a nearby Curry place: Nepal style. That was delicious!\nAfter the training we got time for own training, and also could ask the teacher.\nExam Eventually we were sitting down in the exam. All others left after 10min, I took 30minutes. Today I looked at this link and as I had station 30, it seems like I passed.\nSo I can now consider to get a ham radio, obtain a license to operate the radio, and can then use that. Along with that, a callsign will be handed out, worldwide unique. I can also try to have the license rewritten to a German license of the same level. Also funny: when you register a radio station in Japan, you get to pay a fee of 300円 per year. You get a letter to pay that.. maybe the letter eats already most of the fee. I will bring this up from now on when I talk about dogs being taxed in Germany.\nI learned that actually one can also sit down in English HAM radio exams in Kanto area, and if obtained these can then be rewritten to a Japanese license.\nComparing German and Japanese HAM licenses Quite some differences! In Germany we have just 3 levels. \u0026ldquo;N\u0026rdquo; is quite new, the lowest level. Even there, you can just use 10m/2m/70cm with 6.1W/10W EIRP. frequency areas for use, and can send with up to 100W PEP. In Japan, we have 4 classes. The 4th/lowest one can send on almost all frequencies, like the highest one. Depending on frequency, 10 or 20W. Graph of which class in Japan can use which frequencies.\nAlso different: in Germany you can apply to get a callsign assigned when you have passed the exam. In Japan, after passing the exam, you just get a certification paper (as plastic card, like a drivers license). With that, you can apply to \u0026ldquo;operate a radio station\u0026rdquo; (無線局開局), and a callsign is assigned to that station. You have to pay 300円 per year for \u0026ldquo;using the ether\u0026rdquo;. The permission to operate the station is valid for 5 years, can be renewed.\nNext Steps One week after the exam I was able to see online that I passed, 2 weeks later I got an email - along with an invitation to Hamutte, an online forum for Japanese HAM licensed people. So that\u0026rsquo;s a place like a builetin board, to ask questions and so on.\nI\u0026rsquo;m now waiting for my license card, and can then apply to run a radio station. Will probably order a handheld 70cm/2m device, and a SoftwareDefinedRadio device, details/data collection.\n","date":"2024-03-01T12:37:19+01:00","image":"/posts/2024/03/japan-ham-radio-training/00_Jard_papersb.jpg","permalink":"/posts/2024/03/japan-ham-radio-training/","title":"Japan Ham Radio Training"},{"content":"When I came to Japan in 2016, I had my German drivers license translated by the Japanese Automobile Federation. For ~4000 yen I got translations. Like some other expats from Germany, I got the idea that I would then have to run through normal Japanese drivers school and pass the exam.\nPublic transport in Tokyo is great, so I kept it at this.\nNow in my 8th year in Japan, I learned that drivers licenses from certain countries can be rewritten into Japanese licenses - Germany is one of the accepted countries.\nOn a Friday in February 2024, I had a day off and looked into this. Looking up 運転免許証 (\u0026lsquo;drivers license\u0026rsquo;) showed that a police station in 15min walking distance handles these. They opened 8:30 in the morning, so I walked over, and learned that this station just deals with license renewals. First difference: drivers license in Japan need to be renewed.\nI had seen a bigger office related to drivers licenses, and got there by bike in 25min. This was the districts main office, but I learned there that just 3 offices in the Tokyo area handle rewrite operations. The next was was the Samezu Driver\u0026rsquo;s License Center (鮫洲運転免許試験場) in Shinagawa, I cycled there in 50min.\nTurns out that plum and other trees start blooming. :)\nEntered the huge building, they even have a dedicated room for bicycle parking. The procedures started. I handed over:\nface image (I had it taken at a foto booth) my German license - I still have a paper one which I got in 1998. The pins holding my face image are rusty.. travel passport translation of my passport into Japanese, I had that since 2016, it was done by the Japanese Automobile Federation proof that I had lived for some time in Germany after obtaining the German drivers license. I used papers, written in English, showing that I was employed in Germany. This was apparently requested as a sign that I had also driven cars in Germany. proof of residency (住民票). This turned out to be a fun one.. After handing over the papers, I waited 30 or 45min, and got further instructions.\nProof of residency/住民票 The \u0026lsquo;proof of residency\u0026rsquo; is in most cases issued by the local town hall. One hands over the residence-card (在留カード), fills out a formular, waits 15min, pays 300yen, and receives the \u0026lsquo;proof of residency\u0026rsquo;. The paper can have optional features: depending on what you want to use the paper for, you might need to get references to your citizenship (\u0026lsquo;German\u0026rsquo; for me), your assigned number (マイナンバー, it\u0026rsquo;s for example used to identify you when paying taxes) and some other \u0026lsquo;features\u0026rsquo;. Of course, you want to hand out the \u0026lsquo;proof of residency\u0026rsquo; with a minimal set of details, just the details the receiver needs to know. On the other hand, you also need to know that set. When I was at the town hall, I said \u0026ldquo;I need this for the drivers license\u0026rdquo;, hoping for suggestions for the right set of features.\nNow, I got the wrong set.. the paper needed in my case a note about my citizenship, which it did not have. I pictured myself already cycling 60min to my local town hall. Now, we have some level of digitilization in Japan: I have a \u0026lsquo;mynumber card\u0026rsquo;, a personalized card, which I can use for example to get the \u0026lsquo;proof of residency\u0026rsquo; printed at convenience stores.\nOn Thursday, I went to the town hall to get the \u0026lsquo;proof of residency\u0026rsquo;. I went in person to get their advice regarding the options I needed on the paper. Going to the drivers license centers on Friday, I did not have the \u0026lsquo;mynumber card\u0026rsquo; with me, so printing at a store was not an option. Luckily, there was a town hall nearby - so it\u0026rsquo;s possible to use ones outside of the district where I live.\nFurther procedures 10min by bike to the Shinagawa Yashino district center (品川区八潮地域センター), and I got the paper printed there. After a lunch at a nearby Indian restaurant I got back, completed papers. I had to pay ~5000yen, pass an eyetest, and had a picture taken. Then again waiting until the card was printed, receiving the card (\u0026lsquo;Contratulations! Please check the details on the card!). It\u0026rsquo;s an RFID card, you got to choose 2 pins and after entering them on readers which are provided, you can verify your details.\nGreat idea: on the backside you can directly choose options to donate organs: to not donate, to donate after the brain stops working, and more.\nAs this is Japan: when leaving the license center, one also has to apprechiate the small shop which has fan goodies and figures of the official Japan police character. 🥰\n","date":"2024-02-04T11:27:05+01:00","image":"/posts/2024/02/drivers-license-in-japan/20240201_035322338_residence_confirmation_waiting2_1280.jpg","permalink":"/posts/2024/02/drivers-license-in-japan/","title":"Drivers License adventures in Japan"},{"content":"How do Germany and Japan think about World War 2 (WW2)? Why are some of my friends not visiting the Yasukuni shrine in Tokyo? This is my attempt to sort out the information I received about the 2nd World War, and observations how people think about that time period today - in Germany and Japan.\nEast German views on WW2 As most probably know, Germany had been separated after the war. I grew up in East Germany, under guidance of the Soviet Union it was a Socialist republic. From the East German school system I received my first education about World War 2 (WW2): the Russians had fought down the German Nazi regime (along with also America and other Western countries). This allowed then the creation of the Socialist German Republic.\nThis picture got extended by my parents: my hometown Mühlhausen is located in the middle of nowadays Germany, with East and West combined again. Towards the end of WW2, the allied forces were nearing from the west, and Soviet forces from the east. Initially, the Soviets took Berlin, and the allies took my hometown. The citizens were happy: not only about the end of the war, but also that the allied marched in and not the Soviets. The concerns regarding the Soviets were based on stories of what happened so far in occupied areas, along with many years of propaganda which showed Communism as the evil itself.\nThe allied were not happy about the Soviets getting whole Berlin, further agreements were signed which led to the allies pulling out of big areas, including my hometown - which got taken by the Soviets. In exchange the allies got parts of Berlin, the well known separation of Berlin into sectors happened.\nWhile people were not happy with that initially, the socialist East Germany got founded, and people made the best out of it. The Soviet bloc by itself looked at Germany on one hand as a strategic buffer in case America/Europe would start an attack. On the other hand they were aware that it was also nicely located in Europe, likely to see frequent visits from West Germany and other countries: so quite some resources did go into restoration of East Germany, to show \u0026ldquo;what Socialism can do\u0026rdquo;. West Germany was rebuilt much faster with America and others helping via the Marshall plan, but also East Germany was doing better than many other Socialist countries. Details are in Memories of East Germany and the German Unification.\nWest German views on WW2 Around 1989, Germany got approval to reunite again. In many aspects this was more \u0026ldquo;East Germany joining West Germany\u0026rdquo; than equally combining both parts. Up to 1989, East German citizens were thinking of Soviets as brothers, were learning Russian as first foreign language in school, we had pen pals in countries of the Soviet bloc, and so on. With the German unification, the views of the West were completely taken over - an impressive experience for me. Nothing on the facts of the life in the past dozens of years in East Germany had changed, but over night the Soviets were no longer the brothers, but were said to be \u0026ldquo;these suspicious creatures over there, who probably were just waiting to invade us\u0026rdquo;.\nMy education changed, school topics were now dictated by the west. I was in the last school year learning Russian as first foreign language, but now English got first foreign language. Also views on WW2 changed: while acknowledging that Russia had also been in the war, it was mostly the allies who freed Germany from the Nazi regime. As I later realized, Germany\u0026rsquo;s views on WW2 are matching with Americas views, in the aspects I can think of.\nMy knowledge on Japans role in WW2 over time After the German unification I learned about Japans participation in WW2. Japans motivation to enter the war is not investigated in depth: pupils learn about Japans assault at Pearl Harbor, and entering the war with the goal of extending the Japanese imperium.\nAfter moving to Japan, I learned more about the Japanese views on WW2. Nowadays, Germany as well as Japan are forbidden to own nuclear weapons. Both heavily rely on the USA as an ally for protection, Japan even more than Germany - which additionally can rely on the European Union.\nStarting with the year 1633, Japan implemented a policy of isolation from the rest of the world. Under the isolation, trade and contacts with other countries were at a minimum. While the world unfolded, Japan further developed arts, tea ceremony, and so on.\nThis period ended around 1852 with the arrival of the \u0026ldquo;Black ships\u0026rdquo; (黒船来航), led by Commodore Perry from America. The rest of the world had advanced their military and Perry forced Japan now to open up their country. In towns like Hakodate one can see traces of the \u0026ldquo;back ships\u0026rdquo; event, at these occasions it\u0026rsquo;s shown as a very positive event. The opening did lead to the Meiji restoration, a period of industrialization and with many changes.\nJapan looked at Asia at that time: most other countries had been occupied by western countries and had become colonies. Simply doing nothing, Japan would also be likely to end up as a colony - so extending to other areas could also be seen as a way to prevent that. The bigger the own territory and own resources, the more likely it is that attempts to be colonized can be fought off.\nJapan did incorporate Hokkaido, the big island in the north of the main island. Okinawa also got included into Japan.\n1894/1895: the first war between China and Japan, resulting in Japan extending its territory over the Korea area.\nIn 1895, Japan invaded Taiwan and extended its territory.\nIn 1904/1905, Russia and Japan were at war, ending in a victory for Japan: Korea, Sachalin and some ports in China now became territory of Japan.\nInterestingly, Japan is keeping also this view on history active, they can be studied for example at the Yushukan museum in Tokyo.\nInteresting fact: also for Japan, dividing the country was discussed like for Germany: North to Russia and South under American influence. This was overthrown eventually. Okinawa, the big island in the south of Japan, was until 1972 under American administration.\nThe role of Germany and Japan in the world Germany is nowadays deeply embedded in the European Union. It\u0026rsquo;s one of the financial heavy weights of the EU, also having much voice due to this. Neighboring countries do not really carry grudge from the WW2 time - but the topic comes up from time to time. For example, when the Greek financial situation did not fullfill European expectations some years ago, Greek also asked Germany for financial reparations which allegedly were left from WW2.\nJapan is an island, and also politicly more isolated. It\u0026rsquo;s not in some kind of EU, and feeling in danger of attacks from North Korea and China, Japan relies heavily on America. In exchange, America uses various places in Japan as military bases. Japan also fears that it might be pulled into a conflict between mainland China and Taiwan.\nAfter invading Korea and China before WW2, both of these countries give Japan the feeling of \u0026ldquo;unfinished business\u0026rdquo;. From time to time, both countries ask for \u0026ldquo;official apologies\u0026rdquo;. It seems like Japan has offered these, but they were not accepted by China/Korea, maybe because the countries want to keep the current state of \u0026ldquo;unfinished business feeling\u0026rdquo;.\nOn top, there are disputes with Korea, China and Russia about islands. When the Ukraine invasion started, Japan apparently felt like Russia was busy right now and started to remind of the islands they feel are part of Japan.\nAn interesting fact is that Taiwan, one of the countries who got invaded by Japan, is favorably looking at this. With China, both countries also have a common opponent. China (Xi) has \u0026lsquo;promised\u0026rsquo; to invade Taiwan at some point, and Japan feels strongly that it would be pulled into this invasion.\nDealing with / thinking about the war Some war related aspects are especially interesting.\nAir Raids and anime In WW2, Japan had experienced many air raids. These have been reflected in anime movies, Joachim Alt has presented on these in slides and video (all in German). Joachim looked at various anime around the topic and sees a gap between history and memories:\nactual war: aimed at expansion of territory, using 3.5mio soldiers memories: more defensive, remembering the air raids on 60mio civilians in Japan most important memory about the war at the home front: the air raids The Diaries of Anne Frank These are very well known in Germany: they are often read in schools. An example for people standing up against the oppression of the Nazi regime. Germans read this and remember that previous Germans did this, and it must not happen again.\nIn Japan, the diaries of Anne Frank are also known, and also available as manga (comic book). The interesting part is that Japan manages to read the diaries, and see itself in the role of Anne Frank, oppressed by the opposing military forces in WW2. The western mind sees this and sees a strong contradiction here, as Japan was on the side of Germany (the oppressors of Anne Frank) in WW2.\nThe Yasukuni shrine The Yasukuni (literally \u0026ldquo;peaceful country\u0026rdquo;) shrine is located in Tokyo. It is dedicated to those who died in the service of Japan, including war criminals.\nThe Japanese prime minister is in official mission not visiting the shrine, if he does then it provokes voices from China and Korea. I have friends from China who oppose to visit the Yasukuni shrine. The Yashukan, with explanations on Japans war history, is located next to the shrine. After recent renovations, many explanations are also in English.\nForbidden insignia/books/symbols In Germany, the public display of various Nazi symbols is forbidden, most known symbol is the Hakenkreuz, a type of swastika symbol. In contrast, swastika are since ancient times used all over Asia, mostly rotated in the opposite direction from the Hakenkreuz. With the Nazis using the symbol, there have been activities to purge the symbol from some maps in Japan, to \u0026ldquo;not put visitors at discomfort\u0026rdquo;.\nIn Germany, it was until recently forbidden to own/buy/sell Hitlers \u0026ldquo;Mein Kampf\u0026rdquo;, to prevent the spreading of propaganda which is opposing the German constitution. Some sources seem to report it as allowed now.\nOn Japan, the \u0026ldquo;rising sun flag\u0026rdquo; (red dot in the middle, sun beams into all directions). The flag has been used since many hundred years and was not just introduced around WW2. Many Japanese right wing parties use the flag nowadays, it is not forbidden but seen critically.\nOppenheimer The 2023 movie \u0026ldquo;Oppenheimer\u0026rdquo; is about the development leading to the nuclear bomb. Japan was the only country to experience 2 nuclear bombs, so while the movie was released all over the world in 2023, it will run in Japan in 2024.\nEDIT: After the initial version, year of the start of isolation, and of the invastion of Taiwan were fixed.\n","date":"2024-01-13T01:26:27+01:00","image":"/posts/2024/01/ww2-history-irritations/20240113_023632087_yasukuni_shrine_1280.jpg","permalink":"/posts/2024/01/ww2-history-irritations/","title":"World War 2 History Views"},{"content":"Sorting my pictures from this year, I got reminded of the various pieces of tech I got into contact with..\nThe tech year started with writing an article in how Kanji are input with normal keyboards. The area between Japan, Kanji and technology stays interesting for me.\nAsiaBSDCon In April, we had AsiaBSDcon in Tokyo. I\u0026rsquo;m much more active in the Linux world, but BSD covers other interesting areas, for example NetBSD with support for older hardware. The event was even within cycling distance. 2 days full of presentations and meeting some interesting folks. Took the opportunity to look at OpenBSD/NetBSD/FreeBSD for desktop usage, used OpenBSD while at the conference, shared the experience in a mini talk. Details are here.\nBike navigation with OpenSource In May, we had the \u0026lsquo;Golden week\u0026rsquo;, a series of national holidays in Japan. I found a chain of software which helped me navigate on a bike trip to a friends house in Tochigiken, where I helped with rice planting.\nbrouter-web is great for planning, allows download of the planned route, and software OSMand then allows importing of the route on Android and iphone mobiles. I used an Android and an iOS mobile sequentially for navigation, and recharged with a Thinkpad in the backpack. Software syncthing allowed to share pictures taken on the trip. Data for the trip and links, like details of the hotels etc. were stored in markup files, also synced with syncthing to the phones, and accessed there with software Obsidian. I captured more details in this article.\nRISC-V I got hands on a Star64 board with RISC-V cpu, looked at usability as Jellyfin- or webserver, compared to i.e. the Raspi. Quite interesting hardware, and the hardware features which are already supported are usable! Details are here.\n![FreeBSD booting](0/FreeBSD booting.jpg) ![Star64 RISC-V board](0/Star64 RISC-V board.jpg) ![Steamdeck running portal](0/Steamdeck running portal.jpg)\nRHCE cert In June, I managed to renew my Red Hat Certified Engineer (RHCE) cert. I\u0026rsquo;m taking the RHCE certifications since more than 20 years, in the last years much Ansible knowledge is required - I use it to install virtual guest machines to replicate customer issues, but apart from that I\u0026rsquo;m not using it. Took me 6 months of training along to the normal work to learn for the cert, and I also failed on the first attempt.\nRetro summer In July, I tried to revive some hardware from the attic. The Robotron EC1834, a 8086 clone from former East Germany, does not light up the monitor I got along with the system. Now in Christmas time I got further: the video signal from the system is CGA/EGA compatible, not good for VGA monitors. I got a MCE2HDMI device which should help to access the signal - all under the assumption that the signal from the system is ok and usable at all. I get something now, but no proper sync on the signal. Details are in this thread. I got to light up an Amiga 1000 for the first time - having no keyboard for the system though. Also had to use the Amiga 500 to create a floppy disk with the kickstart code. I would love to get hands on an A1200 or A600 at some point. Details on the Robotron and Amiga activity are here.\nFor colleagues I created a short presentation with the more or less obscure hardware I got my hands on so far, slides are here.\nAt the company, we had the \u0026lsquo;family day\u0026rsquo; event, and I tried to create a map of the office in the style of classic 2D role playing games like \u0026lsquo;Chrono Trigger\u0026rsquo;. I failed.. more details here.\nMastodon hosting and talk Mastodon is an OpenSource micro blogging software, a bit like Twitter. Many people around me have moved to Mastodon, I\u0026rsquo;m also using it heavily. In the last months I looked into hosting a Mastodon instance myself. I\u0026rsquo;m not a big fan of containers, but as software like Mastodon or Lychee (image gallery hosting) are composed of increadibly small software pieces which all need to fit together. Putting these into a container allows easier updating: if an upgrade of the components in the container fails, one can simply fall back to the older state. I got Mastodon to run in a container controlled with podman, details are here.\nRight now, I\u0026rsquo;m not planning to move my Mastodon account to the own instance. I registered a new domain for the own Mastodon instance, and use it for playing. @tracks is an account for posting game tracks, @languagelog is a bot posting updates from the great languagelog site. Writing this Mastodon bot was a nice experience! While Twitter is shutting down their APIs, all of this is easy with Mastodon.\nIn November, I gave a talk on Mastodon at the Tokyo Nerd Nite event: slides, recording and images are available.\n![Asahi Linux install](1/Asahi Linux install.jpg) ![Asahi Linux](1/Asahi Linux.jpg) ![Hard Off retro hardware](1/Hardoff retro hardware.jpg)\nAppleSilicon/Asahi Linux At the beginning of 2023, I was up for company hardware refresh and decided to go with an Apple Macook pro this time. Apple is shipping these with AppleSilicon CPUs, an own ARM design which has a good reputation technology wise. I have colleagues using such hardware for their work in support, but for myself I did only consider the hardware with Linux natively running, not Mac OSX. The decision was basically\nto go again for an Intel based laptop or an Apple Macbook, knowing that the hardware was not yet fully supported on Linux as of requesting the hardware refresh I asked for the Macbook, got it in summer/autumn, and now almost all features are supported. HDMI output was missing until November, I investigated a chain of\nrecording the screen contents, encode it with ffmpeg and have it streamed to a different system which would display it. Works quite nicely, just a tiny bit to slow to feel \u0026ldquo;fluent\u0026rdquo; for example when browsing web pages. But good enough for showing videos. The details are here, this might come in handy in other situations too. Direct HDMI output from the Macbook got implemented in November. Speaker output came in December, but was not essential for me as I use a Sony headset via bluetooth for conf calls, that worked fine already. Now just some pain points remaining:\nright now no support of monitor output via USB-C, but that\u0026rsquo;s just for convenience. I can use an HDMI cable to connect monitors. As this is aarch64 and not Intel, I can not easily spin up KVM guests to replicate customer issues. Spinning up aarch64 RHEL9 works, but older RHEL runs into issues. Need to investigate emulation for this. I\u0026rsquo;m not a fan of the keyboard and trackpad, being used to Thinkpad keyboards and the trackpoint since years. Considering to get this keyboard for work. I\u0026rsquo;m very happy with the System76 launch keyboard, it\u0026rsquo;s just much more expensive. Apart from that, the Macbook is quite nice! Amazing battery lifetime, nice screen. The Asahi project, which ports Linux to the AppleSilicon, also announced now Fedora as their new main distro. Details for running Linux on the Mac are here.\nSelf hosting in Japan? My services like mail- and webserver are running on a virtual server in Germany, latency of 230-250ms from Japan. In Japan, I have 1gbit fiber into my room, and could upgrade it to 10gbit. I seriously consider running services in the flat in Japan 24/7. Services reachable from the internet via IPv4 would be an issue: my ISP is not giving me a static IPv4 address, and also the router does not allow port forwarding. I would need to switch the ISP ontop of the fiber, as media and ISP are separate in Japan. For outgoing v4 connections, there is no issue. For IPv6: I get a whole segment routed and can directly offer services via IPv6. This made me setup an experiment: I created a DNS AAAA record and offered webservices via tcp/443 via v6. For the same domain I created a DNS A-record, which I pointed at my server in Germany, and via a VPN forwarded the request to Japan. So requests for the domain via v4/v6 landing on the same webserver in Japan, just with latency differences. Running for example a Mastodon server like this should be fine, also the latency is not important in that case. For now, not hosting in Japan.. but keeping the idea in mind.\nSteam Deck, OSSummit In December, we had the Open Source Summit Japan in Tokyo, first time since beginning of Corona. I\u0026rsquo;ve been at the event since many years, in the last years at the Red Hat booth. I have the habit of bringing in more rare hardware with Linux, in 2019 that was the Nintendo Switch gaming console running Fedora.\nIn summer, the Valve Steam Deck got available in Japan and I ordered one. It\u0026rsquo;s a Linux based gaming console with x86 hardware. Linux is running natively, and with some software layers also almost all Windows games can run - it\u0026rsquo;s a huge hit. I spent more time tuning Linux on the system than actual playing, and to the OSSummit I brought the Steam deck running Fedora, and powered the booth monitor. 3D acceleration works out of the box, Valve is a good citizen and brought patches into the upstream projects.\nTrying out the Steam deck is really fun: it comes natively with a Linux distro from Valve, the vendor. 2 years ago, I wrote code for the Performance Co-Pilot (PCP) project, this code allows measurement of electrical power consumption. Various distros include the code, so I could just log onto the Steam deck, install the PCP package, and then measure power consumption while playing a game. Nice! Curious about pmda-denki? This article has details.\nWith Fedora/Steam deck, I used Plasma/KDE again. Getting the consoles internal screen and the external one via HDMI (which would be the booth monitor) to work was doable, but then trying to setup remote control for that external monitor was hard - I did not get that up after some hours.\nI also brought the Macbook running Asahi Fedora remix to OSSummit. Was able to chat with many folks, in Japanese and English, and even German!\nAlso great to catch up with colleagues at the booth, and past colleagues and partners who were around. One specialist even asked about Red Hats strategies and code for measuring power consumption!\n![Nerd nite Mastodon](2/Nerd nite Mastodon.jpg) ![Robotron screen](2/Roboton screen.jpg) ![Streamdeck Fedora monitor](2/Steamdeck Fedora monitor.jpg) ![Delay measurements](2/Steamdeck remote delay measurements.jpg) ![Sun Ultra 5 neofetch](\u0026lsquo;2/Sun Ultra5 neofetch.jpg)\nGermany trip Back in Germany, I tried to setup Solaris on a Sun Ultra 5 system - to no avail. For network installation, one has to provide the Solaris install data via NFS version 1, I installed Fedora core 1 from 2003 in a KVM guest as NFS server, bridged that network to the network interface of the Sun.. and still it won\u0026rsquo;t boot Solaris. Ah well, now the system is happily running NetBSD 9.3. Great to see current operating systems still supporting this old hardware.\nI had planned to get LED stripes with individually addressable LEDs, had been researching into this a bit. Bought 2 stripes which turned out to not have individually addressable LEDs, I will take one of these with me to Japan to light up the room in the evening. Indivitually addressable requires also a Raspi/arduino/ESP32 or such and a more complex setup.\nOther project in the last 2 weeks: I got a cheap zigbee2usb device, and 5 zigbee temperature sensors. First testing went well: I connected the zigbee2usb device to the Thinkpad, and the sensors all reported the measured temperature. Then I moved the zigbee2usb device to a Raspi4, sensors still were fine. On that Raspi4, I also host a 2.4Ghz wlan access point, spanning up a network for some \u0026rsquo;less trustable\u0026rsquo; devices like smart power plugs which measure power consumption. As soon as the Raspi4 starts hosting that network, the zigbee sensors go silent. Turns out that the wlan and zigbee are in the same frequency area and interfere. Separating the zigbee2usb device from the Raspi with a USB cable helps with this - but still one of the smart plugs has now lost contact. Zigbee devices actually also can act as \u0026lsquo;router\u0026rsquo;, but only Zigbee devices like light bulbs or power plugs which have constant power supply. So I could consider to replace the wlan connected smart plugs with Zigbee ones, and do a mesh network. The temperature sensors are now getting recorded into PCP archive files, and visualized with Grafana.\nSoftware which stayed Most stuff stayed :)\nI\u0026rsquo;m still a happy user of Wayland with the Sway window manager, also on the Apple Macbook. Tiny Tiny RSS as feed reader is my most heavily used service. For podcasts, I use the FiiO M3K with rockbox.\nMore direct news via the Mastodon account.\n","date":"2023-12-24T18:57:47+01:00","image":"/posts/2023/12/2023-tech-games/SteamdeckRemoteDelay.jpg","permalink":"/posts/2023/12/2023-tech-games/","title":"2023 Tech adventures: Hard- and Software"},{"content":"Background Most flat rental agreements in Tokyo are limited to 2 years, and can then be renewed. Renewal is mostly uncomplicated, the tenant just signs again a paper. There is a catch though: in most cases, an estate broker was doing the initial flat search, and at contract renewal they receive again money from the renter: a monthly rent. I was living in this flat for almost 6 years, renewal came up - so thinking of this extra expense made me look for a new flat.\nMy first flat searches in Tokyo were directly via internet sites like Sumo. Eventually, I ended up with a company specialized in renting to foreigners. Their biggest advantage: no guarantor required. When renting, one has to put down a sum of usually 2 month\u0026rsquo;s rent, which would be used to restore the flat if you move out without properly canceling the contract. On top of that, most contracts also require a guarantor: someone who can jump in if that sum is not enough. For young Japanese moving to the big cities, parents sign up a guarantor. In some cases, a different company can take the role of the guarantor, charging the amount of an extra month\u0026rsquo;s rent.\nFor my second flat, I used a real estate broker, and then did 2 renewals, each one for a month\u0026rsquo;s rent. Various things made me look for something different: the family owning the flat did live in a house directly in front. Starting in spring, they took their house down and had it rebuilt - all along with noise which forced me to work from the office almost every day. The fact that we tenants paid for their house, and were not offered a temporary reduction of our rent, did not make the experience better..\nSo I was in the market again.\nMy Search Criteria I considered moving thousands of km away to Kyushu, but eventually settled with \u0026lsquo;Tokyo, again\u0026rsquo; for these reasons:\nAbility to commute to the office. Contact with colleagues is good, the food around Ebisu is great. Commuting by bike is also refreshing - so far my commute by either bike or train took 25 minutes. Infrastructure: so many nice places in Tokyo I like to revisit, also hard off stores, big electronic supermarkets, Akihabara. Meetups with Germany/Japan communities, Tokyo Linux user group, a few friends, and sometimes friends from Germany coming over Airport within range This time, I was working with the Urban Renaissance Agency (独立行政法人都市再生機構, aka \u0026ldquo;UR\u0026rdquo;). It\u0026rsquo;s a government-affiliated company with a good reputation, giving me the feeling that at least I\u0026rsquo;m not paying for someone\u0026rsquo;s new house, but paying for society. UR prices are usually cheaper than normal or normal, but they have some great features:\nno key money (礼金), literally \u0026ldquo;thank you money\u0026rdquo;. This goes to the landlord or the real estate broker, and the tenant does not get it back no renewal fee (更新料), to be paid at renwal, every 2 years no guarantor required. For my second flat, I asked a colleague to jump in, and not having to ask someone this time is a real relief. So I\u0026rsquo;m not owing someone to do this. Many Western companies like Google also take this for their employees, but not my employer. The actual search Besides the actual criteria, I was quite free about the location of the new flat so pulled up the map from UR, and looked at their free estate. Based on this I was looking at flats with 1 or 2 rooms, which ideally would allow me again to commute by bike to work. On Saturdays, I cycled to objects marked free. All of these were parts of bigger housing complexes (danchi/団地), and as the management office was open I asked for a key and could have a look. I looked at 5 flats. 3 were really nice, but when I contacted the office to grab the flat it was already taken - so these flats are really popular.\nI ended up with this:\nA 2-room flat, 46qm. One room Western style with wooden flooring, and one room Japanese style with tatami mats (和室) 7th floor in a 13-floor building Commuting to the company by bike in ~45min (instead of the 25min from old flat) The paperwork Until the finished contract, some office visits were required. First visit to a small office, to get the flat reserved. I had chosen that one as it was next to me, but for the contract work I needed to go to the bigger Shibuya office.\nWith an appointment I went to the Shibuya office. In 1 hour I got various explanations, quite interesting. The building I\u0026rsquo;m living in is ~50 years old. I was informed that an investigation had not found any asbestos in the walls. In case of earthquakes, the building might need repairs. There are also floods and typhoons, so a hazard map and instructions were also handed out.\nIn the flat, when using up resources to a certain level, I have to pay - above the level the renting company is paying.\nUsually, the contract is then prepared, the tenant gets a paper with bank details, goes to the bank and returns with a certification paper that the money has been transferred. On that basis, the tenant completes the contract - in signing with his hanko/stamp. This bank transfer did not work in my case. I use Shinsei Bank, not offering transfers at the counter. So I fetched the money from the ATM, divided by multiple days as it was extending my daily limit. I had to pay:\nThe security deposit (敷金), a sum of 2 month rents: 21.0400円 (1337€). This money is used to restore the flat in case I move out leaving a mess. To the least, after moving out cleaning of the flat will be paid from that. Rent: my monthly rent is 105200円. As I did not receive the keys on the 1st, I had to pay rent only for the remaining days of the month. Together 29.6560円 (1885€). After paying this sum, I got the contract and a paper which entitled me to receive the keys to the flat.\nFor comparison: the old flat with 27qm I had 10.0000円 monthly rent. I had to provide a guarantor, had to do an extra insurance (20100円 for 2 years). Renewal all 2 years did cost an extra sum worth a month\u0026rsquo;s rent. All in all, I was putting down 48.2740円 (3068€) initially for the old flat.\nThe Utilities I had contracts running for the various resources, had to order them newly now:\nElectricity: I ordered power for the new flat without issues, and after moving called them by phone to stop power. One downside: I had a water-powered energy plan in the old flat. That\u0026rsquo;s no longer offered, and could not be taken over to the new flat. Basic start/stop of power works per internet, but also me calling them could not get the power plan to sustain. Gas: I just use this for showering. Ordering for new flat via internet was no issue, but one has to be in the flat for setting it up with a technician, so also to agree on day/date. To stop supply in the old flat I called them. Water: I also started this via Internet, and called via phone to stop supply for the old flat. Internet: I had 1GBit fiber up into my flat for the old room, and initiated via internet to get that to the new flat. Nice surprise when they called me: \u0026ldquo;You pay now ~4500円 (28€) for 1GBit, how about 10Gbit for ~6500円 (41€)?\u0026rdquo;. I was about to agree, but then heard that it would need 3 weeks to set up while an appointment for setting up 1GBit in the room was possible in 1 week. Although, I just have laptops, Steam Deck and low-power devices around - nothing with \u0026gt;1GBit connectivity. 10GBit network card in the PCI slot of the RISC-V board would have been possible.. but more for hack value than really useful. My new flat still had no fiber in the room, I was charged 1.6000円 (101€) for having the fiber pulled through the house into the room. The next tenant will not have to pay that as the fiber is now here - I had that situation in my old flat. Of course, workers entering the flat take off their shoes, wear a mask. When they have a small ladder, the ladder legs are prepared to not leave scratches.\nI had ordered power/gas/water with overlapping time spans in both flats, just Internet was switched over without overlap - then it was time to move my belongings.\nPhysically Moving Over I realize that since 2001, when I moved to Munich, I have spent most of my life in small rented rooms, knowing I would move out eventually.\nMy goal is to eventually live again in something I own - but I\u0026rsquo;m not yet sure whether that will be in Germany or Japan.\nWith that in mind, I try to keep my belongings in Japan to the minimum, for example, I have no refrigerator. My table (from Ikea) can be dismounted, then there is a chair. I sleep not on a solid and heavy bed, but Japanese style on futon, comparable to a mattress. Easy to stow away over day, and to transport. Just the washing machine made this a bigger topic: it would require transport by a bigger car. 6 years ago, I had bought a used machine, and considered now to buy a new one - which can include services to dispose of the old machine. The issue: these companies expect to fetch the old machine from the same place as where they deliver the new machine.\nEventually, I looked up \u0026lsquo;引っ越し\u0026rsquo; (\u0026ldquo;moving flat\u0026rdquo;) on the internet, and saw a small company. I did send 4 pictures of my belongings (desk/bed/washing machine) and got a cost estimate of 1.3000円 (82€), and was very happy with that. In 2 hours, we drove the pieces to the new flat, including washing machine.\nFun fact: the moving company had rented the car at a company called 赤帽, translated \u0026lsquo;red hat\u0026rsquo;, which matches the name of my employer.\nRegistrations At the post office, I filled out a paper for having letters etc. forwarded to my new address the next time. My utilities like power/gas/water/internet were already aware of the new address as they supply me directly.\nAt the town hall I had to announce my move - as I moved inside the same ward (Tokyo center is divided into 23 of these), this required only visiting one town hall. I also have a \u0026lsquo;my number card\u0026rsquo;, a card with a certificate with which I can prove my identity for some services. At the town hall, also my address stored on the card got changed.\nFurther places to announce the new address, all done via Internet: bank, credit card company, mobile provider.\nSetting up the new flat For now, I mostly live out of the boxes I used for transport. The lockers in the flat have shelves made from wood with small splinters, so I put plastic onto these and put my belongings on top.\nRegarding the tatami mats: I sleep on top of these now, and will see how it goes. That room has a special smell due to the mats. I will also have to do something about cleaning: the mats can not be wiped with cloth and water, I might have to get a vacuum cleaner for this - or there is a method to lift these up, put them outside and clean them.\nFlat specialties The entrance area: one is expected to take off shoes here and never enter the actual flat with streetwear. House shoes can be used in the normal rooms, but not on top of the tatami mattresses. Following tradition tightly, one would also store extra slippers in the toilet, dedicated to be used there.\nThe bath has also some specialties over European baths. For Japanese traditional style washing, one would use a flat stool (i.e. 20cm height), and sit down there for washing. I\u0026rsquo;m showering while staying in the room. The bathtub features a solid cover: these are meant to keep water warm. Traditionally, Japanese use the bathtub often. The family would then first wash in front of the tube, then use the tube for a while, and keep the water in for the next family member. The cover is meant to keep the water warm.\nThe walls: they are meant to be returned as they are handed over. Stores in Japan have all kinds of hooks that have a cover of use adhesive and can be attached to the wall (hoping they are solid enough to stay, but can later be removed without traces). Other hooks have one or more needles in the back, trying to attach to the wall.\nBike handling is really great here: in the past, I had my bike in front of the house, with a bike cover on top to protect from rain. Here, it\u0026rsquo;s commonly accepted to take the bike with the elevator up to the floor where your room is, and store the bike then in a room. So it\u0026rsquo;s protected from rain, and a good place to do repairs.\nPower outlets are also special: they do not have an easy to use pin for \u0026lsquo;ground\u0026rsquo;. Normal outlets have just 2 stripes of metal, 110V AC. One of the 2 is a bit longer and has a an additional function, but for example for wasching machines where shortcuts can be really bad, you want a real ground. For this, an additional small piece is provided: you open the lid, there is a screw, and then you screw the ground cable..\nWhat\u0026rsquo;s left: organizing payments of the monthly rent. UR is offering to conveniently pull this from your bank account - if you have a compatible bank, which I have not. Banks in Japan are not as compatible with each other as they are in Europe. I have friends who solved this situation by opening a new bank account. Together with maybe an employer who can not pay you on your existing bank account, you quickly end up juggling around money between multiple accounts. For now, it seems like I can get away with monthly transfering the rent from my bank account to UR.\n","date":"2023-09-18T13:52:04+02:00","image":"/posts/2023/09/moving-flat-in-tokyo/20230818_024441936_west_washitsub.jpg","permalink":"/posts/2023/09/moving-flat-in-tokyo/","title":"Moving Flat in Tokyo"},{"content":"I was born in former East Germany. At the German unification in 1990, I was 12 years old. A recent TV documentary made me rethink on my East German memory, my take aways about that time - also including how my parents and East German friends perceived that time. How influenced unification and the time after my thinking and my life?\nEast Germany memories Just a view pieces, but they form a picture. Many foreigners nowadays think of socialist East Germany as a \u0026lsquo;kind of North Korea\u0026rsquo;, by all I experienced and took away from talking with friends, that does not really apply.\nEast Germany was socialist, so what gets produced was almost completely steered by the government. With that implementation, the needs of the citizens were not as satisfied as for example in capitalist West Germany - but while certain products were always short, people in general had enough to eat. Jokes about shortages of goods and Socialism/central management were common, for example:\n\u0026ldquo;What happens when socialism is introduced in the Sahara?\u0026rdquo; \u0026ldquo;For a while, nothing. Then they are running out of sand.\u0026rdquo; My dad worked at the \u0026lsquo;German post\u0026rsquo;, was involved in planning phone lines. With such a position, being able to influence who gets a phone line to their house, we had sometimes better chances to get rare products like banana.\nWhile capitalism did lead to better satisfaction of needs, also the socialist system has strong points. East Germany was short on resources like steel and so on. Let\u0026rsquo;s consider electronic products like toasters for the kitchen. In capitalist systems, the vendor has an incentive to design the product to work for 2 years, the guarantee timespan, and have it fail shortly after. In the socialist system, it was worth to think longer: if you could use that product 4 or 10 years, then all in all you would need much less resources. I still own and use electronic devices produced in East Germany.\nConsidering the whole east bloc, so also Soviet union, Bulgaria and so on, it is often said that East Germany was treated favourably, as there was a high chance of West Germans and other Europeans to come to nearby East Germany - and seeing \u0026lsquo;what socialism could bring together\u0026rsquo;.\nAs for my dreams as I was growing up: I understood enough of the world to realize that realisticly, I would only be able to go to east bloc countries. I thought of America as the \u0026rsquo;land of the free\u0026rsquo; - I would not sign that nowadays..\nMy hometown had an interesting fate after the second world war: at first, the American army was marching in from the West. Later, when the war winners approached Berlin, Russia had taken the biggest part of Berlin. Eventually they handed over parts of Berlin, but the Americans pulled back and my hometown became part of East Germany.\nSurveillance and elections When mentioning East Germany, also surveillance is often mentioned. East Germany was with force keeping their citizens inside. My grand-grandma lived in Hamburg/West Germany, so my grandma was from time to time allowed to visit her. That was also possible as the rest of the family was in East Germany, so chances of her staying in West Germany were low.\nAfter the unification, many actions from the stasi (secret police) became known - but at that time not much was known to normal citizens, most were just living their life in East Germany without ever having contact or even being aware of them.\nEast Germany had elections, but the results were decided already upfront. On paper, there were also multiple political parties. It was important to vote at the elections, not doing that could get you called to your boss and have bad outcomes. As the results were predetermined, the activity was commonly not called to \u0026ldquo;go voting\u0026rdquo; but \u0026ldquo;go folding [the paper]\u0026rdquo;.\nTechnology In terms of technology, the east bloc was behind the west in most areas. Space was one exception, with Gagarin being the first human in space. In East Germany, we got catalogs with products from the west (\u0026ldquo;Otto\u0026rdquo;, \u0026ldquo;Quelle\u0026rdquo;), showing products which could be ordered in the west. Relatives from the west were sending parcels, opening these was celebrated. Chocolate, soap, other things: opening these filled the whole room with a certain smell.\nDrugs East Germany had \u0026lsquo;weak money\u0026rsquo;, not much wanted internationally. As per that, no big organizations saw a market to import drugs, so there was no problem with cocaine and so on. Alcohol and tabaco was legal and sold. The drug problems of the west were often pointed out, as an example of something bad, which socialist East Germany did not care about. With the unification, many of us were worried about being brought into drug addiction. In reality, I think now that most people do the first step themself and use drugs for stress relief or such, and then slip into addiction. Substances like jimson weed (Ger: Stechapfel) were used in some circles in East Germany as drugs.\nFreedom of speech and media As commonly known, there was not much freedom of speech. As East Germany was near to West Germany, it was easy to catch Western radio broadcast, or BBC - which required to know English though. Most Germans also were able to receive West German TV with directed (Yagi-) antennas. There is one issue though: as these are directed antennas facing towards the sender, on top of the house one would clearly see when the antenna was facing towards the west. Thus, many years, people had to setup the antenna below the roof, then potentially having to go for a bigger model due to absorption of the roof. Later (1970?), that did no longer matter and also West facing antennas on the roof were ok. There was at least one area in East Germany which was geographically not able to receive Western media, they were called \u0026ldquo;Tal der Ahnungslosen\u0026rdquo; (\u0026ldquo;valley of the people who do not know\u0026rdquo;).\nUnification time My first memories regarding that were talks of my parents about citizens fleeing to the west, using routes via Czechoslovakia and one or 2 further countries.\nSome days later, we were able to go over the border to West Germany. Our family crossed the border with a car, it\u0026rsquo;s 38km from my hometown Mühlhausen to city Eschwege in West Germany. Over there, people were along the streets, greeting us, handing over chocolate. Everybody got an amount of west money, to buy things.\nIn school, me and others entering school at the same year, we were the last ones to get to learn Russian as their foreign language. I think it was taught quite badly - I live now in Japan and see English here being taught in the same, bad way. After the unification, most of us stopped learning Russian and learned English - of course, our English level was below the level of someone who learned English right from the start.\nOur education options expanded with unification: we could now choose between 3 levels of school. I went to the \u0026lsquo;gymnasium\u0026rsquo;, and graduated with a certificate allowing me to go to the university. My parents were happy for me about these new options. I went to the army for 1 year of duty, that was oblicatory at that time. After that I got a job training, specialized in a certain area of IT - so special that in 2001, after graduating, I did not find any company who could use my skills in the radius of 50km around my home town. So I moved to Munich for work.\nMoney of that time is a fascinating topic. East Germany did not only steer production of goods, but also the price. Milk did cost the same everywhere in East Germany. As a child, you got pocket money from your parents, and also in East Germany you would with that learn that you could eather buy some sweets, or save over months and buy something bigger. The thing hitting all of us, children and grown ups: the market of the west did lead also to different prices. In that time, some people drove dozens of km with their car to buy milk for 80 Pfennig instead of 90 Pfennig, so the difference in price not justifying the additional effort. This was something strange to learn.\nFreedom also changed: in East Germany, it felt like people were more equal regarding money, the gap between rich and poor was not as big. After all, the factories and so on were almost completely owned by \u0026rsquo;the country\u0026rsquo; and not single owners. Most people had some money, but the products were limited. After the unification, most people felt like they had no money any longer (harder to earn now, and the exchanged money was not eachanged at favourable rate), but now one could buy everything. :)\nChildcare changed hugely. In East Germany, efficiency was lower than in West Germany, the governments goal was to have everybody working. There was care for small children, and after school the kids were looked after until evening. In the computer club, I and many others got their first contact with computers. Older children got members in the clubs of the \u0026lsquo;pioneers\u0026rsquo;, giving feelings of community, and providing activities. Of course, these clubs as well as special school lessons like \u0026lsquo;Staatsbuergerkunde\u0026rsquo; where also used to indoctrinate, and convince children of the benefits of the socialist system. After the unification, all of this care for children stopped, a huge interruption: what should children do after school ends in the early afternoon?\nDrugs were hardly known in East Germany, the East German money was not popular enough to import these and earn money.\nThese got introduced after the unification.\nHealth did not change much. East Germany had polyclinics, hospitals for a wide range of treatments, also vaccines were given widely, the health system was quite good. After the unification, health got a bit more a question of money again, it matters whether you are in a private or public health insurance, and so on.\nResources changed too: in East Germany, we children did collect glass bottles and newspapers, and got money for that. There were recycling systems, which were taken down with the reunification. Resources got used much more wasteful after the unification, and now as the climate crisis can no longer be denied, people start to understand that capitalism is leading to wastefulnes.\nLater conclusions It took me some years to come to the biggest realization and learning: in East Germany, the Soviet Union and their people were our friends, or \u0026lsquo;brothers\u0026rsquo; in socialist terms. We learned Russian in school, we had pen pals, we visited each other. Right before the unification, the West German government met Mr.Gorbachev and kindly asked for permission to reunite, which was granted.\nWhile commonly called \u0026lsquo;unification\u0026rsquo;, in law terms this seemed to be more a \u0026lsquo;East Germany joins West Germany\u0026rsquo;, this has been critizided in detail by various specialists. Politically, with the unification all the media changed from one to the next day. While actually nothing had changed, the media was now steered by West Germany and trumpeted now: these strange Russians over there, they are the opponent and will certainly invade us next week. This sudden change in media reporting, which we had to notice because it was the exact opposite of what we heard before, that is my biggest take away from the unification.\nHow different were people in East and West Germany at the unification time? I did not spend to much thinking on that and moved to Munich for work. We all spoke German, what should be different? There was quite some.. everybody had grown up with other ideals, learning English and French instead of Russian. People had listened to different music, seen different manga/comics, different movies, to big part also read different literature. Many years later I started to learn Japanese, and along with that got introduced into the Japanese culture. As strange as it sounds, as that was happening more slowly and I was prepared, it went much more fluent than my move to Munich.\nThe song I associate with the unification: Scorpions: the wind of change. I\u0026rsquo;m living in Japan now, and sing this every now and then in Karaoke bars.\nSince moving to Japan in 2016, discovered some fun parallels between Japan and Socialist East Germany, English version of a slide pack I made for colleagues is here.\nThe picture at the top shows a wall at the \u0026lsquo;propaganda\u0026rsquo; caffee in the area of Tokyo where I live. This here is at the Odakyu trainstation in Shimokita, the style remindinds me of East Germany whenever I pass by: When I\u0026rsquo;m asked about the national anthem, the first words coming to mind are still the words of the East German anthem. Would I want East Germany back? Certainly not, but there certainly were also good memories, good approaches to some things.\n","date":"2023-07-28T13:21:59+02:00","image":"/posts/2023/07/gdr-memories-and-german-unification/20171007_143801_propaganda_kaffeb.jpg","permalink":"/posts/2023/07/gdr-memories-and-german-unification/","title":"Memories of East Germany and the German Unification"},{"content":"I have some retro hardware in Germany - in the last weeks I had access, and got to play, try out, find out what works. Capturing the summary here, so I can lookup further activities for the next time I get access, and maybe for others to share ideas for fixing and further debugging.\nRobotron EC1834 I was born in former East Germany, before the German reunification. Around the unification time, I got a \u0026lsquo;Robotron EC1834\u0026rsquo;, one of the last personal computers of East Germany, released 1986.\nNow I got curious and gave it a look.\nThe hardware is really fascinating. East Germany was short on resources, and production of goods was steered by the government. This had many downsides, but also did set some good incentives: producing something solid would last longer than nowadays typical products which are designed to fail after the warranty period ends. Looking at the hardware, I was reminded of that: it looks solid and durable. 256kb RAM on the board, and an extension card with another 256kb. The extension cards are not connected via ISA connectors, but the same protocol is spoken. The east bloc was far behind on semiconductor technology: the CPU is an Intel 8086 clone, produced by Siemens and which got imported into east Germany. Many other chips have \u0026lsquo;сделано в cccp\u0026rsquo;, so \u0026lsquo;made in udssr\u0026rsquo;. The docs are in German, but most descriptions on the board are in English. Caution notes are in German, English and Russian.\nThere is a 20MB harddisk, with sector/head details written ontop by hand. Extra cards: CGA graphic card, memory extension, controller for the 5.25\u0026quot; floppy drive, card with printer port.\nAlthough I did not get to use the box, all in all things look promising: the fans spin up, also the harddisk, and 2 control beeps come out. The monitor is not displaying anything, I suspect that\u0026rsquo;s the issue. The output of the graphic card is a normal sub-d connector with 9 pins, like often used for serial connections. Potential next steps:\nsolder a VGA cable with a sub-d connector. Frequency of the VGA monitor also has to be able to process what the CGA card offers, though. or get a CGA to HDMI converter, these can also do MDA (monocrome) and other modes or investigate with an osci if there is a usable signal from the CGA card at all, so far I\u0026rsquo;m just assuming that No win here.. but things look not bad either. Details on the EC1834 are here.\nCommodore C64 After the German reunification, I also bought a box with 2 Commodore C64, 3 datasettes (players for music cassettes, which can carry data), 2 floppy drives. Tried these out for the first time, with video/audio output to a normal coax cable and with a TV for displaying.\nBoth C64 work, but one has no audio output. Input via floppy or cartridges work. My next quest was to run a C64 demo from the internet. For that, I connected a TEAC 5.25\u0026quot; to an old Linux system running Debian. Seems like Linux should be able to write C64 disks with normal 5.25\u0026quot; drives, but I did not get the drive to run at all. The drive had an old connector which looked much like a normal ISA-bus extension card, I had only one cable with that connector around. The BIOS allowed me to configure the drive, and Linux found it, but when trying to mount floppy disks I only got traces from the driver. Floppy kernel module with debug parameter did not help.\nHappy to see one of the C64 running, but I did not manage to write C64 floppy disks.\nAmiga 500 I had an Amiga 600 when I was young, later sold it to get a PC. My dad gave me this Amiga 500 some years ago, it\u0026rsquo;s fully working. In the last weeks, I took out the extension card to check for leakage: the card has 512kb memory and a realtime clock - which had indeed started to leak. I took it out - now the Amiga reported only it\u0026rsquo;s internal 512kb, and did not detect the extension card. There was a customization on the extension card: 2 wires and a switch which could be used to turn the extension card off. Fiddling with the card had led to the switch no longer working, but after replacing it with a jumper, the extension card is found again.\nOne minor thing is odd about the A500: the power supply is one of the first types of \u0026lsquo;A500 power brick\u0026rsquo;, quite heavy, and when switched on it emits a high frequency noise. A bit like when turning on a cathod ray monitor. I got a hint that the coils could have started oscillating, and hot-glue could be used for fixing. Did not have hot-glue around this time to try out.\nI have an external 3.5\u0026quot; floppy drive which is reported as \u0026ldquo;DF1:BAD\u0026rdquo; from the workbench. Also lubricating the sled and using contact cleaner on the plug did not help.\nThe Amiga floppy drive can handle DD-disks, \u0026lsquo;Double Density\u0026rsquo;, so hold around 800kbyte. Unfortunately, PC floppy drives can not write DD-disks in a way that Amigas can read them - so I could not use floppys for data transfer between PC and Amiga.\nSince christmas last year, I have a serial connection between the Thinkpad running Linux and the Amiga running: with this I can have the PC send ADF disk images, and the Amiga write it - so I can write and execute latest demos for the Amiga, and games.\nI actually also run short on floppy disks: these are no longer produced. I have only a few DD-disks around, these are best suited for the Amiga. Also having HD-disks, which have a different coating, and I noticed that I can also use these on the Amiga - word has it though that they are not as reliable with the Amiga as DD-disks.\nHad some rounds of \u0026ldquo;Supercars 2\u0026rdquo; and \u0026ldquo;Bubble bobble\u0026rdquo;, still great!\nAmiga 1000 My best outcome was with the Amiga 1000: I had obtainesd it years ago, and never turned it on.\nI got the A1000 bare though: no disks with firmware or workbench, no keyboard. I could use the video cable from the Amiga 500, also had power cable around.. and the A1000 worked! Unlike the Amiga 500, the firmware is not on an eeprom by default, so I used the serial connection between Linux PC and Amiga 500 to write kickstart- and workbench disks, and then I could verify the A1000 boots into workbench just fine!\nI could connect the joystick and the mouse which I use with the Amiga 500, worked fine. Also most games and demos worked nicely.. just no keyboard input possible.\nMet some friends, mostly around the IT-world in these days, and one mentioned that he has still 2 keyboards for the A1000 around. Maybe there is a chance to get one, next year.\nFuture Hoping to get further with the EC1834.. maybe Christmas time. Would be really nice to have an 8086 compatible box running again.\nMaybe even with some kind of networking, the system has a modem! You might also have seen the Sun Ultra 5 in the background.. it has NetBSD 1.6 installed and should be upgraded, or even better get Solaris 8, the native system of that era. But while NetBSD can nicely be installed via network, I did not get Solaris installed via network last time I tried.. and the cd-drive has issues.\n","date":"2023-07-23T07:21:11+02:00","image":"/posts/2023/07/retro-summer/05_memory_card3_slim.jpg","permalink":"/posts/2023/07/retro-summer/","title":"Retro Hardware fun of Summer 2023"},{"content":"Transport in Tokyo I\u0026rsquo;m a German, living in Tokyo since 7 years. Cycling is one of the hobbies I took with me to Japan. How to people get around in Tokyo?\nBy train/bus: there are many train stations, e-payment is easy, it\u0026rsquo;s cheap, on time - so this is by far the most heavily used method By car: this is much less used. To register a car, one has to prove that the car has an own parking place. Parking in Tokyo is quite expensive. By bike: this is often used to transport children, groceries, sport. Less so for commuting. As for terms: bicycle in Japanese is 自転車/Jitensha, literally \u0026rsquo;the machine which moves by itself\u0026rsquo;. When talking with Japanese, one would typically not say \u0026lsquo;bike\u0026rsquo;, as this sounds exactly like the Japanese work for motorcyle.\nPurchasing After moving to Japan, I was moving around by foot and public transport, but I quickly also started to search for a bike. The most common bike is the \u0026lsquo;Mamachari\u0026rsquo;, a bigger bike intended to transport groceries or bring chilren to school. This starts around 200eur, but was not appropriate for me. Bike stores are all around, they have also sport-, mountain- and roadbikes. Problem: with my 1.86m I\u0026rsquo;m rather tall by Japanese standars, and bikes in the right frame size were rare.\nI considered shipping my bike from Germany: would cose at least 800€. Bringing the bike as \u0026lsquo;oversized baggage\u0026rsquo; with me in a plane would cost 150€ - but appropriate packaging for transport is a topic, as well as getting the bike to the airport in the first place.\nI investigated the second hand market as used goods in Japan are typically in very good condition - and foreigners (more likely to be tall) also sell there. Eventually I got myself a \u0026lsquo;Giant escape R3\u0026rsquo;, for around 5万円. Turns out \u0026ldquo;French valves\u0026rdquo; are mostly used, in Germany plain car valves are more common, so refilling air at the gas station is easy. Dynamo are not very common here, I got a LED front light, rechargeable battery is attached and supplies 3h light, recharging is done via USB-micro connector.\nThe brand is \u0026lsquo;Giant\u0026rsquo;. In Germany, I always considered this as an excellent name, so was proud to get one for a good price. In Japan, bikes of that brand are not really considered any special - I suspect the difference is simply as the company is rooted in Taiwan and transport to Japan is much cheaper than to Germany. So it\u0026rsquo;s more rare in Germany.\nDirectly at the store I registered the bike with the police: my name and address got mapped to a number, which is imprinted on a sticker on the bike, and probably also a harder to remove frame number. If stolen and found somewhere, the bike can be mapped to me.\nBiking infrastructure, cycling laws, helmet My first goal for a trip was to commute to the office: that took 50min by public transport, and 2h by bike. That was a rough ride: the rules seemed quite frightening. Many cyclists are moving from the walkway to the road without watching over their shoulder. At pedestrian crossings with traffic lights, cars are stopping when thr traffic light is red - but the majority of cyclists is ignoring that when no pedestrians are interfering.\nIn Japan, like in Great Britain, cars drive on the left side of the road. Some roads in Tokyo have bike lanes, but bikes are in general ok to use the footway where normal pedestrians walk. On many streets with biking lanes, one sees parents transporting chilred in the mamachari on the footway.\nAs for laws: after accidents, there is a strong tendency to consider the vulnerability of involved parties. This means that in accidents with car and bike involved, the bike is likely to be seen favourably. For accidents with bike and pedestrian, the pedestrian is likely to be seen favourably.\nAs for helmets: since April 2023, we have 努力義務 to wear a helmet. This is not a \u0026lsquo;you are forced to wear helmet\u0026rsquo;, in the same way as people were not forced to wear mask in the Corona time. Literally, the term says \u0026lsquo;you have to do your part for helmet wearing\u0026rsquo;. Practically, there were only a few helmet wearing bikers in April, but after a few months it\u0026rsquo;s now picking up. Watching what others do and following along is an important part of the society here, I suspect that when a critical mass of people adopts the helmet, it will accelerate acceptance further.\nIn my part of Tokyo, every bike driver has to purchase a liability insurance, that\u0026rsquo;s 1400円 per year. This can be obtained by the local post office, and be payd in cash.\nRepairs I\u0026rsquo;m not doing much maintenance myself apart of cleaning the bike chains, applying fresh oil to the chains, and refilling air at pomp stations. For getting flat tyres replaced and so on, I use repair shops. In Germany, these would take in the bike and have me wait for 3-5 days for repairs. In Tokyo, these shops are short on space and try to not take in your bike over night: for most repairs I can fetch the repaired bike after 30 or 60 minutes. Prices are reasonable, and as this is Japan, no tip (German: Trinkgeld) is given.\nAccidents In summer 2022, I was with the bike waiting in front of a traffic light, I was located on the normal lane which is also used by cars. A car did hit me from the back, apparently the driver had not seen me. The back wheel had to be completely replaced, and I had slight back pain. In that situation, the car driver affered to call the police, but I just wanted to get along and refused this. The car driver left his name and mail address, but interestingly did not hand out his business card.\nI went to ~5 bike shops until I found one which had a wheel for replacement, having to order one would take weeks or months. I informed the car driver via mail about the repair costs of ~150$. He offered to \u0026lsquo;give me 200$, to compensate for the pain\u0026rsquo; - which I refused. I turned out to have back pain for a few weeks, not compensated with 50$, so I asked him to just pay the 150$.\nAll of these mails were plastered with \u0026ldquo;I\u0026rsquo;m sorry\u0026rdquo;, but this is one aspect of Japanese communications I do not like: the driver would say this simply because he was supposed to in this role. I could not evaluate whether this was meant seriously or just following the expectations.\nMy biking occasions/activities In 2018, I moved to Shibuya area - aiming at just the right distance to the office so I could commute by bike. Then came Corona, 1.5 years without commuting to the office.. but now I go to the office most days of the week, 25min per trip.\nApart from that, I go to the Onsen/bath once per week, alternating between 2 very nice bathes. Each trip is ~1 hour. I mostly listen to podcasts when biking, using one in-ear-plug, so I still can take in all traffic noises happening around me.\nFrom 2019-2021, I did the Tokyo metro biketrip: visiting all 292 metro stations in the Tokyo metropolis area by bike, I took a picture of the entrances. 7 new stations had been built after starting the project. The idea was to visit various places around the city - which worked out nicely! I can highly recommend this approach.\nAppart from these, I also go to events by bike, for example the Nibelungen opera performance, to concerts, to do groceries, and so on.\nRoute planning I use a \u0026ldquo;Finn mobile holder\u0026rdquo;, which looks like a rubber strap which you wind around the handlebar of the bike, and can easily attach your smart phone for map guidance.\nI have having to confess it, but Google maps is my \u0026lsquo;quick and dirty\u0026rsquo; app for setting markers on the browser interfeace, then starting the app on the mobile and using the routing functionality. For bigger trips, one can use brouter.de for route planning, then export to files and import these into OSMand on the mobile. With OSMand, one can also use offline maps, saving electrical power on longer trips. Details are here, with my experiences from a recent 250km-in-3days trip.\nIn Japan, also the Yahoo maps app is an alternative to google maps. Both are closed and proprietary, but Yahoo has better maps in some regions.\nSummary If you live in Tokyo or somewhere else in Japan, I can really recommend to get a bike. The local specialities might take a bit to get used to, but in general biking is a nice experience here, one feels safe.\nDid I forget an aspect? Please let me know!\n","date":"2023-07-02T16:08:24+02:00","image":"/posts/2023/07/biking-in-japan/20210928_051017929_rental_bike_800.jpg","permalink":"/posts/2023/07/biking-in-japan/","title":"What's cycling in Tokyo like?"},{"content":"CPU and Memory The Star64 board is a bit bigger than the Raspi4, both are available with 8GB RAM. Both have 4 cores at up to 1.5Ghz, the Star64 has a RISC-V cpu, the Raspi ARM.\nMy first goal was to look at power consumption, and cpu temperature. At 20°C environment temperature, without attached cooler, the Star64 cpu idles at 46°C. The board pulls 4W then. With a single core under full load, 6W are consumed and the temperature rises to 54°C. With all 4 cores under load 6W and 62°C.\nWhen running 4 \u0026lsquo;md5sum /dev/random\u0026rsquo; processes, we see 93% system load and 7% userspace load, that\u0026rsquo;s a bit different from x86 with 70% system load and 30% userspace. The Star64 also has one systemd-udevd process constantly I/O blocked, in runstate D. So even in idle, the load is \u0026lsquo;1\u0026rsquo;.\nhttps://github.com/ThomasKaiser/sbc-bench is a benchmarking suite for small systems/SOCs. When comparing with the results for a Raspi4b, we see that the Raspi is for some operations 5x as fast as the Star64. It also scores better for memory latency tests. For openssl speed test, Raspi4b is 3x as fast. For 7z tests, the Raspi4b is 1.4 times as fast.\nConverting a 4 hour podcast from m4a to wav format took 17sec on a 4.8Ghz Intel cpu, single threaded. It takes 507sec on the 1.5Ghz cpu of the Star64.\nLooking at I/O The Star64 has 3x USB-A connectors speaking USB2, and 1x USB-A connector for USB3. The board also has pins to connect a further USB3 connector, and a further USB4 connector.\nI have an external enclosure which can speak USB3, with a 1TB NVMe inside, but I just got it to work via USB2 ports of the Star64. When connected via USB2, I saw ~70% of the transfer rates I see with other Linux systems. A spinning harddisk could be connected via USB3, and the access speeds were as expected. That drive consumes 5W by itself, so it\u0026rsquo;s good that the Star64 is supplied via extra power supply.\nI saw situations where \u0026rsquo;lsusb\u0026rsquo; did hang, stuck reading bluetooth related data below /sys.\nNetwork speed of the 2 Gbit RJ45 connectors: no surprises, expected throughput and latency.\nSummary While the Star64 might seem usable already for a few things, it\u0026rsquo;s flagged as \u0026lsquo;for developers\u0026rsquo; right now - and that seems justified. It\u0026rsquo;s somehow refreshing after quite some time to deal again with a hardware where you see kernel backtraces, and not yet all hardware is supported.\nThe board has some features like PCI-E slot, 2x Gbit ethernet, which other boards to not have.\nRaw CPU/memory performance is behind the Raspi4b, but I needed to look at benchmarks to notice that - it\u0026rsquo;s sufficient for many use cases. I\u0026rsquo;m surprised the performance is in the same ballpark as ARM at all, considering how young RISC-V is. Performance might also change a bit in the future with better kernel and asm-optimizations for the cpu.\nAs for Linux distros, I have just used the builds from https://github.com/Fishwaldo/meta-pine64 so far, these are Debian based. The Star64 also works with plain Debian/bookworm userland with non-debian kernel. Not yet sure about Fedora. As my NVMe enclosure can not be connected via USB3 right now, playing is restricted to microSD-card/eMMC.\nWhat might also be fun looking at: comparing power efficiency with Raspi and Intel systems. I just have smartplugs running Tasmota here for metering, for the low consumption of Star64 or Raspi that\u0026rsquo;s not accurate enough.\nI also did not look at Star64 for desktop usage. Drivers for display and audio already exist and are said to be usable. My wiki page around the Star64 is here, I will update it with new findings/developments.\n","date":"2023-05-15T13:09:07+02:00","image":"/posts/2023/05/star64-first-look/20230514_23225382_star64b.jpg","permalink":"/posts/2023/05/star64-first-look/","title":"First look at the Star64 w/ RISC-V cpu "},{"content":"The Golden Week This years Golden Week approached, a sequence of national holidays in Japan. Like the last 2 years, a friend invited me to help with planting rice. I live in Tokyo, he lives in Tochigi-ken, north west of Tokyo. In the last years, I did day trips: 2h30min by train, then my friend fetched me by car and we drove another 30min.\nI have a roadbike which I typically use for up to 40km day trips. Google maps estimates 140km if walking over to my friends house, and it is not able to suggest a cycling route. Cycling these 140km in the morning before rice planting would not work out, but the day before rice planting was also a holiday, so: let’s go 122km to Utsunomiya on the first day, stay over night, and go another 30km in the morning before rice planting.\nRoute planning: energy I use no bike computer with power efficient e-ink display, but a mobile phone strapped to the bikes handlebar.\nWhen I did my \u0026lsquo;Tokyo metro cycle project\u0026rsquo;, where I visited all metro stations, I had already exceeded what a single charge of my Pixel 4a phone could do, and had used a second phone. I could use a fully charged mobile for 2 or 3 hours, longer if mobile connection was disabled and if I could turn off the screen for a longer time.\nMy mobiles at hand:\nAndroid Pixel 4a Androud Huawei iphone 13 (company) The plan is to start with these 3 mobiles, and a Thinkpad with 57Wh battery, which can charge a phone when connected.\nRoute planning: software My desired features:\nOpenSource running one software on the Android mobiles and the iphone designing a route on the Thinkpad running Linux, and then transferring it over to the mobiles. Same for \u0026lsquo;points of interest\u0026rsquo;. the phone app should support offline navigation My first part of the route was from home in Tokyo/Shimokita to Utsunomiya, ~120km. Google maps is not even able to draw up a cycling route for that, besides that it\u0026rsquo;s not doing offline routing.\nI ended up with planning the routes with the site https://brouter.de/brouter-web, a free page using OpenStreetmap material.\nOn https://openstreetmap.org, you can access maps with pathes extra suited for cyclists - brouter.de is using these if possible.\nOn the mobiles I use OsmAnd, on Android it\u0026rsquo;s in the Fdroid store. I use Syncthing for file sharing between Linux based devices, it basically provides directories which get synced up between the devices. With that, pictures taken on my Android phone get automatically synced to my Linux desktop. I made a new directory \u0026rsquo;trip\u0026rsquo;, saved the route created with brouter.de as .gpx file, and could then import the route on OsmAnd on the Android phones: \u0026lsquo;plan a route\u0026rsquo;, \u0026lsquo;import track\u0026rsquo;. Then I can start the navigation, click the gear (settings), and \u0026lsquo;follow route\u0026rsquo;.\nSadly, that is not yet implemented on OsmAnd on the iphone.\nNext: points of interest. On my Pixel 4a mobile I stored markers (\u0026lsquo;favorites\u0026rsquo;) for the important places of the route: home, hotel in Utsunomiya, my friends place, one hotel where I will stay when going back. These markers can be exported into a .gpx file, and imported on OsmAnd running on the other phones. The marker import works also on the iphone.\nSumming up, this was the procedure:\nOpen brouter.de/brouter-web, set start and endpoint, have it plan a route. Compare with the route other tools would plan, adjust it to prevent steep elevations. Export the route to .gpx file, copy the file to the other phone(s) Import the .gpx on OsmAnd on the other phones, then navigation can be opened, and below ‘settings’ one can choose ‘Follow track’, where the imported track appears On OsmAnd on one of the phones, store all relevant locations for the trip as ‘Favorites’, then export these as .gpx file and import them on the other phones. Store notes on Obsidian on Linux, use a location which gets synced to the phones and can be opened from Obsidian on the phones On the phones, zoom into the locations you will travel to, so you get offered to download the maps of the local areas How to get files to the iphone? I did send mails with attachment to a mail account, which I did access from the iphone.\nAn other good candidate is bikemap.net.\nConvenient syncing of \u0026lsquo;points of interest\u0026rsquo; between the browser on Linux and on the mobiles stays unsolved.\nSyncing of notes: preparing the trip, I started to use Obsidian. On Linux and on Android I use the Obsidian software, which is basically a fancy editor to create markup files. I use syncthing to sync the markup files between all devices, so with that I can modify my trip notes on Linux and later access that on the Android phone - convenient!\nStarting the Trip My tour for the first day was planned with 122km, the detailed route is here. The trip got exported to .gpx, then imported on OsmAnd and then \u0026rsquo;navitation along the route\u0026rsquo; started. The trip took 7h30min, plus 30min for lunch break. I started with the Pixel 4, when used up I switched to the Huawei - and that run out of power just as I arrived at the hotel in Utsunomiya, the goal for the day.\nOsmAnd tells you for how long you will follow the current road without changing roads, I often deactivated the screen then and just activated it 10min later. Both mobiles allow fingerprint authentication, so just putting the finger at the backside of the mobile unlocked the screen. Recharging the Pixel 4 with the Thinkpad worked nicely, so in the evening I was using it when cycling through Utsunomiya and taking pictures.\nIn the 3 days of heavy usage, I had only one crash of OsmAnd - that\u0026rsquo;s reasonably stable to me. One oddity: the tracks I had transferred from brouter.de were basically a long line of coordinated. On the road, when I happened to miss one of these then occasionally OsmAnd hinted me to turn around to \u0026lsquo;catch\u0026rsquo; also that missed point. In these cases, stopping navigating and starting it freshly fixed the issue. This happened ~5 times in the 3 day trip.\nOn the second day, I started at 6:20, a 35km trip to my friends place. In general, I could see that the planned route tried to stay away from big/heavy used roads. On the last part here, I was routed into the woods, so steep that I had to walk instead of cycling. After investigation of the map, I took a different streed - the website had evaluated \u0026ldquo;stay away from big roads, even if it\u0026rsquo;s much steeper\u0026rdquo; differently than I did. For future planning, I will pay more attention to this - once recognized, it\u0026rsquo;s easy to modify the route on brouter.de and then export the modified route.\nRice planting I arrived around 9:30. My friends house is located really nicely: these is a big valley, in the middle there are rice paddys. To the left and to the right, there are roads and farm houses, and the farmers live there.\nThe actual rice planting is now mostly done by machines. Weeks before the planting, young rice plants are grown in green houses, with the goal of the plants having just the right size at the time of planting. The containers where they are grown have the right size to be loaded into the rice planter machine. The rice planter is quite impressive: it has 4 wheels, just of right size to not sink into the paddy. The paddy is water flooded for planting, the planter then drives over the field, and tiny mechanic arms take young plants and put them into the soil. The rice planter machines came up some dozens of years, before that the planting took much longer. Also now, the machine is leaving deep trenches in the soil, I helped with flattening these trenches out.\nWe all used big boots. My feet are bigger than average Japanese feet, so I had brought in my own boots. In the old times, people had done this without boots - but then you could catch bloodsuckers.\nWe had one break in the morning, then one hour lunch break, with a great lunch. As asual, I did not manage to properly eat the fish.\nAt 3pm we were done with all 4 rice paddys. I started route 3 with 48km, and arrived at the hotel in Mibu at 18:20.\nBack to Tokyo My last route was 113km. This was not as long as my route on the first day, but this time I had wind blowing into my face, so this took longer.\nThe planned route guided me along some rivers, that was one. The biggest piece was following ~15km along a river. I also found 2 biker rests, kind of restaurant located near to roads heavily used by bikers, very nice to see.\nOne thing to change for future trips: besides my belt bag, I was carrying boots, Thinkpad, clothes and so on in my backpack, summing up to 10kg and pulling down the shoulders for many hours. Side backs to be mounted to the bike would be a relief here.. but I would almost never need them. Maybe some stores lend these?\nThis was a nice challenge, happy to have done it - considering to do this again.\n","date":"2023-05-01T09:39:42+02:00","image":"/posts/2023/05/bike-trip-japan/20230502_231143457_bike_housefront.jpg","permalink":"/posts/2023/05/bike-trip-japan/","title":"Bike trip planning and Rice Planting"},{"content":"1: What\u0026rsquo;s BSD + my BSD history I mostly use Linux, but since the time of the Linux Install parties with the Thüringer Linux User Group, I\u0026rsquo;m also using BSD. I was lucky enough to meet many BSD folks when I lived in Munich.\nWhat\u0026rsquo;s BSD, after all? My summary: Unix operating systems, mostly with code bases older than Linux. The biggest 3 variants, each with it\u0026rsquo;s own goals:\nemphasizing security: OpenBSD emphasizing performance, and desktop usage: FreeBSD emphasizing code cleanness and purity, as well as running on many different architectures: NetBSD Linux distros are constructed of the Linux kernel, and many GNU software pieces like glibc and utilities ontop. The BSDs include not only their kernel, but also their own set of utilities. Much software like Apache webserver or Xorg is usable on both, BSD and Linux.\nThe development of Linux and the BSD\u0026rsquo;s is happening differently: Linux has more active developers, the codebase is huge compared to the BSD\u0026rsquo;s, things change often. BSD sees less movement, so you are likely to find a config file in the same place, also in new versions.\nAt my previous employer I was running OpenBSD on Sun sparc hardware, also now I have a Sun UltraSPARC 5 running NetBSD. To the folks at AsiaBSDCon I advertised the Sega Dreamcast, Playstation 2 and Playstation 3: all of these are still available at \u0026lsquo;hard off\u0026rsquo; stores in Tokyo, and can run BSD.\n2: Preparing for AsiaBSDCon: installing BSD on the Thinkpad I\u0026rsquo;ve been at various Linux conferences, as well as conferences dedicated to OpenSource in general, like FOSDEM. AsiaBSDCon was announced for Tokyo, 45min in cycling distance from where I live, so I registered.\nUsing Linux on the destop since many years, I challened myself to get one of OpenBSD/FreeBSD/NetBSD running on a Thinkpad, so I could use it at the conference.\nMy requirements:\nMy homedir needs to be encrypted (it’s 2023, right? When a laptop is lost, the data should not be easily readible) support for the NVMe support, and wlan Xorg, Firefox, audio output Nice to have:\nsuspend to memory UEFI Japanese fonts, to read pages with Kanji Japanese input method, so ability to input Kanji In short, the 3 variants can do all of this. I started to investigate disk encryption: possible with all BSD\u0026rsquo;s, but it seemed so complex for NetBSD that I skipped that one. My NVMe and wlan were nicely supported on both FreeBSD/OpenBSD, also Xorg with Firefox were easy to setup, and Japanese fonts. Suspend to memory worked also on both, though not with closing the lid but manually executing \u0026lsquo;zzz\u0026rsquo;.\nJapanese input method was easier to setup on FreeBSD than on OpenBSD. All 3 BSD\u0026rsquo;s have excellent handbooks and man pages, just installation of the Japanese input method was not really well documented.\nSo, equipped with OpenBSD on my Thinkpad I started the first conference day.\n3: AsiaBSDCon 2023 The con happened in a university. Normal attendees did pay for attendence, and there were sponsors. BSDCon covered flight costs for the presentors, I guess most money was spent there.\nPresentations started, and were quite interesting for me. If a user wants to do something, Linux has often many ways and software pieces available. For example there are many file systems, with their own features and level of stability. The BSD\u0026rsquo;s have for many tasks one single piece of software, but that is then covering a majority of the use cases/most frequently requested functions.\nBetween the sessions, it\u0026rsquo;s easy to get to talk with other attendees. Chatting is as nice as with Linux folks and colleagues, one feels on common ground with most things. It\u0026rsquo;s also great to go into politics: fantastic chats comparing Japan/Germany/USA. Or hearing about the state of things in Hong Kong, from someone living there. Also specialists from Ukraina, now living in the US.\nNetBSD is available for many architectures, there was also a both showing many of these. For the first time I saw a Sharp X68030 system, a PC from the DOS era, produced in Japan.\nI learned much in the sessions. The power architecture is also used in desktop systems, supported by FreeBSD. One presentor had written an input method for entering Kanji - but not for Xorg, his input method workds directly with the terminal in plain text mode - fascinating idea!\nOn Saturday evening, we had a banquette/dinner, in a building attached to a beautiful shrine. It\u0026rsquo;s the last days of cherry blooming here in Tokyo, and there was also a dancing event going on.\nFood and talks at the dinner were very nice.\n4: Ideas I took home with me One interesting idea was about storage. We all know the Fio tool, with which we can do I/O load on storage. But which I/O load to choose? Reading, or writing, or both - in which percentages? Which block sizes? Discussing with an other participant, we noticed that this is missing: tool which for example monitors I/O on a production system for an hour and afterwards tells us which I/O patterns it had seen fly by - that could then be usable to do the same load with fio.\nBSD vs. Linux: so far, I assumed one big driver for companies to use BSD instead of Linux: taking the code, extending it, and due to the license not having to give back the modified code. I met NetBSD developers from a Japanese ISP who had an other reason: \u0026ldquo;When we decided for an OS for our router hardware, Linux would crash and just was not stable enough\u0026rdquo;.\nAlso, with BSD there is less movement in the code than with Linux. I experienced that with the BSD@desktop testing myself: you can come back after many years and find the config files in the same place. This makes updating over a long time and many releases easier.\nI also had a nice chat with a TrueNAS specialist. TrueNAS was called FreeNAS in the past, they are an appliance offering storage. Historically they were running on BSD, but following customer requests they also offer TrueNAS now based on Linux: customers wanted to run Linux containers, put the spare cpu cycles of the storage into usage. Most interesting: for most scenarios the TrueNAS folks see BSD in advantage over Linux, performance wise. They are doing performance comparisons. Surely, they have run and optimized the BSD based systems since many years. This really really asking for a more open comparison. For example: a party, maybe an ISP, should offer 2 systems of identical hardware. One installed with BSD, one with Linux. Then there would be time span where folks over the internet could request benchmarks to be run, and where everybody could suggest optimizations, for example as pull requests into a git, and the requests would end up in Ansible playbooks to be implemented on the systems. After for example 2 months, a final workload/benchmark would then be run against both systems. We know capture the flag contests.. this would be an optimization contest.\nAlternatively: offering for example 5 systems of the same type at an event, giving various communities (OpenBSD, NetBSD, Linux distros etc.) each one system, allowing them to optimize for one day for a use case, and at the end of the day doing a final run of a workload via network - and publishing the results.\nWe talked much about file systems: ZFS is one of the big topics, in Linux and BSD world. Due to it\u0026rsquo;s license, ZFS is not getting included into the upstream Linux kernel. When customers ask for ZFS, we from the Linux world try to understand which actual features they are after and suggest alternatives (ask me if interested..). ZFS is available with FreeBSD and NetBSD. The most interesting piece I learned: it\u0026rsquo;s not available natively with OpenBSD due to license issues.\n5: Japan/logistic things I was lucky having to commute just 45min by bike to the event, and that the weather was nice. Starting from April 1st, Japan is asking bikers to wear a helmet. The used term in Japanese is 努力義務, literally \u0026ldquo;you have to do your best to conform\u0026rdquo;. I pondered for a while what this would practically mean: bikers can now be stopped by the police and be reminded to wear a helmet, but there will be no punishment.\nAround half of the visitors of BSDCon were foreigners, but we also had many Japanese.\nLive streaming was available while the conference was going on, but as of April 4th no recordings seem available. I also did not see uploaded papers - so if you are interested in one of the topics I can scan that part from the printed proceedings. I shared my BSD@desktop experiences later as a mini presentation, the slides are here.\nLooking forward to upcoming conferences!\n","date":"2023-04-03T14:12:08+02:00","image":"/posts/2023/04/asia-bsd-con/20230401_bsdbag_flat.jpg","permalink":"/posts/2023/04/asia-bsd-con/","title":"AsiaBSDCon, BSD and Linux"},{"content":"When I started to learn Japanese, I tried to look up Japanese input - and it felt much like an in-club: not having the right search terms, not having a good understanding of the writing system either, and without someone around to ask you were lost. Let\u0026rsquo;s look into this, along with some details about the Japanese language.\nThe Latin alphabet which is widely used has just 26 characters, so normal keyboards simply map one letter to each key. The shift key allows to choose upper/lower case. In Japanese, one should know 2000 Kanji to read a newspaper. These are composed out of one or more glyphs, this picture shows an old keyboard which actually attempted to map multiple glyphs to single keys, and then use modifier keys to select one of these glyphs:\nSuch a keyboard can not be used in laptops, is not user friendly, is not allowing to add new Kanji - so let\u0026rsquo;s look at what\u0026rsquo;s done in reality.\n1: Background, The Japanese writing system Today, Kana-to-Kanji conversion is most widely used for Japanese input. As hardware, normal keybards with US-American or European key layouts can be used.\nThe Japanese writing system is composed of 3 scripts:\nHiragana: consists of 50 syllables. Natively, the look like for example ま, み, む, め, も. We can try to express the sound with Latin characters: ma, mi, mu, me, mo. When someone masters the hiragana, then already all Japanese words/sentences can be expressed. Katakana: consists also of 50 syllables, all of them have a counterpart in the Hiragana alphabet which sounds similar. Katakana look like this: マ, ミ, ム, メ, モ. Also these 5 syllables are spelled ma, mi, mu, me, mo. Katakana are often used to express spelling of names, or words which initially came from other languages. Kanji: These are the most complex pieces, comparable to words in English or other languages. Kanji are composed from one or more glyphs. 車 (spelled: kuruma) is a Kanji with a single glyph, it means \u0026lsquo;car\u0026rsquo; 救急車 (spelled: kyukyusha) is a Kanji with 3 glyphs, meaning \u0026lsquo;ambulance\u0026rsquo; If Hiragana can already express everything, why bother with the other 2 systems? We would just use a managable amount of 50 different characters and could be done. That would be a bit like using ENGLISH IN UPPERCASE ONLY. BASICALLY WORKS, BUT IS HARD TO READ. Also, Japanese has many homophones, so words which sound quite similar but have different meanings. So one word written in Hiragana can mean different things, but writing it in Kanji makes reading and understanding easier. Besides, using only Hiragana would just be lame - Japanese is full of beauty and culture, and one needs all 3 alphabets to experience that.\nTypical Japanese sentences are written with Kanji, along with Hiragana, and also Katakana.\n2: Fonts First precondition: the system needs the right fonts to display the characters we want to input. Nowadays, mostly unicode is used to refer to the characters which should be displayed. The system needs a font which includes these unicode characters. If all characters in this article here are properly displayed and do not end up as a \u0026lsquo;white box\u0026rsquo;, that is a good sign. Also visiting the wikipedia page of the target language is a quick test to verify if fonts are missing: Wikipedia Japan.\nEven if here everything looks ok, we should be prepared to hit Kanji or other characters in the future for which we have no font installed. With the Latin alphabet, we have 26 characters and can quickly verify if at least one font covers these. For Japanese, we are dealing with the core of the ~2000 most used Kanji which are more likely to be covered, and then many more.\n3: Required software Hardware wise, a normal keyboard with for example US-American layout can be used.\nLet\u0026rsquo;s look at the software. We will need to deal with multiple input modes, in most cases one mode for inputting the Latin alphabet and one mode for Japanese.\nWe need to use 2 kinds of new software:\none to allow Japanese input, so to interpret our key strokes, translate them into Japanese and send that to the underlying application.\nExamples on Linux: Mozc, KKS or Anthy. and one application for switching between the input modes. At least, one wants to be able to switch between English and Japanese input methods. On Linux, ibus or fcitx are typical input switchers. On Mac OS X or windows, the software comes with the operating system. On Linux, the packages can be installed with the whole distro when selecting Japanese as language, or afterwards. On Fedora Linux, I install package fcitx5-mozc - this is the Japanese input software, and the installation also triggers the installation of the input switcher.\n4: Actual Keyboard Input When the window manager starts, I open a terminal or a graphical application like Firefox - and with my settings the Japanese input mode is active by default. A small icon indicates the currently active mode.\nWhen pressing a key combination like \u0026ldquo;shift\u0026rdquo; + \u0026ldquo;space\u0026rdquo;, the input switcher software toggles between the input methods. Initially, the hotkey needs to be configured, for me in starting the fcitx5-config-qt utility.\nSetting up further modes for example for German and Russian is easy.\nWith the Japanese mode active, let\u0026rsquo;s input a sentence. As per my configuration, I can now press Latin characters, and get them converted to Hiragana: pressing the 2 keys \u0026lsquo;ma\u0026rsquo; leads to ま. Just keeping typing, we get full Hiragana sentences:\nきょうはあたたかいですね。 Pressing the same keys while in English input mode, we get this:\nkyouhaatatakaidesune. To get Katakana and Kanji, we can input some Hiragana, and then press \u0026lsquo;space\u0026rsquo;. With this, the input engine looks at a dictionary, and takes guesses what we might mean. Japanese has many homophones, so a combination of Hiragana could potentially mean various different Kanji - because of this we are presented with a graphical menu and can select the Kanji we intend to write here.\nきょう is our first Kanji, we invocate completion and get 今日, a Kanji composed from 2 glyphs. The first part, 今, means \u0026rsquo;now\u0026rsquo;, and the second part \u0026lsquo;day\u0026rsquo;, together they mean \u0026rsquo;today\u0026rsquo;. Following this, we can input the full sentence:\n今日は暖かいですね。- It\u0026#39;s warm today, isn\u0026#39;t it? So there we have it: we basically use keys labeled with the Latin characters, to express in Hiragana the Japanese words we want to write down, and the input methods is then offering us Japanese words which sound like what we want to type, we select what we are looking for and have written a sentence.\n5: More language details If this is your first contact with Kanji, some more details. From the above you might think that like for Latin letters, as soon as we see a glyph like 今, we know the meaning and how its spelled - but that does not apply. For example, this glyph by itself is spoken \u0026lsquo;ima\u0026rsquo; and means now. The Kanji 今日 contains the glyph, but is spelled \u0026lsquo;kyou\u0026rsquo;, so no traces of the spelling before. The meaning of 今日 is \u0026rsquo;today\u0026rsquo;.\nThe Kanji 今年 again has the glyph, is spelled \u0026lsquo;kotoshi\u0026rsquo;, and means \u0026rsquo;this year\u0026rsquo;. So the spelling of the very same glyph has changed again. Almost all glyphs have more than one spelling. When reading, one needs to look at the whole compound of glyphs, and then judge on the correct spelling.\nOne of the reasons for this is that Japanese Kanji have their roots in China: at one point in the past they were used in China for writing, and Japanese was only spoken. Kanji were then introduced in Japan, and depending on the circumstances, Chinese or Japanese readings are used for Kanji.\nReading English or German words is not that different, though. You probably have seen children reading texts in deciphering letter by letter, spelling them slowly and then concluding on the whole word. When we master reading, we also look at the whole word and spell it - just spelling the letters one by one is not counting as \u0026lsquo;fluent reading\u0026rsquo;.\n","date":"2023-02-23T13:04:17+01:00","image":"/posts/2023/02/japanese-keyboard-input/languages2b.png","permalink":"/posts/2023/02/japanese-keyboard-input/","title":"How does Japanese Keyboard Input work?"},{"content":"Some weeks ago, architect Florian Busch presented in German on architecture in Tokyo. Picking up on his details, I attended an event where we visited one of the buildings he designed, and learend some things.\nBuilding in Tokyo Tokyo was built on a river delta, with this it got right from the start many rivers. Over time the rivers got removed, and constructs started to prevent floods. Still as of today, most buildings in Tokyo are located below the rivers. Especially in August Tokyo often sees Typhoons, along with these the rivers rise and sometimes flood houses.\nBuildings here were initially made mostly from wood - fires then swept away big parts of the city. Brick buildings were intended to solve this, but these did not withstand the frequent earth quakes.\nThere are\nbuildings for living, mostly built for 25 to 35 years of usage commercial buildings, these are higher and designed for longer use As small buildings are not used for long, there is also tendency to build cheaper: investing in dual layer windows, or better insulation would pay off more with longer time of use. Business buildings are more likely to have these features. All buildings are now designed with earth quakes in mind.\nDid you notice this tendency in Japanese buildings to have these angles, instead of being shaped like for example a cube? This is to the most part rooted in laws on how much shadow a building is allowed to impose to the next building: the shape ensures the required amount of light:\nThe visited building Many houses for living are here \u0026lsquo;off the shelf\u0026rsquo;, but we were visiting a unique building which is about to get finished, designed by Florian Busch. Why would one get an architect instead of buying \u0026lsquo;off the shelf\u0026rsquo;? It starts with the ground: the place is not shaped like a quare, but more a rectangle. The owner of the ground wants the best possible usage, so one has to decide the number of floors. The city design is limiting overall building hight.\nA building always needs stairs, and with more than 5 floors also an elevator. The ground was not big, so to prevent the need of an elevator they settled with 5 floors. The floors wind up \u0026lsquo;around\u0026rsquo; the building, just like spiral stairs, or in German \u0026lsquo;Wendeltreppe\u0026rsquo;. As a group of ~12 people, we visited all floors and had a look. This building will house a restaurant on one floor, another floor for preparing the meals, and the other floors for living.\nIt takes ~3 years in Tokyo to get this custom designed building done, that would be longer in Germany. The architect is from the first design phases on constantly in contact with various government agencies, discussing what is allowed to be done. In Germany, the architect is more or less setting up the whole plan, then handing it over - and either getting agreement or has to design parts again.\nWhen a building is constructed, plates have to inform the public about all kinds of details. This is the plate in front of the visited building:\nFlorian Busch deciphered the plate of a different construction site for us, very interesting. Who builds which kind of building, which schedule? Which material is used? When the old building was taken down, was there something remarkable used like asbestos? That can give the owners of surrounding buildings hints their houses might have it, too.\nHow to restructure streets? When you are a muncipality like Tokyo and want to restructure a street, let\u0026rsquo;s say it\u0026rsquo;s now 2 lanes and you want it for lanes - for example as it was designated as major street for fleeing at times of deasasters. This is happening here, picture on the left side:\nAs a normal visitor, you notice much unused space along the street. The city has decided to enlarge the street, but can not enforce anybody to sell property. So after setting up the plan, whenever a building is teared down and a new one is to be built, the new plan is honoured: by new planning the area for extension can not be used for houses again. That way, it might take dozens of years, but eventually the space for extension is free. The house on the right side is one of the few ones left: one could argue they stay there hoping the city offers money eventually to move out. Interestingly, it\u0026rsquo;s a real estate shop.\nRestructuring small streets Along major streets, high buildings are allowed - just stepping into a small street nearby we are surrounded by small living houses. Extending the width of the street works here exactly the same: on the following picture one house was built newly, with the enlarged street, and the other one stays as it is - for now.\nChances are that we get guided on a second trip in spring, if interested you can keep an eye on link or ask me for my script to fetch these events as rss-feed.\n","date":"2023-01-23T12:16:52+01:00","image":"/posts/2023/01/tokyo-and-architecture/20230126_013203917_buildingsb.jpg","permalink":"/posts/2023/01/tokyo-and-architecture/","title":"Learnings on Tokyo and Architecture"},{"content":"I\u0026rsquo;m back in Germany for some weeks, and have revived the Commodore Amiga 500. It still works! The gaming experience is different from what we have now, and that is not about better graphics or such. Got me thinking.. why is that?\nThe Technology of the Commodore era I\u0026rsquo;m focusing here on the 2 systems from the Commodore era: the Commodore 64 (\u0026ldquo;C64\u0026rdquo;, released 1982) and the Amiga 500 (released 1987).\nThese systems were capable of displaying colors, and produced (considering their time) really good sound. With joysticks attached, they were often used for games. To get an understanding of their computing power: that was so low, that the chips did not need coolers. Typical available memory: 64kb (C64), 1MB (Amiga). No dedicated chips for 3D, but the Amiga hat blitter chips for making some 2D graphic operations more efficient. Up to 2 joysticks could be attached.\nAt that time, data was read from audio tapes (C64) or floppy disks (C64, Amiga). Software/programs were often so small that one could sit down and directly input the code which was printed in a magazine.\nMy social environment of that era At that time, I was in school. At the beginning of the Commodore era, other pupils had C64, were playing point-and-click adventure games like Zak McKracken. People played at home, talked about the game at school and got new input on what to try out (\u0026ldquo;you need to use the radioactive water with the plant!\u0026rdquo;).\nI got an Amiga, and besides the time of using it alone, there was the experience I want to explore here: 2 or 3 people sitting together in front of an Amiga (or C64) and having fun. There were basically 2 categories of games:\nmultiple players would act at the same time, for example attacking an opponent together, or each one steering a small car on the screen or competing with each other, at the same time or in turns, for example with Golfing games Today, \u0026lsquo;playing together\u0026rsquo; means that everybody is sitting at home in front of their computer, interconnected via the internet, and playing. In the Commodore era, we were sitting together in front of a single computer, playing, drinking beer or coke, talking. C64 and Amiga had no hard disks: data was mostly loaded from floppy disks, this took time. So we decided on a game to play, inserted the disk, it started loading. The programmers also were aware of loading times of their architecture, they provided music or animations already while the actual game was getting loaded - all of this resulting in this \u0026lsquo;Commodore era experience\u0026rsquo;.\nThere is a further piece: we were all in school, most had gotten the C64 or Amiga for present, we had no money to spend on games. Seeing a new game at a friends house, making a copy, taking it home, that was part of the experience. I think nowadays the block buster games are still expensive, but quite some people have decided to get only slightly older titles, for a fraction of the price. Games also got easy to obtain online via steam, while in the Commodore era you would have to buy in computer stores or order via phone and get the game shipped.\nThe described \u0026rsquo;experience\u0026rsquo; is not limited to Commodore systems: I also had great \u0026lsquo;many people in front of one system\u0026rsquo; experiences with Dynablaster and \u0026lsquo;Scorched earth\u0026rsquo;, both x86/MS-DOS. These did not have the \u0026lsquo;we need for the game to load\u0026rsquo; part, and PC\u0026rsquo;s at that time had no sound card, taking down the experience a bit.\nWe might also have to consider graphics and gameplay: graphics were important also at that time, although the standards and capabilities were different from today. Gameplay though is mostly not connected to graphics, and many games of that era are also today fun to play, and immersive.\nWhat other gaming is there? To better understand what was special about that times gaming experience, I would like to compare with 3 other related experiences.\nFirst is current board games: so things like physical Monopoly games. They have physical items like cards or game money, one gathers friends together in real life around a table and plays them. Like with the \u0026lsquo;Commodore era experience\u0026rsquo;, people meet in real world for this. Just switching to a different game is not as easy.\nThen physical tech: I remember my grandpa on occasions like Christmas or Easter setting up a special machine on the table: a small steam machine. It had a water and a fuel tank, after turning on a wheel started spinning and would drive 2 or 3 other small appliances like a hammer, a small version of what a blacksmith is using. Grandpa was operating it, but myself and others were fascinated by this: what\u0026rsquo;s it about that experience? Fascination of seeing technology in action, getting the machine to run, together?\nAlso, LAN parties: a few years later, playing over the internet was still expensive and with high latency. We gathered with sometimes hundreds of people in big halls, bringing in our computer for a weekend full of playing and talking.\nPeople are attracted by one or multiple of these factors also nowadays: the computers of that time were easier to understand than nowadays systems, making it easier to program - but also more challenging as the systems were slow by nowadays standards. At demo parties like Evoke, people challenge each other to create art on such systems.\nReviving the tech I looked into reviving the tech this Christmas: the Amiga 500 starts up. My Samsung TV still has a SCART connector, the Amiga video signal has a decent quality, games run as they did at that time. I got a serial connection to a Linux system setup, so can send disk images in adf file format over that and write them to floppies. x86 systems can not natively write Amiga disks.\nThere are also software emulators like UAE (Unix Amiga Emulator), also emulators for C64 exist. Also FPGA\u0026rsquo;s like MiSTer exist, they are chips which get wired like for example Amiga CPUs.\nSome tech drawbacks:\nMy external floppy drive no longer works, I debugged some time, to no avail My power supply emits a high frequency sound. Opening up, I see no obvious issues like inflated components. Visuals: in the Commodore era, CRTs were used as monitors, pixels were not as sharp as they are today but to some degree blurry. Floppy disks are the medium of that era - and they are no longer produced now. They are scarce.. But then, there is also a trend to augment with new tech: replacing the floppy with sd-cards, using extensions to directly get the video signal and convert it to HDMI, and so on.\nSummary Anybody else having experienced this as a special era of gaming? Has it gone on for you and evolved into something else, or has it ended? Should this be made available for others to experience?\nMastodon thread for comments\n","date":"2022-12-29T20:35:12+01:00","image":"/posts/2022/12/gaming-in-the-commodore-era/20221218_102942852_amigawide.jpg","permalink":"/posts/2022/12/gaming-in-the-commodore-era/","title":"Thoughts on Gaming in the Commodore Era"},{"content":"My classic software selection In 2016 I moved to Japan, and started to share experiences via multiple ways, all with their own strong points.\nJapan life observations is a static html document, rendered by asciidoc from a file in markup format. I find this useful for text, with a few images inbetween. The pictures are backed by 2 kinds of albums: fgallery: small and lean, but an abandoned project, also not dealing nicely with bigger numbers of images, and not providing a map and other features Lychee: having all kinds of features, but also relying on many single components. The wiki to easily modify content These services are updated quite frequently over the year, announced via Mastodon or via RSS feeds for these components.\n..and then there is the fotobook which I\u0026rsquo;m producing once per year. This has almost exclusively pictures. I have it physically printed for relatives who do not use the internet, and a pdf version is published for some friends.\nIn these past 6 years, this was my workflow:\nSorting the pictures of the past year: via Syncthing the pictures land in a directory on my Fedora system. ~30% get over the year sorted into a Lychee album and put online, the other 70% now get sorted, some deleted, post processed with Gimp and ImageMagick. A few stay at high resolution, the others are resized to 1920x1080 pixels. I use \u0026lsquo;qiv\u0026rsquo; for viewing and deleting images. Next, I fetched a proprietary software from a German company. At least native Linux software, which then allowed me to compose the pictures into a fotobook, typically with 32 pages. When finished, the software allows to directly upload the fotobook and have it printed, and then one by one send it out directly to the recipients. I also created a pdf version, but that Linux software is focused on uploading the data to the company - creation of a high quality version is not in scope. My workaround was to have all pages of the fotobook displayed one by one, take screenshots, and then compose these into a pdf file. The outcome is a 6-10MB pdf file of low quality. Refining the workflow The first step, going through the years pictures, has not changed this year. I need to select myself the resolution I want the picture in, for high quality for later printing/usage as screen background, or low resolution to minimize occupied space.\nBiggest pain point in the past was the low quality pdf, and the fact that updating the pdf after changed in the fotobook took much effort.\nMy new workflow is based on composing the fotobook with scribus, an Open Source software often used to layout flyers, documents which have text and images - and where one wants to be very flexible in deciding the layout, where text should flow around images, and so on.\nIf one has documents with more text, then LaTeX and co. get more interesting: they typeset following good common rules and allow the writer to focus on the contents. Just in my case, there is almost no text, and I want full control over the placement of the images.\nThe new workflow in detail Of course, starting to work with a new software like scribus takes a bit of time, but having a visual editor (like gimp, Inkscape etc.) makes things easier, I came up with a fotobook after some hours.\nExporting to pdf gave me nearly a heart attack: the exported pdf file with ~30 pages had 180MB in size. For printing at a company, the size would not matter, but that file would be to big for friends to \u0026lsquo;just download and have a look\u0026rsquo;.\nDots per inch (dpi) is an important unit. For printing, using material with 300dpi or more should lead to acceptable results. A print of a low resolution image on a big sheet of paper will look bad. When exporting to pdf in scribus, options to compress lossless are the default, resulting in big files.\nThe new workflow allows me to tune the output quality versus the file size, here you can access various versions:\nmax quality, 72MB: This is what I sent to the company for printing. 300dpi, 43MB: hard to distinguish from the 72MB version 200dpi, 25MB: quality down a bit, when zooming in 100dpi, 8MB: quality down even more - but still looks way better than what I got the last years from the screenshots. These are links, you can compare the versions directly. \u0026ldquo;pdfimages -list filename\u0026rdquo; shows that the maximum quality file has images with up to 900dpi, for the other versions it confirms that the images were indeed scaled down.\nI recommend that you start with asking a printing company to just print a single book, instead of directly ordering 10. That gives you a chance to correct issues, this is the first version I received:\nNow I modified the pdf and wait for delivery of the second print.\nSummary I love the flexibility I gained: my source file (scribus format) is now open, I can easily modify it, create a public version with faces of family members and friends cut out. The resulting pdf file has by default excellent quality (but is big), and with options I can reduce the quality and get smaller pdf files.\nSo creation of the image book took this year longer than last year, this is in part due to the new workflow and will improve a bit in the future.\nSomething I really like about my LaTeX and asciidoc workflows is that I have a plain textfile and then have the output rendered from that.\nBesides commandline \u0026lsquo;diff\u0026rsquo;, I really like the graphical tool \u0026lsquo;diffuse\u0026rsquo; for this, it\u0026rsquo;s packaged in Fedora. With my new workflow, the scribus file is the \u0026lsquo;one source\u0026rsquo;. To compare multiple scribus files, I need to open them in parallel and visually compare how they are rendered.. that\u0026rsquo;s not as nice.\n","date":"2022-12-20T18:44:46+01:00","image":"/posts/2022/12/imagebooks-the-open-way/20221224_scribus2d.jpg","permalink":"/posts/2022/12/imagebooks-the-open-way/","title":"Reworked image book workflow"},{"content":"What\u0026rsquo;s this about? This is my attempt of an overview of my current social networks/news fetching setup.\nEver explained a colleague or friend a technical problem, and along with giving the explanations understood the issue better, leading to a solution? Summarizing my current social network/news fetching setup forces me to rethink it, and can give ideas what to improve.\nAlso, the increasing numbers of Mastodon users show that also many people rethink their habits and tools, so this can also provide food for thought for others.\nRSS feeds are the base Initially I had websites interesting for me open in browser tabs, and would frequently reload these, and then if these were complex pages also have to scroll down and try to understand if something changed. That improved very much with RSS feeds: these are files provided by for example a website, representing the website in a machine readable way. When websites offer RSS feeds, I can frequently fetch their feed, and easily decide if something changed.\nI run Tiny Tiny RSS (ttrss) for that, an open source software which frequently fetches feeds and notifies me about changes. Many pages offer feeds, I also offer them for my wiki and other sites. My main interface to ttrss is my webbrowser, my first tab in Firefox accesses the ttrss instance I am running. I also access ttrss from Android mobile, for example when I\u0026rsquo;m on the train: apps are then accessing the ttrss instance, showing the feeds, and one can click the link to read the full news article on the website.\nFor example I read a feed with news from tagesschau.de, and see headlines of various news presented in my ttrss interface. With key combinations I can jump to the next news entry, and if something sounds interesting I can open that news on tagesschau.de directly in a new browser tab.\nEverything is a feed! Websites themself are in the best position to offer feeds of their contents, but many opt into not doing that, maybe they want to force people to watch their site directly so they can distribute more ads, or can collect more data. One of the best projects around this is rss-bridge. One can install that code on an own webserver, and then access URLs provided by rss-bridge as feeds. rss-bridge under the hood is then contacting the webserver, extracting data from there and offering it as feed. So with this, for example the dilbert webcomics become available as feed, and I can read them via ttrss.\nI do not consume all articles directly: in the morning I skim over new items presented by ttrss, open some in the browser and have a look. Many of these articles are then interesting, but so long that they need more time to read than I can invest right now - for these items, ttrss allows to mark the article with an asterisk. These items are then marked as \u0026lsquo;read\u0026rsquo; in ttrss, but appear in a section called \u0026lsquo;starred articles\u0026rsquo; where I consume them later.\nYes, really everything can become a feed! I\u0026rsquo;m still around on the chat network IrcNET, which uses the IRC protocol. An RSS feed is just a text file with some XML, so a small script I created is just frequently opening the logfile from these IRC channels, taking out posts which look like a URL, and then these are ending up in my ttrss instance.\nMastodon as social network Besides the browser tab with ttrss, I have a tab with Mastodon open.\nGoogle plus closed down, I left Facebook years ago, but Mastodon has stuck. Many of my interests are around open source, technology and language learning - we have nice communities around these on Mastodon.\nI also occasionally post on Twitter, and like to follow some people who are not on Mastodon: also for these, rss-bridge can easily make the Twitter posts of users available as feeds, and my ttrss offers these in the most convenient way. Things I learned from that: when I was watching my twitter timeline, I missed updates from people I \u0026lsquo;follow\u0026rsquo; via the Twitter-follow-function. Now, I catch all posts from these users.\nFor me, having Mastodon separated \u0026lsquo;feels right\u0026rsquo;, but in case you want to see posts of Twitter users in your Mastodon timeline, this would work:\nrss-bridge to get the tweets as feeds Then using this or other software to get the feeds into Mastodon. What about Podcasts? So for podcasts, I use no dedicated fetcher/app. I use ttrss to fetch the feeds for podcasts, get notified there. For example, when a new episode of the Bilingual news podcast (weekly talk about news, in Japanese and English), I mark the ttrss item with the \u0026lsquo;star\u0026rsquo;, and in a terminal I run script \u0026rsquo;latest_selfhosted.sh\u0026rsquo; which fetches the last episode, drops it in a directory for podcasts. After a few days soem podcasts gathered there, then I connect my mp3player via USB and move the files to the player. The player runs \u0026lsquo;rockbox\u0026rsquo; and can also set bookmarks. When I finish a podcast file, I remove the file, and later in the ttrss interface I remove the \u0026lsquo;star\u0026rsquo; for that entry - in many cases I also open links belonging to that podcasts feed item, like links related to the podcasts content.\nThings I love about the setup For me, the current state works very well. The feeds are openly standardized as rss/atom, and Tiny Tiny RSS as well as the browser and Android app I use for access are open source, with all it\u0026rsquo;s benefits. For communication, with Mastodon I\u0026rsquo;m now happy that there is no longer a single company able to enforce their own rules.\nI feel like Mastodon could really help to also make new connections: when we entered Corona, many from the Tokyo Linux user group (also other groups had the idea) tried to volunteer, to work with schools to have them try out open source software to lift lessons to the internet. The issue turned out to get connections to teachers and schools. In the end, I had 2 sessions with a school which turned out great (and did lead to this documentation), but really this should be done on bigger scale. In Mastodon I see instances appearing with focus on teaching, and it seems like getting such contacts might be easier. I also got into touch with old friends, after years.\nThings I would like to improve Just some recent thoughts on what could be improved:\nMastodon timeline, quite much content.\nFor ttrss, I can nicely pull new items as I have time. For Mastodon, I have a long newsfeed, and the more people I follow, the more new items get available. Consuming the timeline is harder.. but I also would not want that as feed in ttrss, that would somehow not fit for me. Separation by topics is an issue. I have currently a single Mastodon account, and my posts cover the wide range of open source, Japan, language learning and more. People might be interested in just one of these, but they get all.\nIn reverse, I also decide to follow (or not) others from the signal/noise ratio of what seems interesting. In the morning, when opening the browser tab with the Mastodon timeline, the timeline is scrolled up showing the most recent entry. Would be convenient to have it scrolled down to the last entry I did already see, so then I can directly read up the new entries. Attribution. So for Mastodon, all people I follow see that, but with rss-bridge I actually create feeds from the twitter users tweets-page, that is independent of me \u0026lsquo;following\u0026rsquo; them on twitter. If many people do this, it might result in actually very interesting content from twitter users not being reflected in their follower-number. Not feeling sorry there for the twitter algorythms, but the user themself.. The unsolved problem of money making: some podcasts I listen to start to ask for money based on proof-of-work blockchains, which I can not agree to - that\u0026rsquo;s just further accelerating the climate crisis. But having a good other mechanism to help content creators for podcasts etc. would be nice. ","date":"2022-11-06T03:05:46+01:00","image":"/posts/2022/11/social-network-and-news-fetching-as-of-2022/mastodon3.jpg","permalink":"/posts/2022/11/social-network-and-news-fetching-as-of-2022/","title":"State of Social Network and News Fetching as of 2022"},{"content":"TL\u0026rsquo;DNR Both smartplugs (reporting electrical consumption of a connected device) and the laptop battery allow consumption measurements. Results are in the same ballpark, but also differences of 20% can be seen.\nSmartplugs for Linux system power measurement So far, I only measured power Linux system power consumption with the help of the internal battery, details here and here. Now I got 3 smart plugs for Japanese power outlets which can measure power, and wanted to use these for more experiments. All 3 devices are setup in installing an Android app on the mobile phone, activating bluetooth, and then the smartplugs hook into the local network. While you have the smartplugs directly next to you, all of these vendors intend you to use their Android apps to turn the smartplug on/off, or read the current measured power consumption of connected devices.\nFor my measurements, I want that power consumption as plain numbers, for scripting ang graphing. It was hard enough to find smartplugs for Japanese power outlets which also can measure consumption, and after playing a bit with these only one allows reading of the consumption via API. For this, I would ideally talk with the smartplug directly via local wifi, but only get that the API from the vendors cloud on the internet. So with that, a script can now query the consumption.\nDetails of the 3 devices are on the wiki. The one on the left, Switchbot プラグミニ スマートプラグ, has the consumption available via API. The Shelly smartplug I use in Germany offers it directly on the device, and additionally also has a temperature sensor.\nThere is quite some delay between changes in power consumption and seing the change reflected via API, 3-4 seconds.\nWhy? What could be interesting to look at? Previously I had tried to find out how much energy single components in a laptop consume: NIC, display, cpu, NVME-disk and so on. For that, I had looked at how quickly the battery discharges and calculated the consumption of the components.\nBut how well do the measurements of the battery and the smartplug match up?\nMeasurement setup Consumption measurement via battery is only available when the system runs on battery, and measurement of the smartplug only when the system is on the power grid. So I need to create load scenarios on the system, which cause consumption, and then apply the scenarios to both ways of measurement, after each other.\nTo really get similar scenarios, things needed to be scripted. I use one system to control (\u0026lsquo;controller\u0026rsquo;), and a second system undergoing measurement (SUT, System Under Test). An Ansible playbook is preparing the SUT:\ninstalling powertop (to easily optimize for low power consumption), upower (to verify battery state), and Performance Co-Pilot (to access metrics like power consumption, and archive them for later examination) Activating pmda-denki and pmda-sensors, these provide the metrics for consumption/temperature) enable remote access for PCP\u0026rsquo;s pmcd. The daemon runs on the SUT, and offers the metrics. The controller system is querying the metrics from there and archiving them for examination. Then, scripts can be run which for example run \u0026lsquo;openssl speed\u0026rsquo; on the SUT, and observe the consumption and other metrics. When the test is done for both battery and smartplug measurement, the resulting data gets dumped from PCP\u0026rsquo;s archive files, and fed into gnuplot.\nSome results Looking at Thinkpad T460s with Fedora and comparing what smartplug/battery report:\nSystem state report smartplug report battery idle, screen off 1.6W 1.7W idle, screen on 5.6W 4.3W cpu loaded 27W 22W I wondered if maybe the bios is tuning down the cpus when on battery, although I had disabled that in the bios. The RAPL readings help to understand that aspect: they report consumption estimations for CPU, memory and gpu.\nThe green and violet curves are the RAPL metrics, so these report here for both battery/smartplug operation comparable consumption, even a bit higher for battery.\nThe blue curve is the consumption reported by the smartplug, and the yellow line is the consumption as reported via the battery, in many cases as /sys/class/power_supply/BAT0/power_now. This seems to be computed from observing the battery level.\nThe orange line is calculated within PCP/pmda-denki: in one minute timeframes it looks at the current battery charge level and does an own computation of the current power drain, this is meant as a cross check against power_now/to average over longer time instead of jumping up and down like power_now.\nIdle system Looking at the idle system, green/purple are again the RAPL metrics. Ontop of that we have consumers like nvme-disk, NIC and so on. The yellow and orange computations are battery based and quite constant, just the smartplug has here interesting spikes.\nFuture Might be interesting to look at the impact of kernel mitigations for the spectre/meltdown family of issues on consumption. Doing a load which allows also to measure performance would be interesting, to compute power efficiency. Also various tuning profiled (tuned) should have influence.\nI should also do that for an Apple silicon system running Linux.. the one I played with in the past is now used by my parents.\nAlready now with consumption measurement, one could output how much co2 was exhausted if the energy was created in burning coal, and so on.\nIdeas what would be interesting to measure? Please let me know.\n","date":"2022-11-03T10:11:10+01:00","image":"/posts/2022/11/powermeasurement-with-battery-and-smartplug/20220803_043840650_jp_smartplugs.jpg","permalink":"/posts/2022/11/powermeasurement-with-battery-and-smartplug/","title":"Power measurement: Laptop Battery vs. Smartplug"},{"content":"For me, Japan deals with culture, technology and religion in different ways from what I was used to. I will try to grab these specialities here.\nCombining Culture \u0026amp; Technology I started to wonder about this when I came to Matsuyama, and saw the Dogo onsen. There it was, with more than 1000 years of history one of the oldest bathes in Japan. Just in the middle of new buildings, a nice combination of traditional old and new things.\nAnother example if impressive combination of culture and technology is this robot. Technology of robots which are fore example used in car production has advances, so comparing the capabilities, power and precision of such a robot with something like traditional sword usage is just great.\nSpinning Religion further Making fun might be a bad phrase.. thinking about religion outside the box might be better. I had good discussions with religious people believing in Christianity, but shining Christian religious figures in a different light is forbidden by all I know.\nThings seem more flexible in Japan. That starts with most Japanese taking parts of multiple religions, Buddhism and Shinto. While Christianity has this exclusivity rule \u0026lsquo;you must not believe in other gods\u0026rsquo;, Japanese take depending on the occasion from both religions: marriage from Shinto, funeral from Buddhism. Also religious places from both are frequently visited, sometimes these are located next to each other.\nHaving own thoughts about religious figures in Christianity is very limited, but more flexible in Japan. These are guardians at the Yakuō-in Sanmon Gate (高尾山 薬王院 山門) near Tokyo, in front of a buddhist temple:\nNow, how do you think these look like when they are now on duty, guarding the entrance? An artist wondered to, and made then great pictures of \u0026ldquo;guardians on their day off\u0026rdquo;. I just love how people can go ahead and do this kind of thing with religious figures!\nFlexibility with eating Food is modified, extended everywhere, constantly. But here I appreciate the way of enjoying food: eating nagashi somen. In Japan, we see more emphasis on the group, in the western world more on the individual. Imagine a pipe of 10cm diameter, with the upper half cut away, so looking like a drain for rain water. At fairs and festivals, such a pipe is setup with a slope, and drinkable water flows down. In that water, thin noodles made of wheat flour are inserted, the Somen. Ice pieces are also inserted, so the Somen moves downwards with the water. The participants then take place at the sides of the side, can take out Somen with chopsticks, dip into a cup with soy sauce they have in their other hand, and then eat the Somen. What a fantastic idea to eat together in this way!\nNext: ramen, a kind of soup, is famous in Japan, so one would think that exotic experiments are forbidden? Nope, there is even chocolate ramen..\nTechnology and Religion I work at a company involved in software development. One of my first encounters of how Japanese bring technology and religion together was when we were about to release a piece of software. One of our partners from Japan flew to America to discuss technical topics, and they brought along a kind of religious talisman, which was from a shrine, for giving the software a good start. The talisman was placed in the room for the whole week, while the talks were going on.\nIt does not end there: the Kanda Myojin shrine is specialized and has talisman (\u0026lsquo;omamori\u0026rsquo;) available for technology pieces like computers.\nDo you know Daruma?\nThese are a kind of doll in Japan, mostly red, looking like a head.\nThey are available with white eyes, when having a wish then the owner paints one eye, and when the wish is full filled the Daruma gets his second eye. I really loved to see Jaxa (kind of Japanese NASA) landing the Hayabusa spacecraft on an asteroid - do you notice the Daruma on the right side?\nJapan - a fine and flexible combination of tech, culture and religion.\n","date":"2022-07-28T14:29:49+02:00","image":"/posts/2022/07/japan-dealing-with-tech-culture-religion/20170715_141937_nagashisoumen.jpg","permalink":"/posts/2022/07/japan-dealing-with-tech-culture-religion/","title":"Japans flexibility around Culture, Technology and Religion"},{"content":"Off for a daytrip to the tea garden Takanashi (高梨茶園) in the Tanzawa area. After 60 min train ride from Tokyo I arrive at Hadano station, where I had been before when going to the Tou-no-dake mountains. After 15 min by bus and a walk, there is the tea garden. The trip is organized by oag, we are 23 people.\nThe first thing we see: a flag saying 新茶, so \u0026lsquo;fresh tea\u0026rsquo;. Family Takanashi is introduced: a man in his 60ies, his son in his 30ies with his wife, and the grandson of maybe one year. We taste a cold, green tea, are divided into 2 groups. Takanashi family has relations to Germany: they export tea, and presented at German exhitions. We sit down in front of tea plants and listen to the senior Mr. Takanashi.\nTea cultivation details What\u0026rsquo;s better than listening to someone who is really into the topic? Not much.. so in his childhood, tabaco was planted here instead of tea. That had many downsides: already the children got to breath tabako, the cultivation was difficult, and extra houses were needed for trying tabaco. So his father decided to cultivate green tea. Besides Shizuoka and Kagoshima, Tanzawa is known for tea. Once planted, the tea plants \u0026rsquo;live\u0026rsquo; 50 or 60 years! We visited beginning of June and had just missed the first harvest, so when leaves were cut from the plants and fabricated into tea. There is also a second harvest in a few weeks.\nUrban legends get debunked: I had learned that the green tea in pet bottles, sold in Japans convenience stores and drink machines, is from China - not true. While he prefers fresh tea, Mr. Takanashi drinks also from pet bottles if there is nothing else. He also welcomes sweets with green tea, and other innovative products.\nThe tea we tasted was brewed with 70-80°C, and was ready 1min40seconds after filling the tea into the water. That\u0026rsquo;s actually pretty much how I prepare my tea in the morning: I fill green tea into a small tea bag, give that into hot water well below boiling temperature, and drink that after ~2 minutes.\nThe tea factory After that, we move into the factory building. Of course, one use house slippers for a tea factory. :) Surprisingly, all of the machines here are made in Japan. After harvest, tea gets cleaned, then steamed, and then rubbed, optionally roasted. So theoretically, what would we get from just picking the leaves and giving them into hot water? No taste! The taste just appeard when \u0026lsquo;rubbing\u0026rsquo; the leaves, either with hands or with machines.\nAs one might imagine: the factory halls are full of green tea smell. When drinking green tea, most people feel the activating effect of the ingreedints, and we feel that already from just breathing the air here! So the workers here also still react to that, they did not somehow get used to it, it\u0026rsquo;s also for them stimulating. They can sleep non the less, after a full day of work. As the tea has to be processed as quickly as possible, they just sleep 3-4 hours in the harvest time.\nEventually we step out of the building, drink tea again, have lunch. We buy tea. The young kid is playing with freshly roased tea.. and my nerd brain remembers the \u0026lsquo;Dune\u0026rsquo; movies, with the planet Arrakis and the \u0026lsquo;spice\u0026rsquo; being everywhere, people being in contact with it so it event changes the colour of their eyes. I guess that won\u0026rsquo;t happen for green tea. ^^\nEver noticed these \u0026lsquo;fans\u0026rsquo; above tea plants? Turns out these are indeed to mix the air, to keep frost away from the plants - hard to imagine, with the fan located so high above the plants.\nWe then walked up to a well nearby, and drove back to Tokyo.\n","date":"2022-06-08T14:01:57+02:00","image":"/posts/2022/06/tea-farm-trip/20220605_044848426_teafields.jpg","permalink":"/posts/2022/06/tea-farm-trip/","title":"Visiting a tea farm in Japan"},{"content":"My past views on Apple hardware In 2007, some time after Apple had rebased MacOS on unix, I bought a Macbook and had a look. MacOS was not convenient for me, it did not even offer virtual screens at that time - I had been using these with Windowmaker on Linux since years. I tried native Linux on that Intel-CPU based hardware, it was not bad, but I preferred IBM Thinkpads for better driver availability, price and because of the red Trackpoint/nipple. My mum got the Macbook, I still felt much better with that than recommending her some Windows laptop. She stayed with Macbooks, before going to Japan I got her a Macbook in 2015, which she is still using.\nSo now with Apple silicon available, and Linux starting to get usableon the hardware, it was a good time to look again. I got a Macbook Air with 8GB RAM, 8 cores, 512GB SSD at Bic camera in Tokyo.\nA look at the hardware and MacOS X Wow, the Macbook Air is really thin. The hardware and MacOS X are tightly integrated, and convenient: after turning on, a personalization wizard starts. One presses a key with integrated fingerprint sensor multiple times, and can later use that for authentication. Setting English as system language and optional Japanese language input was easy. One leaves some voice samples to train Siri, which can then can answer simple questions like \u0026ldquo;What\u0026rsquo;s the weather\u0026rdquo; or \u0026ldquo;What\u0026rsquo;s the population of Germany?\u0026rdquo;. Asking like that is actually quicker than looking it up with a search engine - but I would not want to use such tech with proprietary tech, just with OpenSource backends. Siri can also not cope when multiple languages are used.\nI type Japanese characters on Linux since many years, in concert with English and German. On Linux, we typically use key combinations like \u0026ldquo;CTRL\u0026rdquo; + \u0026ldquo;space\u0026rdquo; to switch between Jap/Eng/Ger input modes, that\u0026rsquo;s really nicely done on the Macbook Airs keyboard: a dedicated key right of the spacebar activates Japanese, a dedicated key left of the spacebar jumps back to English. Besides that, I had hard times using the keyboard: the CTRL-key is much higher than on other keyboards, also other keys are off.\nThe MacOS X desktop features are quite nice: on Linux I\u0026rsquo;m a fan of dark themes and the dark mode in browsers. On MacOS X, I was surprised to see the whole desktop switch to a dark mode in the evening, very nice!\nAs temperatures rise here in Tokyo, the Thinkpad T590 I use for most things is despite all kinds of optimizations often spinning up the fan - the Macbook is fantastic in getting warm but not being noisy at all. Really impressive.\nTrying MacOS X software I\u0026rsquo;m involved in the Performance Co-Pilot (PCP) project, a suite for performance measuring. PCP was usable on older MacOS X, so I gave it a try on this box with MacOS X 12. Fetching and compiling required some other pieces like \u0026lsquo;git\u0026rsquo;, I got these from the brew project, that has come a long way since the last time I used it. My second try would have been pkgsrc which I know from NetBSD.\nI brought changes upstream to get PCP better into shape, but did not get it to compile on current MacOS X. Instead of Apples LLVM, one has to get GCC from brew, bison and so on.. here are details on where I got stuck.\nNext on my list was Nebula: the way I use it is to have a central system, and have clients like this Macbook connect to that central system, and they then have access to an IP space, i.e. a 10.0.0.0/24 network. The clients can also be behind a NAT to establish the connection. The clients can then \u0026lsquo;see\u0026rsquo; each other, so this is a nice idea for logging into systems to provide remote support. Nebula from MacOS X worked nicely for me, I just did not get automatic starting after reboots setup.\nLinux on Apple Silicon The big thing was of course to run Linux: Asahi Linux has an installer which makes it easy to setup dualboot with MaxOS. Installing Asahilinux went nicely, it\u0026rsquo;s arch based, comes with KDE. I then tried to install Fedora as a third system: that did not work out, the scripts are not prepared for a 3rd system. On Apple Silicon, no simple UEFI partitions are used, but more complex containers. The framework is set and implemented by Apple, and Hector Martin has done a great description of that, and suggested how Linux and other OS\u0026rsquo;s can utilize the system to be bootable.\nSo Fedora 35 runs nicely: plain Fedora userland, just with not the upstream kernel but the one from the Asahi Linux project. Last year I contributed to code in PCP to measure electrical power consumption, now I updated that code for Linux on Apple silicon. The Linux kernel is already reporting the current battery charge level and current consumption, with the PCP updates these metrics can now be recorded and visualized. Upcoming PCP 6.0 will be the first release with the changes.\nI did some measurements, and even though some details of cpu control are missing, the power consumption of Linux on the Macbook is in the same area as with the Lenovo Thinkpads.\nI use the i3 windowmanager on the Macbook, sway which I normally use is quite similiar. Virtualization also already works: I installed Fedora36 with the plain Fedora kernel as guest. Just as for speed, the guest speed is usable for trying things out, but operations like installation or \u0026lsquo;dnf update\u0026rsquo; take their time: I suspect that is related to our currently imperfect dealing with the M1 cpu. Seems like the host is 2-70 times as fast as the guest, depending on the exact operation. When multiple cores can be assigned to the guest, that can help with the difference of course.\nFurther notes are on the wiki.\nLinux@Apple Silicon: what works/works not? The detailed list is here at the Asahilinux site, but speaking broadly: Linux with windowmanager is nicely usable, browsers. Sound via internal speaker is disabled, microphone jack works. Wireless works but I had hickups every now and then. The GPU is not yet supported, but the M1 does make up for that with pure cpu power, so that window managers work smoothly. Charging also works on Linux, ethernet too - the Macbook Air just has 2 USB-C connectors, so I would need an adapter for that. So, as daily driver all things are in order and reasonably stable.\nI had wondered about getting a Mac mini with M1 chip instead of the Macbook Air, that would also provide 10GB ethernet.\nSummary This trip to Macbooks went nicer than my last one: with ARM base, the hardware feels much better than Intel based hardware. The keyboard is awkward to me, but with my external System 76 Launch keyboard that would not be a problem. MacOS X got more usable and some nice features - I just miss my flexibility from Linux, and the ability to dive into the code by myself if there are issues, which one has with OpenSource systems.\nThe ARM hardware can really have a future, working without fan noise is really nice. Pine powerbook and other hardware is also available.\nI will play a bit more with the Macbook in the next weeks, and in summer in Germany hand it over to my mum.\n","date":"2022-05-15T09:28:23+02:00","image":"/posts/2022/05/apple-silicon/20220515_071319640_mac_htopc.jpg","permalink":"/posts/2022/05/apple-silicon/","title":"A fresh look at Apple Silicon/Linux/MacOS X"},{"content":"The human factor One night I woke up from shaking. It\u0026rsquo;s immediately crystal clear: an earthquake. The big question, as always: will this be the big one? How does one feel? Frightened and helpless: one can not do anything against it. It\u0026rsquo;s shaking, and you can\u0026rsquo;t do anything.\nHow bad is that fear? I think I did not hear of this as a phobia so far, but I would not be surprised if people move for example to Hokkaido, the north part of Japan, as there are fewer earthquakes.\nI live in Tokyo. There is a cycle of roughly 30 years between major earthquakes, and the next one is due. Everybody knows. Schools do drills on what to do in case of earthquakes, all households get papers with instructions where to go to in case of a catastrophy. People are asked to keep water bottles around, food, a backpack with some things so the flat can be left in case of an emergency.\nWhat\u0026rsquo;s an earthquake, actually? Japan is on the crossing of tectonic plates. Moving towards each other, while one is moving below the other plate, it\u0026rsquo;s taking the other plate a bit with itself downwards - until the other plate accumulated so much pressure that it\u0026rsquo;s moving upwards suddenly, causing the quake.\nWe see notifications of the epicenter, sometimes these are under the see, sometimes on land.\nAre people prepared? Japanese live with it since \u0026lsquo;forever\u0026rsquo;. So from mindset people are prepared to some degree. Besides earthquakes we also have our share of super storms and floods. I fear that this level of emergencies is one of the reasons climate change seems to be taken not seriously so far: from birth one is used to emergencies and results like people loosing their home etc., making one think \u0026ldquo;it\u0026rsquo;s god given, can\u0026rsquo;t be changed\u0026rdquo;.\nJapanese have traditionally built houses with much wood, there were huge fires in cities, and firebombs in WW2 did lead to catastrophic results. Normal living houses in central and south Japan have thinner walls than houses in Europe, they are mostly built to withstand magnitude 8 earthquakes.\nSo all prepared? Not really.. the big East Japan earthquake 2011 resulted in a tsunami, and a meltdown in Fukushima. Also smaller quakes lead to many deadly accidents, broken streets, train tracks, houses and so on. Broken water pipes mean interrupted water supply, power supply stops due to interrupted lines and damaged power plants - last weeks earthquake with epicenter in the east of Fujishima was magnitude 4 in Tokyo and damaged coal power plants, so in the following week the \u0026ldquo;power usage counter\u0026rdquo; for Tokyo was over 100%.\nJapan is only producing 40% of the consumed food in the country, so support lines from other countries and in the country are very important.\nWhat is done, with technology? You can\u0026rsquo;t forecast earthquakes by hours or days, which would be ideal. In the country and on sea there are sensors, detecting first signs of quakes. The ripples of earthquakes move slower than electric energy, so that helps a bit with notifications. Algorithms have to understand if a quake is serious enough to result in notifications - and for which area. The most effective notification is via mobile phone cell broadcast, ETWS.\nI have a private phone always turned on, and a company provided phone. At a past earthquake, I got notifications on both of these - the screen lit up, there was a buzzing sound, and a voice said \u0026rsquo;earthquake\u0026rsquo; in Japanese. This was so loud, I even heard it from the flat next to mine.\nFor this notification to work, the mobile and the provider must support it. Most mobiles sold in Japan are sim locked, the provider has his hands deep in the phone, deployed his config, on a sim locked iphone from the company I received the alerts in the past. With the Pixel 4a, I did not receive alerts so far - I bought it unlocked.\nThese warnings are not restricted to mobiles, they also go to TV and other media. Besides earthquakes, also floods etc. can lead to notifications. The Jアラート (J-alert) backend can also notify about incoming threats like missiles.\nETWS, the mobile broadcast, is notifying quite wide, at least whole mobile cells. When we had the super typhoon Hagibis in 2019, whole areas received the ETWS warning on their mobile, and were then for details looking up on the internet whether their specific street/area was recommended to evacuate or not.\nSpecial experiences One time I was standing next to an open window. My rented room is on a small hill, one can look down on parts of the city. It was afternoon, and I started to hear a kind of rumbling sound down from the city - I guess that was from earh and everything ontop vibrating. At first, the source of that sound was not really clear, just after some seconds the vibrations reached me directly, this was an earthquake of 2 or 3 at my location.\nIn 2018 I had deployed my mobile phone at the company office, configured to take one picture per minute. Around 22:00 the last people left and the office went dark. Just that night there was an earthquake, and one can actually see one frame dislocated by one or two pixels. One also sees how lights in the flats around lit up.\nPictures and the timelapse are here.\nIn 2019, we had the supertyphoon Hagibis, pictures are here.\n","date":"2022-03-24T12:08:41+01:00","image":"/posts/2022/03/earthquakes-floods-japan/20191012_spaceview.jpg","permalink":"/posts/2022/03/earthquakes-floods-japan/","title":"Earthquakes, floods and Japan: human factor and technology"},{"content":"Here I will try to sum up steps and some backgrounds about running Linux on the Playstation4. The guides I read were sparse on background details, aimed at saying \u0026lsquo;click here and there, and then you have a Linux\u0026rsquo;. But for debugging one needs to understand more of the background. The guides also use Windows systems for the preparations, I try to provide the details so a reader can perform the steps on all systems. The install steps could actually be simpler when done from Linux, but I will stick to the procedure also used in other guides.\nWhy would one run Linux on the PS4? I\u0026rsquo;m oldschool in thinking that when I purchase hardware, I should have full control over it. When I got the Playstation 4 pro, I basically just played \u0026lsquo;Control\u0026rsquo;, and then it was collecting dust - until now, that I can run Linux. People have various motivations to run Linux on the PS4:\nTo use the PS4 their media client: to play videos, music, Netflix etc. As media server: the PS4 can provide services. One can make video/ music files available to the PS4 via usb3 or network, and then run Kodi, Jellyfin, Plex Media server etc. to make the media files available to be played via local network, or on the Ps4 itself For gaming: the PS4 has 8 cores, 8GB of RAM, and a GPU powerful enough for many 1080p games. As the cpu is x86 architecture, Linux games can be installed, for example via Steam. With that, also Windows games can in many cases easily be run, thanks to Proton. 3D acceleration works, with Vulkan. Because it makes just a useful desktop with enough power for internet surfing, simple development This is the PS4 running the fantastic WebGL wipeouthttps://phoboslab.org/wipeout/ code:\nTrying out Linux also comes with low risk: external storage is used, the internal disk not touched. As the internal storage is a rotating disk, using a fast external storage also gives you better performance.\nWhy NOT run Linux on the PS4? If you are looking for a simple media player, just get a Raspi4. It boots up without jailbreak, has enough power to play media. You should have some Linux experience, and be willing to spend time for debugging. The current solution requires you to jailbreak the PS4 again after each reboot. The jailbreak works with overflows, and is not working every time: you might end up trying for 10min until the jailbreak works. If you use the PS4/Linux as media center solution and your family is waiting for watching a movie, then this might be to long for you.\nThe Generic Steps \u0026amp; Preparations When the PS4 boots, the firmware starts only code signes by \u0026rsquo;trusted\u0026rsquo; parties. The code we want to run has not been signed, so are first running the normal system, and then using weaknesses to get our foot into the door, and run our code.\nSecurity holes which can be exploited like this existed for various firmware versions, when they got known newer firmware versions with fixes have been released. The recent jailbreak requires firmware level 9.0, if your firmware is lower you can upgrade to 9.0 and then use the jailbreak. If you run higher versions, you are out of luck for the moment.\nPreparations:\nI use a PS4 CUH7200B B02 1TB, a usb keyboard (System76 Launch) and usb mouse.\nyou need a small usb stick, a small image is written onto the stick, no partitions etc. are created. If your stick is at /dev/sdx:\n# dd if=exfathax.img of=/dev/sdx a further stick should be 16GB or bigger. It will be used for the Linux system. I strongly recommend here to use an SSD connected via usb3, or an NVMe-enclosure with an NVMe inside. Plain usb-sticks have quite bad performance for small writes/reads, and that is exactly what we do when installing, running and updating Linux. This media should be prepared with a DOS-style partition table, a single partition spanning the media, FAT32 formatted. The following command does this on Linux with gnu-parted, if the stick/SSD is available at /dev/sda:\n# parted -s /dev/sda mklabel msdos # parted -s /dev/sda mkpart primary 0% 100% # mkfs.vfat /dev/sda1 # mkdir -p /mnt/tmp # mount /dev/sda1 /mnt/tmp Then, we need to copy 3 files to /mnt/tmp:\nthe Linux kernel, filename bzImage the initrd, filename initramfs.cpio.gz an image with the Linux distro to be installed, file name psxitarch.tar.xz For a start, the \u0026lsquo;Fedora 32\u0026rsquo; distro image, a kernel and the initrd can be be obtained from the description section of this tutorial video. As for the kernel: these depend on the type of your hardware, you might have to try out various kernels to find one working. You can also compile your own kernel and boot that - but debugging that is challenging, as monitor output might not be working. The \u0026lsquo;5.3.7 Gentoo gc41cdf12d-dirty\u0026rsquo; works for me.\nFor now, with these 3 files on the second stick/SSD, the preparations are done.\nThe Detailed steps to boot Linux The generic workflow is like this:\nThe PS4 needs to be updated to firmware 9.0, and then be prevented from updating the firmware, in setting custom DNS servers which do not resolve the systems used for updating I recommend to watch this video, to set DNS servers as per the video, access the PS4 webbrowser as per video. Then, via the PS4 webbrowser, the URL from the video is accessed, code loaded and executed. As it executes, it asks to insert the usb-stick with the filesystem-exploit. If succeeding, you are asked to take out the stick, and get a message that the jailbreak code is now running. For me, the site for the manual exploit and GoldHenV2b2 works best. 2 things are happening: one is the actual exploit (utilizing a weakness to run own code), and then running own payload. One of the possible payloads is to start Linux kernels. Next, video mode preparations need to be done: the screen needs to be set to 1080p, no HDCP, full RGB color space. Then, you should connect a usb keyboard and mouse, your usb stick/SSD with the Linux distro, and run \u0026lsquo;Linux/1GB\u0026rsquo; from the same site.\nThis allocates 1GB of the 8GB memory to the GPU, and leaves the rest for the Linux kernel. If loading is successful, you see a message \u0026lsquo;waiting for clients\u0026rsquo;, then the screen turns black. If the kernel loads properly, the PS4\u0026rsquo;s LED turns blue and violet, and we get via HDMI console output from the booting Linux kernel. If that does not happen, most likely the kernel you use does not fit your hardware, try a different kernel. This should now have started the kernel and run the initrd, and dropped us into a shell, and we can run # exec install-psxitarch.sh The script will copy the files to memory remove the partitions create a small partition with FAT32, and populate it with kernel/initrd and create a second partition spanning the rest of the device, for the Linux distro When the install script finishes, you drop again to the shell, where you can run # exec start-psxitarch.sh You should see various systemd services getting started, eventually the screen turning black. At that point, I need to disconnect my HDMI cable and connect it again. I have 2 devices to process the HDMI signal from the ps4: an IODATA LCD-MF276XDB monitor and a simple HDMI grabber adapter which presents the HDMI signal via usb like a webcam to a computer. The grabber is easier in picking up the ps4 signal at this point. After disconnect/connect of the HDMI cable, I need to switch to the text console with \u0026lt;ctrl\u0026gt;+\u0026lt;alt\u0026gt;+\u0026lt;F2\u0026gt;, and then back to the graphical mode with \u0026lt;ctrl\u0026gt;+\u0026lt;alt\u0026gt;+\u0026lt;F1\u0026gt;. At that point, you should see a login manager, and be able to login - password ps4linux for the distro mentioned before. After the first login The first task should be to open a terminal, and use\n# sudo systemctl enable --now sshd to make SSH login possible. That allows to login, watch logs. Then the distro should be updated - excluding packages as per details in the readme file which you find on the desktop.\nCompiling an own kernel I managed to compile an own kernel, from upstream 5.3.18 sources plus 400kb of patches which introduce a new architecture, which were never upstreamed. The default configs are designed for a monolithic kernel without modules. Compiles fine with Fedora 32 userland, but on Fedora 35 userland binutils is to new. Making an own distro might be fun, with about to be released Fedora 36 userland.\nFurther points Here is a collection of further links and hints.\nThe jailbreak is not persitant, you will have to jailbreak the ps4 after each reboot. There are projects which automate this: a Raspi or ESP32-S2 can span up a wlan network, emulate usb devices, steering the pointer to have the PS4 load the jailbreak, and to offer the special file system for the jailbreak.\nI tried 2 distros: the ITmania one, Fedora 32 based. There is also the ENCOM one, Fedora 35 userland. The current distros do not use YUM-repos for the packages with patches extra for the PS4, that might be useful to setup.\nUsing the internal disk of the PS4 for storage would be interesting, but as the PS4 only executes signed code after boot, the native system needs to stay in place to boot the PS4 up. That would change if Linux loaders could be signed.\n","date":"2022-03-02T14:16:20+01:00","image":"/posts/2022/03/linux-on-ps4/20220303_124228600_blackmesa.jpg","permalink":"/posts/2022/03/linux-on-ps4/","title":"Running Linux on the Playstation 4"},{"content":"Why would one do a trip in January? Due to Corona, left over holiday was stacking up, and everything not taken by end of February would be cut off - so I had to go in the cold season.\nWhere to go? The north of Japan is coldest and has snow by January, Okinava in the south is only reachable via plane or ship, so I aimed at Kyushu. Main reagion I wanted to visit: the Yufuin area, known for Onsen, Japanese bathes. A recent survey lists Yufuin among the top 10 Japanese onsen areas. Here is a map and the numbers of pictures I took, Kyushu is on the left.\nTravel Getting there from Tokyo: plane was out of question for environment reasons, and as trains are great. I took the Shinkansen highspeed train to Hakata at the North of Kyushu, 1100km in under 5 hours is impressive. Of course, I got a \u0026rsquo;lunch pack\u0026rsquo; at the train station, called obentou. These are available in many flavours. On the map you can even see at which spot I took the picture of the obentou.\nThe \u0026ldquo;Yufuin no mori\u0026rdquo; train runs from Hakata, a special train which only operates in Kyushu. The train is aimed at tourism: sweets only available on that train are sold, the interiour is nice, staff is offering to take pictures and chatty.\nYufuin ryokan When booking the stay, I was taken back and forth. Ryokan are the traditional Japanese style hotels one associates with Yufuin, these would be the canonical choice. Yet, when I had seen ryokan houses in Touhoku, I understood these as cold and not really good for winter. Hotels would be more likely to be built solid with warmer rooms. I decided for the ryokan, and my stay turned out to be a quite new building, windows even have 2 layers of glass. The room floor had tatami mats, the hotel had a great inhouse bath where one could from the pool see Mount Yufuin, the mountain watching over the city. Also the breakfast was just fantastic.\nMnt Yufuin The day after arrival I wanted to hike mnt Yufuin, described as \u0026lsquo;90min hike\u0026rsquo;. Just walking over to the foot of the mountain took 2 hours, could have been quicker if I had taken a bus there. It was cloudy that day, the mountain top constantly in a cloud. After starting to walk up, trails turned out to have snow and ice, eventually it started to snow. I do not own requipment like spikes for the shows, and 20min before the summit I decided to return. I had late lunch in the city, and got a great small cheese cake with caramel flavour, which I ate while drinking green tea at the ryokan.\nYufuin sightseeing Sightseeing of Yufuin on the next day. There are a few long streets with shops, art atelies, museums. Towards north I wanted to hike Hidake/飛岳. Android app Osmand (which uses openstreetmap material) showed 2 pathes to the summit. The one from the north would mostly have been along a road, so I opted into the south one - and 10min before that summit again had the way vanish/could not advance further. Ah well.\nKyushu is well known for growing mandarines, and indeed they were cheaper here. There is also Kabosu/かぼす, a kind of lemon which the Oita area in the east of Kyushu is known for.\nThe ryokan also had local food for breakfast, one day a jely made from Kabosu.\nOita The next stop was Oita: reachable in ~1 hour via local train. Oita is bigger than Yufuin, near the sea. I discovered a restaurant with Sri lanka food, that was so good that I ate both days of my stay in Oita there. The tourism office at the train station offers maps with walks through the city, these were nice.\nThe Tenkai hot spring is really great: while soaking in the hot water, you can enjoy views of the sea. The body soap available for washing contained also coal, so it was black. 炭 was written on the label, the area is known for coal.\nKokura and trip back After that, I did spend a night in Kokura/Kitakyushu. The railway museum is quite nice. At the hotel breakfast, I got Karashi mentaiko/ 辛子明太子, another speciality of Kyushu: marinated salted cod roe. The Corona infection numbers were then going up, it was unclear which counter measures would get decided, so after a stop in Osaka I went back to Tokyo.\nOne day after I had left Kokura, the whole area Oita/Yufuin/Kokura were hit by an earthquake, magnitude 5+ in the Oita area.\nThe pictures of the trip are here.\n","date":"2022-01-25T02:15:52+01:00","image":"/posts/2022/01/onsen-trip-to-kyushu/20220117_230901960_breakfast.jpg","permalink":"/posts/2022/01/onsen-trip-to-kyushu/","title":"Onsen Trip to Kyushu in January"},{"content":"Why would one learn foreign languages? Over my journey of learning Japanese until now, I had various stages of motivation. All started with a \u0026lsquo;I like the country, I want to go there again for holiday, and learn the language a bit\u0026rsquo;. Right after starting to learn, progress is quick, that motivates very much. To me, the big achievements so far were these:\nability to give a self introduction ability to tell what I did in the past week telling a joke in the foreign language, and having the other person(s) laughing meeting the first person where my Japanese level was slightly better than their English level, so settling with a talk in Japanese situations in your life when the canonical word to express it comes out in the language you learn, not your native one giving a presentation in Japanese - in a direct talk one recognizes if the own words are not understood, but in a presentation that is very hard Others have goals like \u0026lsquo;I want to pass the Japanese language proficiency testi (JLPT)\u0026rsquo;, which can be taken in 5 different levels. For me, taking the first levels of JLPT was interesting, but towards the higher levels the testes skills (quick reading and answering questions) does not match with my learning goals - I try to optimize for better speaking and understanding in conversations.\nIs your god alive? But what keeps you addicted? Most importantly, a German learning Japanese means to also dive into a different culture. The language concepts are rooted in the culture: unuit have many different words for snow because there was or is a need for these. Meeting language learners, I settled with the idea of 2 big language families:\nEuropean based ones (English, German, ..) Asian ones (Chinese, Japanese, Korean, ..) When I as a native German speaker learned Russian and English in school, then these concepts and the culture was not to far away from Germany, learning was relatively easy. Dipping ones toe for the first time into the other language pool is then harder, one needs to look deeper into culture and different concepts.\nExactly this is where the fun starts! You notice differences, question the background, and also think about your native language. My best example for this are the Japanese verbs for \u0026rsquo;to be\u0026rsquo;. There are 2:\niru/いる, used \u0026lsquo;for living things\u0026rsquo;, like humans or animals aru/ある, used \u0026lsquo;for not living things\u0026rsquo;, like a chair or a car Pretty simple definition, right? So then you start to use these, and notice Japanese talking about the god living in Shrines, and saying \u0026lsquo;Kamisama ga imasu\u0026rsquo;. The English expression is simply \u0026lsquo;gods exist\u0026rsquo;, but knowing about the difference of the 2 possible verbs, one notices: the verb for living things is used here! So, which verb would you use for the god as referenced in Christianity? Noticing such differences just blows my mind.\nBy the way, there seems also to be a change in references to robots recently: 10 years ago I saw many references to these with aru, so as nonliving things. Now with speaking capabilities, reactions and so on, references with \u0026lsquo;iru\u0026rsquo;, so as living things are increasing.\nLevels of respect In German, when you address someone, you have to choose between basically 2 layers:\n\u0026lsquo;Du\u0026rsquo;, the more familiar level \u0026lsquo;Sie\u0026rsquo;, the more formal level This can already lead to \u0026lsquo;fun\u0026rsquo; in many sitiations: for example when you are daily interacting with a customer and address him with \u0026lsquo;Du\u0026rsquo;, and his boss enters the room, there is an awkwardness in addressing the boss in the higher level.\nEntering Japanese studies, this seems like just a childs playground. Japanese society had developed into many more layers, not only fill words are changed to show respect, but whole verbs were newly created to express an activity in just a different level of respect. When I look at something, I would use a different verb than when asking a customer if he has seen my email, and for the emperor again other verbs would be used.\nWords just used by one gender Some things would typically be expressed in one way by men, and in one way by women. Men are likely to refer to meat as niku/肉, while women would be seen as rude to not say oniku/お肉. Words to refer to one self are also different for the genders. When a couple of a male foreigner and female Japanese lives abroad, and a male child gets only to hear his mothers (female style) Japanese, then this is leaving noticable impact on the boys language skills.\nYou can look at these differences like problems, but they are also fascinating :)\nWords to count In English or German we count things always in the same way: one banana, two bananas. One sheet of paper, two sheets of paper.\nIn Japanese, the way of counting depends on the object: banana ippon, banana nihon would be used, as banana are long, round objects. ichi mai, ni mai, san mai would be used to count the sheets of paper.\nThis gets really interesting when for example looking at a Tuna which is caught: still in the water, it\u0026rsquo;s counted \u0026ldquo;ippiki\u0026rdquo;, one living animal counter. When it\u0026rsquo;s caught and the line attached, it\u0026rsquo;s counted \u0026ldquo;ipppon\u0026rdquo;, so one of round/long objects. The corpse on the market is then \u0026ldquo;ichou\u0026rdquo;, and so on. Watch the fun here!\nSo.. how to count phone calls? Like long/round objects! Probably someone was visualizing phone lines like this.\nThere is a word.. for that? Out of different culture, there are words which might not have counterparts with comparable meaning in other languages. For German, a famous example is \u0026lsquo;Schadenfreude\u0026rsquo;.\nI just love Japanese for having words like:\nあげおとり: When you look after a haircut worse then before 金継ぎ: the art of repairing pottery with gold 森林浴: \u0026lsquo;forest bathing\u0026rsquo;, recovery by walking in the woods 辻斬: The act of trying out a new sword on a random stranger I keep a list of these.\nProverbs I\u0026rsquo;m constantly amazed of how deep and fantastic proverbs exist, and it sparks joy to understand the proverbs in their native language. Some examples:\n馬鹿は死ななきゃ治らない, sometimes stupidity will stay all over the lifetime 五月雨式に申し訳ございません, Sorry, this is dragging on and on.. (It\u0026rsquo;s pesky like the rain in May/5th month) 頭かくして尻かくさず, hiding the head but leaving the butt uncovered: trying to cover up something, but forgetting a piece 犬が西向きゃ尾は東, literally: when the dog looks towards west, the tail faces east I collect some here.\nAlong these lines, I have also found some (just a few) pieces of Japanese writing which I like very much, and am just happy of being able to read them natively. For example, Natsume Sousekis the third night.\nKanji Now this won\u0026rsquo;t be relevant for all languages.. but it\u0026rsquo;s certainly for Japanese: Kanji. Europeans or Russians use alphabets with up to 50 characters. In Japanese, we have 2 simple alphabets, each with 48 characters. With these, you can write down already all words. Kanji are used ontop, and texts can contain all 3 kinds of characters.\nFor me, being involved in OpenSource and Linux, it was fascinating to bring Japanese onto the computer. To research input of Japanese characters, to typeset Japanese texts with LaTeX, learn about unicode. Discovering overlap between language learning and existing hobbies is great!\nThere is more.. You learn to know interesting people: as teachers, other students, and speakers of the language you learn. More differences between German and Japanese Gestures are a whole different can of worms! Fun to learn the differences. What keeps you attracted to language learning?\n","date":"2022-01-15T12:42:54+09:00","image":"/posts/2022/01/why-learn-foreign-languages/20220118_044108371_kanjic.jpg","permalink":"/posts/2022/01/why-learn-foreign-languages/","title":"Why would one learn foreign languages?"},{"content":"Christmas in Europe and Japan Before Corona, flights from Japan to Europe in the middle of December were full of couples with one partner from Europe, and the other partner from Japan. Often such couples spend Christmas in Europe: it\u0026rsquo;s the time to meet family, to enjoy a quiet Christmas. With many Christians in Europe, most countries also have national holidays then. On the 24th, Christmas eve, presents are given and received.\nAt the same time in Japan: nothing special. Restaurants and stores decorate in Christmas style - but it\u0026rsquo;s not traditionally important in Japan, there are no national holidays. Some Japanese families go to KFC on Christmas eve, not much else going on.\nTime towards years end After Christmas, the celebrations turn to Japan: here the years end/ years beginning is the important event, due to national holidays and some free days from their company, most people are not working for nearly a week.\nDecember is a busy time in Japan: people start to prepare for the years end. Often people are doing extra cleaning of the house or appartment, to start the new year fresh, without dirt and leftovers from the other year. This is so widely done, that TV weather forecast is reporting on how well the weather fits house cleaning.\nIn Japan, there is a trend of many young people leaving the countryside to work in cities like Tokyo or Osaka, and live nearby. On sequential holidays and especially around years end, there is a strong tendency to return home to the countryside over these days. End of 2020, returning was strongly discuraged because of Corona, but in 2021 we had vaccinations and also new infection numbers down, so many Japanese returned to their family home.\nWhile I send out many Christmas cards, new years cards are sent in Japan - with many rules: they must not arrive in the old year, and if in a family a person deceased in the past year, then sending that family a new years card shows your ignorance of their mourning. Being a German in Japan, I am taking the liberty of doing some things differently, and send out Christmas cards.\nIn Europe, you see Christmas decorations everywhere in December, and in Japan you see typical new years decoration in late December. I can confirm: seing these reminds you of the upcoming quiet new years days, makes you look forward to the upcoming silent days - even though I do not have any family members here in Japan.\nLike many others, I had 6 free days this year, which I spent with cleaning the appartment, 2 times visiting Onsen/bath, hiking on new years day, video chat with family in Germany and friends, investigating some Linux topics.\nCompanies often do years-end or beginning-of-the-year parties, but there were not many this season - because of Corona. Even with infection numbers down, people wear masks, and companies do not want to take a risk. My department did an online drinking party, and we were offered to have the company send us food and drinks via delivery.\nDec 31st On 31st, while in pre-Corona-Europe people were launching fireworks, streets in Japan are quiet. There are some places with celebrating people, and I also heard helicopters, probably out of fear of terror acts, but for most people the end of the year is just a quiet time.\nSpecial meals, most famously: Osechi Ryori. These are a composition of many different meals. Many families create these by themself, or order and get them delivered. I stayed away from these, as I would have to eat all pieces by myself, also the ones I do not like. More details on Osechi Ryouri.\nA further speciality: eating rice cake is said to bring luck. Every year people die from these rice cakes, this year 19 people needed medical treatment, 4 died, that\u0026rsquo;s just the numbers for Tokyo.\nMany stores are closed on the first days of the year - as I do not cook by myself, I find then the few open restaurants to be very crowded.\nNew years activities What\u0026rsquo;s typical to do in the new year? Most importantly: the first shrine visit of the year, called Hatsumode. This year, I went hiking to Jinbasan/Takaosan on January 1st, and along the way did my Hatsomode. People are the queuing up in front of the shrine to pray. Many buy a new talisman, called omamori, and dispose the one which protected them the last year.\nIn the first few days of the new year, people often greet each other with a special greeting: あけまして、おめでとうございます。 ことしも、宜しくおねがいします。 Meaning: \u0026ldquo;the new year started, congratulations. Let\u0026rsquo;s also spend this year in good terms.\u0026rdquo; This can be efficiently be shortened to あけおめ、ことよろ. but the short version should only be used with good friends, as other people could assume that they are not worth your efforts to say the full greeting.\nLucky bags: many stores sell these in the first days of the new year. Typically, these contain a variets of the normal products of the store at a reduced price - but you do not know exactly what\u0026rsquo;s inside. Some of the bags allow you to peek inside to some degree. These are quite popular and often sold out quickly. If you have a big family, you can buy these and give away what you do not need - but as I would have to dispose that, I stay away from these bags and just buy what I really need.\nDo I as a German, used to celebrate Christmas, get caught in the new years mood in Japan? Definitely! The new years decorations on houses are everywhere, so one is constantly reminded of the upcoming quiet days - very nice, much like German Christmas mood.\nWhat\u0026rsquo;s to be missed? At least I miss ginger bread.. it\u0026rsquo;s not even available here in stores specialized on imported products. Stollen (a kind of Christmas cake) is available, imported from Germany.\nMore pictures are in the Years end/ beginning gallery.\n","date":"2022-01-04T13:16:19+01:00","image":"/posts/2022/01/new-year-in-japan/20170102_165500_evening_fujisan_mod2.jpg","permalink":"/posts/2022/01/new-year-in-japan/","title":"What are new year celebrations in Japan like?"},{"content":"I was asked to share my experiences about the System76 launch keyboard, and a Mastodon post would not provide enough space - so a blog article it is.\nMy setting and needs What\u0026rsquo;s my environment like, making me consider a 285$ keyboard?\nSince forever, I did not pay much attention to the keyboards I use.\nIn the last 5 years, I mostly used the keyboard build into Thinkpads, with German and English layout. From a trip to Korea I brought an external keyboard back with me which I used at work, also here I used English layout. I use Linux, and ibus and fcitx5 as input switch methods, so I can input German, English and Japanese characters.\nSince I moved to Japan 5 years ago, I live in a small room in Tokyo, considering carefully which hardware I buy and which not. So far, I constantly used Thinkpads with open lid, in front of a 24\u0026quot; monitor. The Thinkpad keyboard is quite nice, and because of their fantastic track pointer I do not need a mouse. In my setup, I do not use the Thinkpad montor when an external monitor is connected, so it just wastes space on the desk, and collects dust.\nSo, do I need a keyboard at all? After all I try to stay lean, to not buy more things which I do not use eventually.\nWhen we were forced to clean our desks at the office, I had taken the Hangul keyboard home with me, refurbished it and tried it for home use. I liked having real keys, but it was to bulky for me, took to much space.\nThe reports on the System76 launch keyboard looked nice:\nit has an open firmware, can be patched with fwupd an OpenSource software is available to do key remapping it\u0026rsquo;s small: an 80% keyboard. Just look at it: it\u0026rsquo;s reduced to the really essential parts - but these are of high quality. it\u0026rsquo;s not only connected via USB to the computer, it also has a USB3 hub, one can attach a mouse, USB sticks, the mp3 player for uploading podcases and the mobile phone directly. Market watch: which other keyboards are out there? I had seen reviews of the Launch keyboard and considered it as my first choice - but before buying a keyboard for that price I had to do a proper research of what else is on the market.\nMany people build up their own keybards, I did not want to do that. The \u0026lsquo;happy hacking\u0026rsquo; keyboards are everywhere, and similiar in form factor to the Launch. They also are easily obtainable via many online stores, while System76 only sells directly from America. But then, \u0026lsquo;happy hacking\u0026rsquo; feels like a buzzword name to me.\nSystem76 produces open firmware for the keyboard, and then they are also active in other OpenSource areas, for example with an own Linux distro. So, I ordered the Launch keyboard.\nUsing the launch Shipping to Japan: 285$ for the keyboard, 85$ for shipping, plus 1700円 (15$) tax which I had to pay when the keyboard arrived.\nIt\u0026rsquo;s heavy enough to stay on the desk, comes with some replacement key caps. After connecting, I had to reconfigure the Sway window manager, and noticed that one of the 2 shift keys did not slide up after it got pressed down. A tool for removing the key caps came with the box, I removed the key cap. This certain key is quite big, below the cap is one switch and 2 stabilizers. After moving the stabilizers up and down a few times, and again placing the key cap on top, everything works as it should.\nBy default, the LEDs behind the keys shine with a rainbow pattern, constantly moving. The \u0026lsquo;keyboard configurator\u0026rsquo; has some dependencies, but can be compiled on Fedora without issues, and then allows to configure the LEDs: disabling, different animation patterns, brightness, all configurable. On the Thinkpads, the LED makes the key cap inscription readable, but on the Launch, the illumination is not much helping with readability. On the day after receiving the keyboard, I was typing in the LUKS disk encryption passphrase when it was dark - and needed 3 attempts.\nLED brightness can be configured without the software, just pressing certain keys on the keyboard. More complicated changes are done in the software, and is then uploaded to the keyboard.\nI also used the key remapping function of the software: I changed the key caps on the bottom to match the Thinkpad order, details are here. With that change done, I feel now quite comfortable when using the keyboard.\nFascinating the Launch has just exactly the keys I need. The F-keys are there, and I heavily use ctrl + cursor keys for switching virtual screens. If I find out that I need more, then I can still use a modifier-key which I do not use today, and map further keys which I can press together with that modifier key.\nMore restructuring Apart from plain keyboard usage, more on my desk changed.\nSo far, I used the camera in the Thinkpad lid for online conferences, and often also the Thinkpads microphone and speaker. With the new keyboard, I closed the Thinkpad and moved it behind the monitor. That frees up space on the desk, but for now I have to use an external webcam, and for audio a bluetooth connected headset. These 2 pieces are an inconvenience compared to the old setup, but for other video conference participants this improves audio/video quality.\nI had thought that through before buying the keyboard. One nice thing which turned out nicer than expected: the Launch includes a USB3.2 Gen 2 hub, and offers 2 Type A and 2 Type C connectors, hooking up mouse, USB-sticks and so on is convenient, and fast.\n","date":"2021-12-26T02:49:10+01:00","image":"/posts/2021/12/launch-keyboard/20211223_044425696_launch_sun.jpg","permalink":"/posts/2021/12/launch-keyboard/","title":"The Launch Keyboard"},{"content":"What\u0026rsquo;s here? The Linux unplugged 437 podcast announces the software pieces which were most important for the listeners this year. After listening to that, I wondered: how changed for me in 2021 regarding software?\nPodcasts I listen to podcasts when going for lunch or hiking in the mountains around Tokyo, and a big share of the podcasts is tech related. The ones which I discovered and started to listen to regularly this year:\nLinux unplugged has weekly Linux news I\u0026rsquo;m into self hosting services, the Selfhosted podcast gives great ideas for new pieces of software to try out The Iron Sysadmin podcast is around sysadmin things - and has also quite some American views on what\u0026rsquo;s going on in the world Bilingual news - news on societs and technology - also helps me to improve my Japanese, and has both American and Japanese views Self host everything! With that, I refer to the idea of hosting as many services as possible by myself. That is to learn about technology, to be able to directly relate to the topics from the sysadmin world, and because it\u0026rsquo;s just fun to discover and run new software. Some of these services got extended this year:\nWeb: stayed nginx, but the root site is now static and points at the various other software pieces The wiki runs on Dokuwiki as before The blog stays Hugo based, but now on the Binary theme The Japan life observations are based on an a text file and rendered to html with asciidoctor Additionally to the existing fgallery based picture albums, I did setup Lychee on pics.fluxcoil.net, hosting the Hokkaido trip images Many services stayed: postfix for my email, syncthing for file sync, PerformanceCo-Pilot for performance monitoring, Redis/ Grafana for visualization, Tiny tiny RSS (TTRSS) for collecting feeds and news I played with Jitsi for hosting video conferences this year, but have taken it down again, not easy to debug in daily operation. As for social media, I\u0026rsquo;m on twitter, but much more active on the decentralized Mastodon. For now, just using a Mastodon account on a different instance, but one can also host that by oneself.\nA huge change regarding my hosting of this year: I started to use containers. This came in with Lychee: it\u0026rsquo;s a beast with many PHP modules which I wanted to some degree separated, and easy to roll back. For now it\u0026rsquo;s in a Debian container. The newest Lychee version has jumped to PHP 8, which almost no distro has as of now. Also TTRSS runs now in containers, Alpine based.\nMobile phone My main phone stays the Pixel 4a. I\u0026rsquo;m happy with the hardware, and tried to ungoogle it this year with custom images: after much fiddling I\u0026rsquo;m now back on the Google image: pictures taken on the custom images had an unbarably bad quality.\nMy company SIM runs in the Huawei P20 lite: after disabling 4G, it has an amazing battery lifetime. I attempted to use a custom image, but a code is needed to unlock the bootloader - shame on you Huawei, I won\u0026rsquo;t buy from you again.\nContributing code At work, I use a wide spectrum of skills: communication with partners, customers and colleagues, debugging, documenting, writing articles, giving technical presentations and more. Speaking at PCP.Conf 2019, I mentioned the idea of a software piece for Linux to collect metrics about electrical energy consumption. Kind of powertop, but recording the most important pieces. I gave this a try this year, the code is now in upstream as pmda-denki, and will be in RHEL8.6. Denki/電気 is Japanese for \u0026rsquo;electricity\u0026rsquo;, I\u0026rsquo;m proud the project was ok with that name. Details on using pmda-denki and findings are here and here.\nDesktop End of last year I switched to KDE Plasma, after many years on WindowMaker. It\u0026rsquo;s nice.. but it felt a bit heavy, and finally when failing to properly configure an external keyboard, I made the switch to the Sway window manager. Sway is Wayland only, which came with own challenges, but at the moment I\u0026rsquo;m quite happy with it - details are here.\nSway is a tiling window manager. On WindowMaker and Plasma, after start of the window manager, I always started terminals on some virtual screens, arranged them to a size which seemed fitting, and so on. Now, I just start terminals and by default Sway decides on the size - actually quite convenient. I have used gkrellm to display monitor network traffic, system load and so on for many years - no longer using that, but now the most important status details are in the Sway status bar on the top of the screen. I feared I would no longer have a nice overview of the system, but it turned out this is actually all I need. :)\nAlong with the switch to Sway, I switched to fcitx5 as input switcher, and Mozc as Japanese input engine - working really nicely! I experimented with some terminals, but stayed with xterm: works nicely also for Japanese input, and supports the fonts I need. I use mail client (mutt) and IRC client (irssi) from the commandline, also these did not change this year. Also on the desktop: qiv (image viewer), Libreoffice, Inkscape, Gimp.\nAlso desktop: for video watching I use mplayer since many years: reasonable performance, I\u0026rsquo;m comfortable with the keyboard controls. VideoLanClient (vlc) is also installed, but rarely used. Now this year, I have hit a 4k video, which the player needs to scale down to 1920x1080 for my screen - I did not get this done with reasonable frame rates on mplayer, so will use more vlc from now on.\nSmaller devices Rockbox: This year, I discovered the Fiio M3K mp3 player running rockbox as a nice platform to code, where one has to be aware of technical limitations. My goal was to port \u0026lsquo;bad apple\u0026rsquo;, details are here. Rockbox is really recommended, a nice project and a great community!\nRaspi4: I was curious how well USB HDMI grabbers work, got one and played with it. Quite usable, actually!\nNintendo Switch: I tried out the new Ubuntu@Switch version from the switchroot project, Linux support got really far on the Switch. Nintendo also released the new hardware revisions which no longer have the hardware glitch which allows to run Linux. The last Fedora release which I brought to the Switch was 32, the Switchroot project now provides newer builds of Fedora and other distros. Cheers azkali and gavin!\nGames I read the RSS feed from GamingOnLinux, and keep a short list of games around which I am interested in. There was much movement this year, especially the SteamDeck might be big next year. Last year, I also purchased a small hand held console, looks a bit like a playstation portable, which runs Linux and all kinds of SNES/Amiga and so on emulators. I\u0026rsquo;m not using it at all.. if you are in Tokyo and think it can be of use to you, let me know.\nI spent most of my gaming hours this year in Black Mesa, a Half life clone, and ParticleFleet.\nUpcoming Looking forward to some things:\nI ordered the System76 launch keyboard, which can nicely be updated with fwupd, looking forward to using it. Would like to play a bit with the new Apple aarch64 based hardware - but it\u0026rsquo;s to expensive just for playing with it a bit, and Thinkpads are my daily drivers currently. Curious what the Linux based SteamDeck (hand held gaming machine from Valve) will be like. Also the Pine note, an e-ink device meant for Linux looks quite nice. There is my list of project ideas, and while pmda-denki and \u0026lsquo;getting bad apple onto rockbox\u0026rsquo; came off the list, the list got extended with fresh ideas. Let\u0026rsquo;s make it a good year 2022 for OpenSource! :)\n","date":"2021-12-22T12:35:35+01:00","image":"/posts/2021/12/2021-opensource-linux/20211102_191636759_keyboard_night_small.jpg","permalink":"/posts/2021/12/2021-opensource-linux/","title":"How OpenSource software influenced me in 2021"},{"content":"How are society and the low Corona infection numbers in Japan related? I was looking forward to quite some things when moving to Japan, to the food, bathing culture and more. I was also signing up to other aspects, like a society where people are watching each other, more than in Germany.\nA first flavour of that is in TV series about Japanese life: older woman with a worried face picking the trash bags of others from the outside, verifying if trash was correctly separated and the rules got followed. People are also, compared to other countries, very sensitive to noise in the house. One of the fears of renting a room to a foreigner are loud parties, which would disturb others.\nI was aware of the \u0026lsquo;watching\u0026rsquo; already before moving here, and made my piece with it. I also had to cut down some things, for example I did not dare to buy an electrical guitar here in a rented room in Tokyo, fearing the noise for others. A friend with electrical(!) drums did get hinted that others heard his drums.\nMe being me, I keept thinking: now if you have a society where people are watching each other to some degree.. there surely must be good aspects?\nComparing the Corona situation in Germany and Japan, I found an answer: in times like this. the \u0026lsquo;watching\u0026rsquo; helps. This occurred to me last week Saturday, when I was at the top of Mnt.Nabewariyama/鍋割山. Daily new Corona infections in Tokyo are below 30 since weeks, but here hikers were lined up in front of a hut, waiting to buy \u0026lsquo;Yamaudon\u0026rsquo; for lunch. Despite infections down, and all at the fresh air, all wearing a mask. For comparison, in Germany, with numbers exploding, there was recently a soccer game - and you saw a wall of fans, almost noone wearing a mask. Why do people wear a mask here? It\u0026rsquo;s less that all know in detail how infection happens, but because a majority wears a mask.\nAbout vaccinations: Japan was very late. Germany started around beginning of 2021, Japan started middle of the year, and vaccination rate is now the highest in the G7 countries. Also in Japan, we have mistrust in the government and so on, but once enough people were vaccinated, people wanted to do their part.\nI think I saw the good part of the \u0026lsquo;watching each other\u0026rsquo;.\n","date":"2021-12-06T12:48:15+01:00","image":"/posts/2021/12/society-watching-and-corona-in-Japan/20211204_023640735_nabewaliyama_fujisan.jpg","permalink":"/posts/2021/12/society-watching-and-corona-in-Japan/","title":"Society Watching and Corona in Japan"},{"content":"..but why? My employer asked us TAM\u0026rsquo;s in Japan to fetch our belongings from the office, so I got home with the keyboard I bought in Korea, with Korean Hangul - and 3 big places of rust.\nI made a weekend project to remove the rust and repaint, worked ok. Then: hooking it up to my Thinkpad, currently running KDE Plasma. Basic configuration worked ok, Plasma was coping with the Thinkpad having German layout and the Koran one having English layout, plus extra keys. But then, I tried to configure keys for changing the virtual screen, and was not able to do that. The error was about the focus being configured to follow the mouse, while the desired key combination was not compatible to that. After 30 min of trying around I gave up.\nThere are further points about Plasma for me: after initial installation via Fedora package group, I also had a search engine installed which took down performance as it indexed by complete homedir- I do not need that, I can search my files directly when needed, and then more targetted than a search/index machine can do. Also, things like attaching the mp3 player by default opening a window and offering actions - no, I am fine with doing a plain \u0026lsquo;mount /mnt/walkman\u0026rsquo; then as user, and moving the new podcasts around on the shell. Things like this make me wonder if Plasma is doing more handholding than I need.\nSway! I was running sway since quite some time on the company provided Thinkpad, but now it became my main driver on the private Thinkpad L480. The basic config ~/.config/sway/config was copied over from the company system. Configuring the external keyboard took just 5min, including the key bindings to move to next/previous virtual screen.\nWhen I started Plasma, it took quite some seconds, Sway in contrast is small and quick. With Plasma, I started to execute gkrellm, which shows stats like cpu/network load, then I started terminal instances, and moved them around with the mouse to always the same sizes and places on the screen.\nWith Sway, I have currently just the most important monitors, and directly in the status bar at the top: network bandwith in/out, cpu load, consumed memory, fan rotation, battery charge level, time. As Sway is tiling, when opening the first terminal it consumes the whole screen. Opening the second terminal, both get equal shares of the screen by default. So effectively, I spend less time of changing sizes of windows than I did on Plasma.\nCurrent painpoints I live in Japan, and the default Input Method framework on Fedora/RHEL would be ibus, but due to a bug, it is not yet usable on Sway. fcitx works reasonably well.\nBiggest issue is screen sharing with video chat software. So far I used Plasma with Xorg, and all of Skype (native Linux application), Bluejeans, Jitsi, Google meet (all services usable via Firefox or Chrome) were able to grab the whole screen and share it in the video confcall. Sway is Wayland based, and apparently whole screen sharing is not yet possible.\nIs that important at all? For me, yes. I have sessions with colleagues or Japanese schools where it\u0026rsquo;s easiest to just share the screen, then together do an internet search, or I run \u0026lsquo;dia\u0026rsquo; (an OpenSource diagram software) and draw up something to underline an explanation. Or I just share the screen and then run an image viewer to show family members pictures taken in the last week.\nSo what to do on Sway for this? So far, I came up with 2 workarounds:\nusing v4l2loopback and wf-recorder to record the current screen, and make it available as device, for example /dev/video0. That can then be chosen as webcam, and transmits the screen. pro: transmits the full screen, when I change to other virtual screens, it follows con: it puts quite some load onto the system con: seems mostly to be presented with bad quality at clients second method, webtops container. This basically runs a window manager in a container, and you access that desktop via web browser - and as Chrome and many video chat softwares allow sharing of browser tabs, this allows sharing with the others. con: requires to setup and run an extra container con: whatever you want to do, you need to do it in that container instead of your normal desktop. So might require installation of extra packages there, etc. pro: less cpu load than above workaround pro: good quality for the client My Sway notes and hints are here in the wiki.\nNext steps I\u0026rsquo;m now in the market for looking again at terminal emulators. Using xterm since many years, I did do far not manage to get URLs in xterm clickable, so that double-click of one character of https://example.org marks the whole URL, copies it into the buffer and opens it. Also increasing/decreasing font size would be nice, and as I also use irssi (IRC client) and mutt (mail client) from the terminal, I need it to deal properly with unicode and Japanese fonts. \u0026rsquo;terminator\u0026rsquo; looks like a good candidate.\nLet\u0026rsquo;s see how it goes!\n","date":"2021-11-07T12:59:25+01:00","image":"/posts/2021/11/moving-to-sway-windowmanager/sway-screen.jpg","permalink":"/posts/2021/11/moving-to-sway-windowmanager/","title":"Moving to the Sway Windowmanager"},{"content":"Japan online lesson experience and online learning I was recently talking for 1 hour with a Japanese class and teachers, about the internet, Open Source, and what online lessons can look like.\nState when Corona started When Corona started, I heard from various people about the state of online lessons at schoole: it was really bad. In Germany as well as in Japan.\nKnowledge on how to do lessons via Internet instead of teaching in front of students and children Internet connectivity, devices for access I was worried about that. Schools in both Germany and Japan were facing these questions, in danger of just getting visited by Google or Microsoft sales, and then buying \u0026lsquo;something\u0026rsquo;, which might mean tax payers money spent on software or services which lead to vendor lockin, personal data being at risk, and more. To be clear: although I use very much Open Source, here I merely would like the people in schools to make educated decisions, instead of buying something from whoever knocks their door first.\nThe Tokyo Linux user group mailing list was the place I took this to. Linux user groups are organizations which are spread all over the world. In the first days of the Linux operating system, when internet connectivity was scarse, LUGs were there to help: on mailing lists, regulars tables, and doing \u0026lsquo;Linux install parties\u0026rsquo; where you could bring in your hardware and walk out with a Linux installation.\nFeedback summary: others confirmed the sad state of online learning, were also willing to spend time on helping schools, but were like me unsure about a good interface to offer advice.\nI have friends in Germany who were involved in activities of technically affine people giving lessons, even before Corona, called \u0026lsquo;Chaos macht Schule\u0026rsquo;.\nThe first online lesson In October 2021, I had a one hour session with a school class in Hyogo-ken, Japan.\nI was invited in person, which I refused considering the travel would be many hours. So, the lesson happened online, using a normal video chat software which transmits audio, video, and allows to share ones screen.\nWhy me? I am proud that I was not invited as a foreigner, but as a technical specialist, to bring in \u0026lsquo;fresh perspectives\u0026rsquo; on things.\nThe first question was a wide topic: what is the internet? I shared my screen, started \u0026lsquo;dia\u0026rsquo;, an Open Source diagram software, and draw a client, an ISP routing packets, and web servers.\nAlong with that, terms like \u0026lsquo;system\u0026rsquo;, \u0026lsquo;service\u0026rsquo; and so on came up. Then \u0026lsquo;why should I as a student learn how to use the internet?\u0026rsquo;. \u0026lsquo;What is Open source?\u0026rsquo; \u0026lsquo;How can the internet help me to teach?\u0026rsquo; \u0026lsquo;Which software exists, and which fits my needs best?\u0026rsquo;\nNext steps The answers to these questions are actually complex, I started to collect bullet points as answers online, will use these points at the second lesson which has already been scheduled.\nI started a subfolder \u0026rsquo;en\u0026rsquo; in the wiki, potentially I will do for example a Japanese translation. For the first lesson I created slides as pdf. If I end up with a Japanese translation and get a feeling that would be used, I might render the markup files into a pdf file with pandoc. If I get really much input, I might move these documents to a git instance and accept change requests as pull requests.\nCan you help with the following? I started to write up the answers, input is welcome. For sure, there are more schools than this one in this position, having these questions. How to connect them with technical people who do not have an intention of selling something, but can give objective advice? Marketing people of companies might not be the best ones for this, maybe members of Linux user groups are good candidates: we are paying taxes, many of us have children, so we want the money spent well, and the children to receive good education. ","date":"2021-10-22T14:57:57+02:00","permalink":"/posts/2021/10/japan-online-lesson/","title":"Japan Online Lesson"},{"content":"My view so far So.. my view on Linux containers, managed with i.e. podman or docker so far were this: complicated, and for servers.\nI run Linux for myself on desktops, and on servers, where I host things like DNS/bind, webserver, Wikis, Tiny tiny RSS for collecting RSS feeds, and so on. I run these services on a virtual system in Germany.\nSo far, almost all of these services were running natively: that makes it easy for me to update them as I trust the Linux distro maintainers of the distro to fix security issues when they get known.\nFor work, I have looked at containers, but not considered them for my servers. Introducing containers there would just make things more complicated: instead of just patching the host system, I would need to patch all of the containers by themself.\nFirst omen.. TTRSS Then, Tiny tiny RSS, which I am a huge fan of, started to move to containers: the default way of installing is now container based. Want to install natively? You got to dissect the container scripts then..\nNew challenge: Lychee image gallery Not just since I moved to Japan, I like to take pictures and make them available. My journey of image software started with Album-generator from MarginalHacks, Perl software which generates nice static albums like this. Next step was fgallery (example), which is unfortunatelly not continued, and I did not get warm with the followup software.\nSo new candidate: Lychee. I fired up a KVM guest with Debian, and installed it. Installation works fine, and an album which I imported for testing works great, just one pain point: while installing, it installs \u0026gt;130 php modules. This really puts me at unease: the more code is running, the bigger are chances that security issues with the code get known. PHP, and 130 modules which need to be kept an eye on.. running something like that on my system makes me think. Pixelfed, another candidate for a gallery software, is not better either..\nGetting more opinions So what to do about this? I walked over to tlug.jp, started a thread, and got great advice. Summing up some ideas from myself and the thread:\nI could rent a completely new KVM guest, and on a different KVM hypervisor. With that, only my pictures would be at danger: if the system gets hacked via security issues in Lychee, just the image data can be obtained. I could apply things like SELinux, which wrap around i.e. the PHP processes, and can with proper policies restrict the processes to just what they really need. This lowers the chances of an intruder doing something bad, and one can also use this to detect when something unusual happens. Linux containers. Various opinions exist, I think they provide a \u0026lsquo;weaker\u0026rsquo; security protection than KVM guests, but it\u0026rsquo;s better than nothing. Ontop, I can now do this workflow: I can setup the container in my workstation running Fedora, import images, and when satisfied I transfer the container to my server and take the service online there.\nAlso: having an application like Lychee in a container, allows me to easily commit the current state to an image, then do an upgrade, and if that fails roll back to the image I had taken. Also, having applications in the container makes updates of the host more easy.\nStill not sure if I will stick with Lychee.. but it seems like containers are not only for Enterprise but can also help me. My Lychee gallery is here.\n","date":"2021-09-15T13:50:10+02:00","permalink":"/posts/2021/09/containers-for-selfhosters/","title":"Containers are also for Selfhosters"},{"content":"What is this about? Rockbox is an OpenSource operating system for mp3 players. I had many mp3 players which were lacking functions which I would really have liked. Trivial things like allowing playback without pause between tracks, or loudness normalization. It does not stop there: Doom has been ported, one can play mod and sid files from the Commodore Amiga/C64 era, and much more.\nRockbox is a replacement firmware for various mp3 players. Much time for research is required to port to new mp3 players, some time ago I got a FiiO M3K when the Rockbox port appeared. My Rockbox/M3K notes are here.\nWhy code something for an mp3 player? We see a trend of people collecting and coding for retro computers. The hardware running Rockbox is mostly to weak to run even Linux. So I wondered how much effort it would take me to so some basic graphical and sound things on the M3K.\nBad apple ..is in some areas of the internet almost a classic to be run on \u0026ldquo;unusual\u0026rdquo; devices, just do a video search for the term. The \u0026ldquo;default\u0026rdquo; animation is this, and it did not exist for the M3K - so I tried to port this.\nBuilding Rockbox Installation of Rockbox, if you just want to use Rockbox and not code, is simple: one has to install a bootblock on the mp3player, unzip a file onto a microsd-card, and that\u0026rsquo;s all.\nFor writing code, I started with doing an own build, so producing the contents of that zip-file for my mp3player myself from the sourcecode. Also quite easy with the instructions from the homepage:\nFirst one builds a toolchain for crosscompiling. My system here is x86_64/Fedora34, and the cpu in the M3K is mips. Then, checkout of the code via git, having PATH point at the just built crosscompiler run configure: this is asking for which target (mp3 player) you want to build for. Then run \u0026lsquo;make zip\u0026rsquo;, and the build of Rockbox appears. Coding environment When running \u0026ldquo;configure\u0026rdquo;, I noticed that also a simulator target was offered! So, after building that, I could run \u0026ldquo;rockbox\u0026rdquo; on my x86_64 system, it would present me an SDL-application, just like the real hardware! One can then on the emulated screen run code, just like on the real player. Only pitfall: performance on x86_64 is higher, which is relevant when playing animations.\nNext, I had to choose between 2 pathes.\nCoding with lua: this is an interpreted language, one deploys a file.lua on the player, and executes it without extra compile steps. So, I could open a file in vim on the x86_64 system, make changes, and then run the script on the emulator without extra compiling. Lua is slower than pure C, but I found the functionality I needed like setting pixels and playing music, so I was going with this. Direct C: faster, needs extra compile step First code So, I started with modifying existing lua scripts from Rockbox, that worked nicely. Setting single pixels, check. Towards getting an animation onto the player, I implemented a simple picture displayer. Format of choice: bmp. Wikipedia has the details. Basically, in the file header one reads the X- and the Y-resolution for the encoded picture. Then the encoding starts, for all pixels, one after the other, the RGB-values, as one byte. Seeing the first bmp picture rendered on the M3K was nice :)\nTowards animation Next step towards the goal: showing pictures as fast as we can after each other, and see which speed we get with that. For this, I fetched above video with youtube-dl, separated sound into an mp3 file, and rendered all frames into single bmp pictures with ImageMagick. The players display is 240 pixels wide, so I converted the ~6500 frames into 240x120, and had the code show this as animation. The native video runs at 30 frames per second (fps), as expected, the result in the emulator was: 10 fps, and on the real hardware even slower. Well ok, this was just the first try.\nReal animation First idea for optimization: so far I was opening ~6500 files and closing them.. that should come down. Also something interesting: looking at the bmp files with \u0026ldquo;hexdump -C\u0026rdquo;, I noticed that all 3 bytes for RGB were at the same value, so storing that just one time would help. So, next try: putting all the pictures into one big file. I was basically writing an own data format here, so I also wrote C-code on the x86_64 system to convert these pictures into one big animation-file. When opening files in Lua, I had settled with reading the whole file and then parsing it - that did not work for this huge animation-file, out-of-memory.\nNext idea: making \u0026ldquo;macro image files\u0026rdquo;, so taking 4 or 8 single frames, and putting them into a macro-image-file. Why 4 or 8? I found that files should stay below 500 or 700kb to not cause trouble when opening. The result was working, but not really giving more fps.\nI experimented with dropping every second frame, but it was still to slow.\nDiffs Most pixels are actually not changing to the next frame. So next idea was to basically write a real animation file format, which is not painting a complete picture each time, but just processing a \u0026ldquo;list of changes from the last frame\u0026rdquo;. Conveniently, the first frame is also black. So with new C-code, I created 6500 diff-files, each one describing the diffs to the previous frame. Then describing 8 frames per file helped further. I did never look deeper into the common video formats, but I recall that in the past you could select via options after how many frames \u0026ldquo;a complete frame\u0026rdquo; should be used instead of \u0026ldquo;diffs\u0026rdquo;.\nSound Animation got a bit better now, so I wanted the sound in. Simple example code playing mp3 worked. My animation code too, but not together.. turns out that memory was the issue, I had to take out optimizations which did cost memory.\nCutting resolution I really did not want to.. but going down to 120x90 pixels made things really smooth, and while I have not implemented sleeps or framedrops to get the animation perfectly into sync, that works now at least to some degree.\nA video of the result is here, you can get the code and instructions how to setup from github.\nIf anybody is interested into writing code for somewhat ressource constrained hardware, mp3 players and Rockbox are recommended :)\n","date":"2021-08-09T10:51:42+02:00","image":"/posts/2021/08/rockbox-badapple-m3k/20210809_100328719_rockbox_badapple.jpg","permalink":"/posts/2021/08/rockbox-badapple-m3k/","title":"Getting Bad Apple onto the FiiO M3K mp3player"},{"content":"In the previous article I talked about external meters and pmda-denki. Now lets see how much power certain hardware components need.\nPmda-denki usage pmda-denki is now in PCP upstream. One can use the code from there, via packages build regularly, or wait for the next PCP release to make it into the Linux distros. With that done, and pmda-denki being installed, metrics regarding electrical power are available. On laptops where pmda-denki can access both sources:\nAs per \u0026lsquo;pmrep .. -t 20\u0026rsquo;, in 20second timesteps a new line is printed. From the left, we get\npackage: sum of all RAPL values, in Watt. RAPL is available on recent Intel and AMD cpus. core: consumed cpu power via RAPL uncore: GPU consumption, if onboard-GPU dram: memory All of these are actually counters from the sys-filesystem, and PCP computes the actual rate. The next 3 values:\nenergy_now_raw: the current battery charge, in Watthours. Read from sys-filesystem. energy_now_rate: rate of charging/discharging, computed by pmda-denki in 60second steps power_now: rate of discharging, computed by firmware, updated frequently Why so many battery values? Let\u0026rsquo;s consider these graphs:\nBoth graphs show the battery charge level over time, while the system consumes energy. Our application is running, and we want to know the amount of energy consumed while the application was running.\npower_now reports current consumption, that metric is updated by the firmware roughly once per second. If we want to know the average consumption over a certain timespan, or if we want to compute the \u0026ldquo;amount of consumed energy\u0026rdquo;, then we need to query this metric at least once per second, store all of the single values (shown as stripes s1-s5), and then calculate based on that.\nOn the right, we use energy_now_raw: this metric has the current battery charge level. The overall consumption between points in time t1 and t2 is the area below the curve. We can simply measure the charge level at times t1 and t2, and then calculate how much energy was consumed in between.\nenergy_now_rate is calculated by pmda-denki, using energy_now_raw it calculates the rate of charging/discharge in 60sec steps.\nSo: power_now shows current consumption, frequently updated. For more accurate measurement, energy_now_raw should be used.\nWhat can we do with these numbers?\nHardware component consumption With the tools mentioned I tried to understand how much single components consume. For cpu we see it directly from RAPL, but not for example for wlan. First step: get a systems consumption down as low as possible, here a Thinkpad T460s. Log in from via ethernet, turn off screen, run \u0026lsquo;powertop \u0026ndash;auto-tune\u0026rsquo;, and \u0026rsquo;tlp start\u0026rsquo;, stop unneeded processes. Measure that, then put single components under stress, and measure again. Results, raw data, graph:\nFrom the left, we see consumption of the idle system. Used hardware: Nintendo Switch (aarch64), Thinkpad L480 and Thinpad T460s. Then consumption while stressing SSD/NVMe, ethernet, wlan, SSD/NVMe + bus/cpu, activated display, and running cpu loads. So far only battery based metrics were used. As we know the idle consumption, we can estimate consumption of the single components:\nFor this, also the RAPL metrics were used - they are not available on ARM. The T460s has an SSD and the L480 an NVMe, thus these differences. The \u0026lsquo;hdparm -T\u0026rsquo; tests stay inside the SSD/NVME, the NVMe does not only need less power but also shows 19GB/sec transfers instead if 8GB/sec for the SSD. For ethernet, wlan and the other tests the results are comparable.\nSoftware For software, these 2 scenarios seem interesting:\nconsumption of one and the same software running a workload, but comaring different versions of the software. It\u0026rsquo;s common to compare speed of new versions, but comparing power would be interesting too in times of climate crisig. energy_now_raw is good for this. taking one certain task, and compare how much power various programs need to perform that task. If you sit in a plane with a laptop on battery and want to watch videos, you want to use the software with the lowest consumption. Let\u0026rsquo;s look deeper into the second scenario.\nGiven task, various softwares raw data\nOn the left, we have the unloaded system again, then the screen with brightness at 50%. Then for task \u0026lsquo;play this h264 encoded file\u0026rsquo; consumption with various software. Specialized software like mplayer/vlc does better than browsers. ARM does energy wise better than x86, that was also expected.\nOn the right, playing this webgl demo on a browser is the task, and it\u0026rsquo;s impressive how good ARM does here. We should be fair here though: I did not look at frames-per-second or other quality signs, so it\u0026rsquo;s possible that chrome renders much more smooth than firefox.\nWhere is software burning power? If we can trust the RAPL values, we can also estimate where software is burning power.\nThis is all from the Thinkpad T460s, as single exception the values from the Wipeout-firefox-L480 test. We see that over all of the webgl and h264 tests, display and \u0026ldquo;other\u0026rdquo; consumption was quite constant. The \u0026ldquo;other\u0026rdquo; is basically \u0026ldquo;everything I can not account more precisely\u0026rdquo;, after looking at overall consumption (from battery metrics) and subtracting the RAPL values. Wipeout-firefox on both Thinkpads shows almost identical consumption patterns. Differences in cpu and GPU consumption between the programs are big.\nHow much can we trust these numbers? Hard to say. As for RAPL, there are studies painting them as \u0026ldquo;not to bad\u0026rdquo;. My approach to get an idea on how good the battery metrics are: I used an external consumer via usb, whose consumption I could nicely measure externally. So I used the idle system, hooked up the consumer via usb, measured the consumption externally, and looked at power_now and energy_now_rate.\nResults: Thinkpad T460s got a conumer attached with 6.1W consumption as per external meter, while power_now reported 7.8W (already subtracted the idle power, consumed by the system). With a 2.4W external consumer, the sytem reported 2.8W.\nFor the L480: 15.2W (bat reported) vs. 12.8W (measured), and 5.1W (reported) vs. 4.4W (measured). So if we assume the external meter as correct, we see that the internal values were 14-22% off - but always in the same direction.. so maybe we have 14-22% loss when providing power to usb. The raw data is here.\nThe software is out, the metrics seem good enough to be used in comparisons - we might see more consumption comparisons in the future. We might also get new power sensors available, providing more metrics. Let me know if there is something which I should have measured!\nSoftware used for this article: PCP, GIMP, mplayer, vlc, chrome/chromium, Firefox, gnuplot.\nLinks The wiki has further details on pmda-denki. Article Take control of your Red Hat Enterprise Linux systems’ power consumption with pmda-denki explains the steps to measure energy consumption for a workload Article Is a slow but low-consumption system more energy efficient? Let’s measure, with RHEL! uses these steps to compare energy efficiency of various architectures (arm, RISC-V, x86, ..) for a workload ","date":"2021-07-04T13:10:58+02:00","image":"/posts/2021/07/measuring-power-consumption-on-linux2/total_consumption_hardware.png","permalink":"/posts/2021/07/measuring-power-consumption-on-linux2/","title":"Power consumption on Linux, measurements"},{"content":"TL\u0026rsquo;DR With a Linux laptop, you can quite accurately measure the systems current overall power consumption while on battery. This allows to compare if the new version of your software uses more power than the old one. On Intel/AMD systems, power consumption of cpu\u0026rsquo;s can be measured with RAPL - good enough to compare cpu workloads. You can read these metrics directly from sysfs and do some simple computations, or use the PCP/pmda-denki tool. pmda\u0026rsquo;s are agents, enabling the Performance Co-Pilot (PCP) software to archive the metrics for historical reference, and more. From what I have seen, it pays off to choose the software you use wisely: for example, the the applications playing h264 consume different amounts of energy. When you run on battery in a train, that can make a difference. Why measure power consumption? We all use \u0026lsquo;powertop\u0026rsquo; on Linux to see the current power consumption. Since some time, I was wondering about further ways of measuring and recording the power consumption of my Thinkpad, and graphing it. How much power is used when idling, when I\u0026rsquo;m not logged in and running powertop? How accurate can one measure the consumption of single components like NVMe disk from the Linux side?\nIf solid long term measurements were possible, we could answer things like:\nWhich software uses least power to show a video file: firefox, chrome, mplayer or vlc? When we release new versions of a software, we often compare performance with the older version, for example the time required to compute certain things. But comparing power consumption of new vs. old version could have an impact on how long the Thinkpad runs on battery, and understanding which version is more power efficient can have a world wide impact! More illustratively, can I compute how much dead dinosaur (oil) converted to electric energy are used for common tasks? Hardware power meters At first, I got 2 external power meters to measure power on it\u0026rsquo;s way into the Thinkpad. I got the Satechi usb-c meter, which is nice - but I have to read the measurements on the display, they can not be transmitted digitally. To improve that, I got the AVHzY CT-3. The details are here, but essentially: I did not manage to read the measured data from Linux, and the CT-3 seems also for 20V output capped to ~7W, not enough to supply the running + charging Thinkpad.\nSo how to measure then? To store recorded values over time, I use the Performance Co-Pilot software. 2019 I wondered about writing pmda-denki slides, pages 25-27, by then considering to use temperature sensors and system load to estimate the consumption. The idea: look at how much the air heats up, and guess based on that how much energy is used to heat it up. That\u0026rsquo;s possible, but not as accurate as what I ended up with.\nThe newly written agent \u0026lsquo;pmda-denki\u0026rsquo; allows to read various metrics related to power. I use the following sources in the system:\nThe cable on the left connects the power supply (not shown), and the external power meter. Inside the Linux system, we have the first sensor, the metric gets available as denki.bat.power_now. When external power is disconnected, this metric shows the current consumption of the whole system. To the right, we have denki.bat.energy_now_rate. This is computed inside the the pmda-denki software, and tells with which rate the internal battery is charging or discharging. Further to the right, we have the denki.rapl.rate.package-0 metric, abbreviated. RAPL is available on recent Intel chips, and shows how much cpus/gpu/dram together consume at the moment. The other 3 metrics show the single consumption for cpus, the gpu and the memory. The 2 metrics on the left are only available on laptops with battery, and only delivering usable metrics when charging or discharging the system. The four metrics on the right are always available on Intel systems.\nIn the middle, we have further components where we can not directly measure their single consumption: NVMe, internal display if active, fan etc. When the system runs on battery, we have the overall consumption from the metrics on the right, see the readings on the left, and then see how much all in all the other components consume.\nSo, what\u0026rsquo;s this \u0026lsquo;denki\u0026rsquo; term? Denki is the roman character representation of the Japanese term 電気, spelled \u0026lsquo;denki\u0026rsquo;, meaning \u0026rsquo;electricity\u0026rsquo;. Living in Japan, I just always had this in mind for the component collecting electric power metrics.\nAny findings from doing measurements? Detailed measurements are in the second article. System consumption when rendering the great Wipeout webgl implementation: Firefox 20W, Google chrome or chromium use 15W.\nWith PCP for recording the data, and redis/grafana for visualization, one can better understand charging/discharging:\nOn the top we see purely the RAPL readings, the bottom additionally has the consumption which needs the battery. This is from a desktop. At 16:55, power supply is unplugged, so energy_now and power_now start to report values: overall consumption is around 6W, the desktop is idle. At the same time, the pure RAPL values report 1.5W.\nAt 17:00, google meet is started, overall consumption is up to 10W, both energy_now and power_now agree on that. RAPL sees 4W.\nAround 17:43, google meet is stopped, and the power supply is connected. We start charging, which my energy_now implementation reports as negative value - power going into the battery. To the right we see how charging gets slower, until the battery is full.\npower_now is quite directly read from the sysfs. energy_now as it comes from sysfs is the current charge of the battery. My code is constantly reading that and then computing effective charge- and discharge rate.\nConclusion pmda-denki is merged in upstream, and should be part of PCP 5.3.2 (expected end of July 2021). The raw values for RAPL and the battery metrics can also be obtained directly from sysfs, pmda-denki is just making these available more conveniently, and allows to record them over time via PCP/pmlogger.\nThere are powertop (for current readings, and hints for optimization), powerstat (vmstat style reports of battery and rapl readings, I used parts of that code) and rapl-plot, also reporting RAPL. libvariorum is trying to be one library implementing power metrics from multiple sources, like for example NVidia cards.\nWhile the system runs on battery, we can measure overall consumption with usable accuracy, so this is a great way for open source projects to verify if their new version needs more or less power than before! Many projects test for syntax errors, testsuites, and performance impact of changes - but we should pay more attention to power consumption. For fighting the climate crisis, and having the laptop running longer on battery.\nThe second article is here.\n","date":"2021-06-07T14:50:52+02:00","image":"/posts/2021/06/measuring-power-consumption-on-linux/denki_black_on_yellow4b.jpg","permalink":"/posts/2021/06/measuring-power-consumption-on-linux/","title":"An approach to get power consumption metrics right on Linux"},{"content":"What I can compare.. Disclaimer: I have just experienced 2 employers in Germany, and 2 employers in Japan. Also, not only the location/country you work in is relevant for \u0026lsquo;what work feels like\u0026rsquo;, but also the company. My first German employer was a big IT service provider, with many customers in Germany, some more spread over Europe and just few in the rest of the world. My second employer in Germany was an American based company with customers all over the world, that felt already different.\nSo, how is work in .de and .jp different? Starts with the daily routine: in Germany most people come to work, and immediately start to work - or at least try to give the impression of working. In Japan, I saw workers arriving every morning and starting with 30 minutes of newspaper reading.\nTry this in Germany, you get fired immediately ;)\nIn average, Japanese workers are clocking more hours as \u0026lsquo;work hours\u0026rsquo;, but the German GDP (Gross domestic product) per worker is a bit higher than in Japan, so you could say that in Germany more \u0026lsquo;work per hour\u0026rsquo; is done, or \u0026lsquo;work density is higher\u0026rsquo;.\nCommuting Being in the office is much more important in Japan than in Germany. Entering the Corona time, German companies were really slow to adopt remote work - but Japanese companies were even slower. In classic Japanese companies, also working overtime is very important, and there is an extra word for \u0026lsquo;death due to overworking\u0026rsquo;, 過労死.\nJapanese companies are also supporting commuting, they are paying public transport fees for employees.\nWork style How are work styles different? In Germany, there are hierarchies, but in general employees are asked to openly and clearly communicate their opinion on various matters. Of course, the boss has the last word on all matters. In traditional Japanese companies, saying your own opinion is often not requested. My employers in Japan were either Germany or America based companies, so the work style in their branches in Japan were \u0026rsquo;tilted\u0026rsquo; towards German or American style. Still, I saw Japanese superiors making decisions which directly affected me, without asking me for my opinion.\nNormal office routine Quite some things are different. In Japan, you can find shelves with office supplies, and while employees are completely trusted to only use the supplies for company usage, the employees have to write down on paper when a pen or other supplies are taken. If refilling was the goal, then just counting from time to time could also help here.\nIn Japan, I was working in big office spaces, with desks of probably more than hundred people. When a coworker asked me via email to lookup something from his desk, I was walking over, and accessing papers etc. as if it was my own desk. In Germany, that would definitely have rised some eyebrows, but in Japan there was full trust in me fully being entitled to do this - a very nice feeling.\nWhen 2 people are in Germany in front of a screen and one enters a password, the other person is explicitly looking away from the keyboard - to not rise any doubt to spy out the password. In Japan, I have not seen that among coworkers - you are trusted to a level where people are not looking away explicitly. I also saw fewer people lock their screen when they leave their workplace.\nThere are also habits of daily office life which can only be observed in either of the 2 countries. In Japan, I was surprised to see and hear colleagues clipping nails, something unthinkable in Germany. I was delighted to learn these were finger nails. :)\nAfter work life In Germany, employees tend to recruit their friends from sport clubs, the club they are playing instruments in, or such. In Japan, much more is done with coworkers. Going to restaurants and drinking after work is quite common - in classic Japanese companies when you are not asked for your feedback on work related things, giving that feedback at drinking time is important and won\u0026rsquo;t be used against the employee.\nRealations to customers Huge differences. I work in an engineer/support role, so I basically work together with a customer or partner to solve an issue. In German, we can with 2 different ways of phrasing sentences either explicitly express respect to the other party (German: siezen), or rather that we know the other person, and \u0026lsquo;just want to get things done\u0026rsquo; (German: duzen). While I would not be surprised to hear that sales people in German stay on the respect-level, with almost all of my partner and customer contacts I interacted on the daily-interaction-level.\nJust to give English-only speakers an impression what this feels like: in English it\u0026rsquo;s harder to express this, but adressing the other party in each sentence with \u0026lsquo;Sir\u0026rsquo;, or \u0026lsquo;Misses\u0026rsquo; is similar.\nNow, how about Japan? Japanese (language) is laughing about the 2 layers of German, adding multiple layers of \u0026rsquo;lifting up the person you are talking to\u0026rsquo;, and also layers of \u0026rsquo;explicitly lowering your own actions\u0026rsquo; with words. For many verbs, like \u0026rsquo;to read\u0026rsquo; multiple forms exist, to express on which level one considers the other party. So in Japan, customers and partners are addressed in the most respectable ways, lifting their positions and lowering the own position.\nThis might feel like \u0026lsquo;you are on the tips of your toes all the time\u0026rsquo;, and it feels a bit like that indeed. A further aspect is that in Japan, the customer/partner will likely refer no to actions of an individual in your company (\u0026lsquo;you said that [..]\u0026rsquo;), which is direct and accusing, but refer to the company: \u0026lsquo;your company expressed that [..]\u0026rsquo;.\nInteractions with colleagues One interesting aspect of interaction with colleagues in Japan: last names are used. I do not even know the first name of many colleagues from the top of my head. For me with a foreign name things are a bit different, details are here.\nWhen talking with colleagues, then one addresses them with showing respect, for example using -san. When talking with colleagues in front of a customer, then the colleagues are considered like a family, and one refers to them without -san. Of course, the customer is addressed with the highest respect.\nActivities with colleagues, also besides drinking, are much more comming in Japan. For example, before Corona we used to go jogging around the emperors palace in Tokyo after work.\nRegarding communication in departments: in Japan, bosses seem to \u0026lsquo;play it safe\u0026rsquo; and only communicate down the absolute minimum which you need to work. I even started to call this 秘密主義, the idea of \u0026lsquo;declaring things secret by default\u0026rsquo;, and rather making it an exception to pass down things. This approach also has good sides though: when sales or other departments escalate things, they are in Japan likely to not go to me directly but first to the boss. Such escalations are in Japan less likely to reach the simple workers, while such pressure/requests would in Europe simply be reached through, in Japan most of this terminates at the boss, leaving the employee \u0026lsquo;room to breathe\u0026rsquo; and concentrate on solving the actual problem.\nBurocracy/flexibility There is much paperwork in Japan, more than in Germany. When entering the company, I received a big folder with many paper sheets, and details on which people to contact for certain things. Most of this was outdated after 6 monthes, and can be looked up at the intranet - but it\u0026rsquo;s Japanese style to hand these out.\nOn various occasions, coworkers have to deny a request - this is natural. For example, \u0026lsquo;Sorry, our vendor printing business cards does not allow us to print a GPG fingerprint\u0026rsquo;. In Germany, the colleague is then more likely to question that, ask the vendor if change is possible, and the background of this rule. In Japan, colleagues are also very willing to help with creative workarounds (\u0026lsquo;we will print the GPG key in the field for the fax number\u0026rsquo;), but asking for the background or changing such rules seems more out of question than in Europe. Asking for the background of rules, questioning things, is in many cases not welcome.\nSummary So where is working better? Depends on the kind of work you are doing, on your personality and also the company you are working in. For myself, I could not imagine working in a classical Japanese style company - but a mixture with directness from the American mother company renders things working for me. :)\n","date":"2021-05-16T14:58:46+02:00","image":"/posts/2021/05/how-differs-worklife-in-japan-and-germany/20160617_075449_tokyo_morning_trainstation.jpg","permalink":"/posts/2021/05/how-differs-worklife-in-japan-and-germany/","title":"How is working in Japan different from working in Germany?"},{"content":"First I worked and payed taxes in Germany. From 2016 on, in Japan. Time to think about the differences in paying taxes, and how they are used. I dislike doing taxes, would prefer if everything would be deducted automatically.. but taxes are there to stay. The differences between both contries are interesting!\nIncome tax in Germany In Germany, income taxes are automatically deducted from the monthly wages. One a year, one has to do a \u0026lsquo;income tax declaration\u0026rsquo;. Some are diving into these papers and do that by themself, others do it themself supported by software. I decided the most painless was for me to visit a tax adviser once a year, who did the tax declaration for me. Quite convenient, and not very expensive.\nFor work, I moved to Munich in the south of Germany, while keeping my house in Thuringia. This was recognized in my tax declaration, I mostly got money back - so a part of the taxes which had been automatically deducted from wages. Also daily commute to the workplace, or buying things for a home office were noted in the tax declaration for my benefit.\nAs part of my work contract, I had received stock from my company - also that was noted in the tax declaration, I had to pay taxes for receiving stock.\nIn Germany, whatever you declare, you have to bring in papers and prove everything. You might say \u0026lsquo;of course!\u0026rsquo; - but go on and read about Japan..\nDepending on how much income one has in Germany, the tax rate is between 14% and 45%.\nIncome tax in Japan Also in Japan, tax income tax is automatically deducted. In contrast to Germany, you are not asked to do a declaration (確定申告): if you had no further income besides normal wages, and also did not spend money on things which you can declare to your benefit, it\u0026rsquo;s perfectly ok to not do a declaration.\nIn the tax declaration season, public transport has banners with a reminder.\nGood to know: when starting work life in Japan, one does not get notified \u0026rsquo;now is tax declaration time\u0026rsquo;. I work now under Japanese contract for my former employer, so also receive stock - and have to declare it. In 2017 I was aware I would have to do tax declaration, but never got any notice with hints on where to do it. You have only 4 weeks to do taxes, so in the last days of that period I did the declaration.\nAs for way of declaration: it can be done online, but with my poor knowledge of kanji I go to places where workers do the tax declaration with you together. For me, the procedure took 2 to 3 hours each time, one year I also had to go 2 times to the consultation place.\nJapan does a funny thing with taxes: hometown tax (ふるさと納税). Many people move to big cities and work there. When paying taxes, they can choose to dedicate the tax to a different place, for example the place one was born. The area to dedicate the tax to can be chosen freely, and these areas can send gifts to \u0026rsquo;thank\u0026rsquo; for dedicating taxes in this way. This results in web sites where people can compare the gifts of various regions, and use that as criteria to dedicate taxes. Often these are local products: some tax payers end up with a fridge full of meat, others opted into getting face masks which were scarce in the tax-pay-season 2020 when Corona was around.\nIn my case, I declare just 2 things: my normal wages from the company, these taxes are deducted automatically. Ontop, I explicitly have to declare the stock I get from the employer each day. The important part is how much the stock shares were worth on the day I received them, computed in yen.\nAs for papers: every year I appear with full papers at the tax declaration consulting/input place. I also leave the building with all of these papers: just the value of the received stock shares is used, and tax computed ontop of that. Trust is given here by default, no proof papers are handed over.\nOn one side, I feel honoured by that, I like it very much. On the other side, the tax office can request the papers from you in the next 5 years, so you have to store them.\nLike in Germany, Japans tax rate varies depending on income, between 5% and 45%.\nSo what\u0026rsquo;s done with the taxes in both countries? So tax load in Germany is higher than in Japan. What is done with the taxes, where are the differences?\nBasic services are covered on both sides: for example police. But let me count up the annecdotal pieces where I spotted differences.\nSome numbers: Germany has 80 mio residents, Japan has 120 mio. Japans annual GDP (GrossDomesticProduct, value of all produced goods and products in a year) is higher than the GDP of Germany, but breaking it down to citizens, Germany has ~46.000$ GDP per person, while Japan has ~40.000$ GDP per person.\nHow much is spent on military? Germany has spent 1.3% of it\u0026rsquo;s GDP on military in 2019, Japan 0.9%. Here are more numbers.\nEducation: in Germany, schools and universities are basically payed from tax money - and that\u0026rsquo;s the way it should be to give all children the same chances, education wise. In Japan, there are schools run by the country and private ones, for all one has to pay money. Ontop of that, it\u0026rsquo;s rare that children do not get cram school, also that is expensive.\nDealing with Corona, both countries did setup help for businesses. In Japan, 10.0000円 (~780€) were offered to all residents to help them and to stimulate business, so \u0026lsquo;helicopter money\u0026rsquo;.\nA further difference: animal shelters. In Germany, also the ones run by the state are keeping the pets around. In Japan, every year thousands of cats and dogs are \u0026lsquo;disposed\u0026rsquo; instead of spending money on keeping them around. Japan also has many pet shops advertising cats and dogs. There are private organizations who aim at keeping pets a home, they operate from donations.\n2 other taxes Just mentioning here the few interesting to me. Of course, there are taxes on tabaco, alcohol etc, I won\u0026rsquo;t dive into these.\nVAT/Sales tax (Mehrwertsteuer, 消費税) is in Germany 19%, and for some services and products like food 7%. In Japan, VAT is currently at 10%.\nResident tax (住民税): this exists only in Japan. This is computed as 10% of the income, a fixed rate independent of how high the income is. This tax is used by the local area/city, it\u0026rsquo;s payed seperately of normal income tax.\n","date":"2021-02-27T01:30:32+01:00","image":"/posts/2021/02/taxes-in-japan-and-germany/20210227_055428424_tax_declaration_reminder_bus2.jpg","permalink":"/posts/2021/02/taxes-in-japan-and-germany/","title":"Taxes in Japan and Germany"},{"content":"Background: I\u0026rsquo;m a German, living in Japan. Details are here.\nThe visitor visa In March 2008, I went to Japan for the first time. For 3 month I would live and work in Tokyo. As German citizen, or being from other 1st world countries, that is easy: for business as well as holiday you can get a visa for staying up to 3 months in Japan. No application in advance required, simply fill out the paper upon arrival in Japan.\nWork and holiday You have to be between 18 and 30 years old for this visa, I \u0026lsquo;discovered\u0026rsquo; Japan to late to be eligible. As the name says, one can travel Japan, and work. Depending on the home country, the numbers might be limited.\nThe 5 year work visa In 2016, I got a work contract in Japan. One can apply for 3 or 5 year visa based on a variety of reasons: for example doing arts, learning about culture or martial arts, for working as missionary to \u0026rsquo;teach\u0026rsquo; religion - I met some of these guys, or for working as a specialist - that applies to me. The application procedure is complicated, my employer payed a contract company to do it. After filing the request, I left my passport (travel passport) for some days at the Japanese embassy in Munich, and got it back with a special imprint/label. When I entered Japan, based on that label I received a \u0026lsquo;residence card\u0026rsquo; (在留カード, literally \u0026ldquo;residence card\u0026rdquo;), registered myself as citizen at the town hall, and the visa procedure was complete. That card was called \u0026lsquo;alien registration card\u0026rsquo; in the past, an interesting term. Japanese do not have a passport, but the \u0026rsquo;travel passport\u0026rsquo; or the drivers license is often used for identification. Also a citizen identification number exists, and also I got one as I pay taxes in Japan.\nJapan has 120 million citizens, 2% of these are foreign residents.\nWork visa extension Now my 5 year visa is about to expire, and I have applied for an extension by 5 years. The extension involves preparation of papers, including proof that I earned money in Japan, that I am employed and so on. My employer kindly payed a contractor again for doing the application. Such a service costs around $750. I was asked to leave my residence card and my German passport at the contractor, to do the application work for me - I opted into going directly to the Immigration office and perform the procedure. I handed over the application papers, residence card and passport - these papers I got back after some minutes. Processing is expected to last 4 or 6 weeks, and if all goes well this should result in a new visa for 5 years, and a new residence card.\nGoing by bike to the immigration office to apply for the 5-year-visa extension, I saw this painting of a forward looking person with a shamizen. Setting the right spirit for the trip! :)\nPermanent residency With \u0026lsquo;permanent residency\u0026rsquo; (PR), one can stay without time limit and just has to renew the \u0026lsquo;residence card\u0026rsquo; from time to time. The right to vote though is not part of PR, one has to become Japanese citizen for that. In the first months of Corona, it got very hard to enter Japan: Japanese citizen were allowed, but also owners of a PR could not enter Japan.\nThe Japanese term for PR is 永住権: 永 is for \u0026rsquo;eternity\u0026rsquo;, 住 means \u0026rsquo;living\u0026rsquo;, and 権 is the \u0026lsquo;right to do something\u0026rsquo;. So literally, the \u0026lsquo;right to live until eternity\u0026rsquo;. Kanji are beautiful, right? :)\nBeing married to a Japanese citizen makes it easy to obtain permanent residency, but the normal procedure is to apply after living for \u0026gt;10 years in Japan. With further regulations, one can file for PR also earlier, that is what I did after living for 4.5 years in Japan. Following a set of rules (details here), one can obtain \u0026lsquo;points\u0026rsquo; for things like\na certain level of proficiency in Japanese (JLPT 1 or 2) working for many years in ones profession receiving high wages having a degree being young If an applicant can prove such points, enough points can accumulate to file for PR.\nIf a document is not in Japanese, a Japanese translation has to be provided. The documents required for PR seem intended to demonstrate that the applicant has been a good citizen:\nproof to be a registered citizen, from the town hall proof to have payed all taxes properly, tax office proof to \u0026lsquo;have money\u0026rsquo;, so not having to rely on the Japanese welfare Also, a guarantor is needed: a person who would give you guidance, and sponser a ticket back to your home country if things to not work out. This point caused me quite some headache, I did not want to ask people to do something like this.\nAn other point: the \u0026lsquo;reason letter\u0026rsquo; (理由書), a description of why you request PR. I was puzzled on what to write here.. but once I started, the page filled up quickly. I wrote about my interactions with local communities for learning Japanese, presenting about socialist East Germany, about singing Beethovens 9th in the chorus of 5000 people in Tokyo.\nFiling the PR turned into a project which took many weeks, my application consisted of 35 pages. I filed for PR in October 2020, and got it approved in March 2021. 4 months after the initial request, 2 further documents were requested. Reading the second reference was quite a challenge:\nThis turned out as 2021年の見込み収入, so \u0026ldquo;statement which earnings are expected for 2021\u0026rdquo;. There are also specialized companies which offer filing PR for $1000, but I wanted to do this on my own.\nThe PR consists of an \u0026lsquo;application paper\u0026rsquo; with fields to fill in, the proofs to provide and the translations. The \u0026lsquo;application paper\u0026rsquo; has to be filled out with Kanji and Romaji characters. Having me fill that out by hand would lead to hard to read results. The website provides the paper as pdf and Microsoft Excel file. The pdf file can only be used for printing. I do not own a copy of Microsoft Excel or Windows, and when opening the excel file in LibreOffice, Romaji and Kanji characters get badly misaligned. To work around this. I imported the pdf into Inkscape (an open-source vector graphics editor), and entered the text ontop - this worked flawlessly.\n@polm23\u0026rsquo;s site has further details on PR.\nBeyond PR Becoming a Japanese citizen, one\ncan vote can quite easily travel to most countries (that applies also to German citizen) has best chances to return to Japan in times of Corona. Important if one has property and family in Japan. The downside: Japan insists on citizen having just a single citizenship - so becoming Japanese one has to discard any other citizenship. Also Germany is increasingly averse of dual citizenship. The thinking of the politicians seems to be \u0026ldquo;people should commit/decide for one country\u0026rdquo;. I suspect they never had the experience of learning a different country, language and culture to a level that one would also like to become citizen, be able to vote and get voted. Seems like for example Australia is offering multiple citizenships.\nturning-japanese.info has details on naturalization/becoming Japanese.\n","date":"2021-02-11T07:15:09+01:00","image":"/posts/2021/02/japan-visa/20210209_221733610_filing_5year_visa_extent_spirit3.jpg","permalink":"/posts/2021/02/japan-visa/","title":"Visa and permanent residency in Japan"},{"content":"So, having seen Norbert and @codewiz mentioning Plasma frequently, I decided to give it a try. My desktop box is Fedora33, which comes with a quite current version of Plasma.\nSome years ago, I tried gnome - but Gnomes \u0026ldquo;let\u0026rsquo;s make it as simple as possible, leave buttons out\u0026rdquo;-strategy does not stick well with me. I don\u0026rsquo;t want a dumbed down window manager.\nI use WindowMaker since dozends of years. It\u0026rsquo;s nice and lean, but does not seem to have a future with Wayland. I use xdm with an old style ~/.xsession file.\nFirst step for Plasma: dnf groupinstall kde-desktop-environment. 1.6GB of installed files, and seeing so much stuff I don\u0026rsquo; intend to use makes me shiver.. but well, worst case I can remove the packages afterwards. I currently run \u0026rsquo;exec wmaker\u0026rsquo; in my .xsession. Plasma inherited from KDE, so that becomes \u0026rsquo;exec /bin/startkde\u0026rsquo; now for me.\nAlso, cleanup time. I am sometimes using my Thinkpad on the road, so only the internal display. At home, my monitor is connected via HDMI, and then I disable the internal screen. Until now, I have some bash code in .xsession which uses \u0026lsquo;xrandr\u0026rsquo; to do that detection, and disable the internal screen if the monitor is connected. Commenting that out, as Plasma will do that.\nUntil now, I also start pulseaudio, the ibus framework for switching to Japanese input, and run redshift for making colours softer in the evening - all that gets disabled as it\u0026rsquo;s implemented in Plasma - needs to be configured there though.\nThen logout from WindowMaker, logging in with xdm, and Plasma is running. Preferring the Thinkpad screen by default, but that is configured quickly. Let\u0026rsquo;s try logout/login.. urgs, when logging out from Plasma the desktop gets stuck, and xdm does not appear. After a reboot, the internal screen is directly disabled, good.\nSwitching global theme to \u0026lsquo;Breeze Dark\u0026rsquo;, and Plasma style to \u0026lsquo;Oxygen\u0026rsquo; fits my colour needs better.\nOn to configuring. Ok, so the \u0026lsquo;virtual screens\u0026rsquo; I heavily use are \u0026lsquo;Desktops\u0026rsquo; here. I create 8, and configure the 2 keys above the Thinkpad cursor keys to switch between the Desktops. Switching to ibus-kkc for Japanese input works directly - nice!\nSince forever I use gkrellm on one of my \u0026lsquo;Desktops\u0026rsquo; to keep an overview on how busy my system is, network throughput and so on, using that also now. After starting it, I see some load on the system, and I/O. Turns out indexing for desktop search was running, process \u0026lsquo;baloo_file_extr\u0026rsquo;. Disabling this at System Settings/Search quiets the system.\nChanging between applications on one Desktop works with \u0026ldquo;alt\u0026rdquo; + \u0026ldquo;tab\u0026rdquo; as usual, good. Dragging an application window to the left or right of the screen is not moving it to the next/previous Desktop.. that\u0026rsquo;s what I\u0026rsquo;m used from WindowMaker. TODO, search how to move apps to other Desktops. I know right moust button click on the title bar has \u0026lsquo;move to desktop\u0026rsquo;, but there might be something simpler.\nNow for terminals. Since years I use xterm with white on black fonts, let\u0026rsquo;s try Konsole. Default font is smaller than my previous config.. but seems usable. Just my mutt, irssi and vim colour schemes are not very usable with the default \u0026ldquo;gray font on grayish background\u0026rdquo;, so getting a customized profile with real black background. Ah, much better! On-the-fly changing of the font size is a nice feature.\nSo, I read an email from a Japanese friend with mutt, mark a Japanese sentence and want to paste it into https://www.deepl.com/translator for translation with the middle-button.. that did somehow not work in my first Plasma session, but since the second.\nNotifications: by default, these were collected and showed in the panel - I\u0026rsquo;m used to see direct popups from Firefox for mastodon actions, that could be configured easily. Notification sounds were off by, I would have preferred \u0026lsquo;on\u0026rsquo; per default.\nA pain point from WindowMaker also exists on Plasma: in firefox, after opening https://9gag.com , when unmuting sound on videos, the sound is to loud. Adjusting that in settings is just adjusting the sound for that single video, and the next one again plays at to high volume. There is a generic setting for firefox in \u0026ldquo;audio -\u0026gt; applications\u0026rdquo;, but it has apparently no influence.\nTODO: I use multiple terminal windows, not a single one. I have \u0026lsquo;Konsole\u0026rsquo; pinned in the panel, I would like to start new instances when clicking it, but it is just minimizing the running Konsole instance.\nOther cool things: I use a dark Plasma theme, firefox title bar is now also dark, and also websites seem to show a dark theme. Nice!\nComments? Please here on Mastodon, or on twitter\n","date":"2020-12-30T00:32:52+01:00","image":"/posts/2020/12/plasma/20201230_plasma.jpg","permalink":"/posts/2020/12/plasma/","title":"The Plasma desktop for a newbie"},{"content":"Onsen? Ryokan? Japan being on the intersection of seismic plates is leading to earthquakes, but also to hot water which is used in bathes, so called Onsen. A Ryokan is a traditional style of hotel in Japan, with tatami mat covered rooms, often a bath attached, including good meals and services like preparing futon beds in the evening for visitors. Japanese sometimes go to Onsenmeguri, which means \u0026rsquo;travel with stay in an onsen'.\nWith infection rates in Europe rising, it became clear that I can not spend this christmas in Germany - I decided for a trip to Japans Northeast, including an area known for Onsen.\nNyuto onsen area I decided to visit the Nyuto onsen area (乳頭温泉郷) in the north- east, in the first week of December. First thought was to stay in a traditional Ryokan, the area has 7 of these, all with an own onsen attached. For that season, some of these were already closed and others would charge an extra \u0026lsquo;heating fee\u0026rsquo; - which reminded me that traditional rooms would not be warm at this time - Touhoku is colder than Tokyo where I live. I booked 2 nights in a hotel with normal rooms with air conditioner, but which also would have the meals of ryokan, and an outdoor bath. To have a look at traditional ryokan, I planned to walk over to a nearby, classic ryokan/onsen.\nGetting to the hotel A bus goes in 45min from the Tazawako train station to the \u0026lsquo;Komogatake grand hotel\u0026rsquo;.\nI hiked to the hotel by foot. Google maps knows about the bus route, including all bus stops in between. The next bus would run in one hour, I did not want to wait and walked with my 2 backpacks along the route of the bus. As I wanted to see lake Tanzawako, I missed the bus 1 hour later. I walked further towards the hotel, and tried to catch the next bus which would according to google maps run another hour later. That exact bus connection did not run as I learned at the bus station, so I again walked towards the hotel. The next bus, an other hour later, was running - but I decided to walk the last 30min, total 4h walk. Androids OSmand app has the best walking directions, it also considers elevation.\nThe western style hotel Since the night before it snowed in the area. The room was western style and warm, using the hotels outdoor bath while the snow was falling was great. The other services were like at a ryokan: yukata clothes were provided for guests, a set for preparing green tea at the room, fantastic dinner and breakfast.\nTsuru no yu - traditional onsen On the next day I went to see and enter a traditional onsen: the Tsuru no yu (鶴の湯). There was quite some snow, for the 1 hour walk I just used streets - there was not much traffic.\nAt the onsen, I had a chance to see the rooms. Nice traditional rooms with tatami mats, a fire place. For me, the western style hotel was the right stay in this season.\nThen I payed a fee and used the bath. All the buildings covered with snow, and again snow falling while bathing was great. There are multiple bathes which can be used, I used one outdoor with water which seemed to be white, like milk. Taking pictures was forbidden, but you can get an impression here.\nTake aways When \u0026ldquo;Onsen meguri\u0026rdquo; (onsen travel) is mentioned in Japan, people think of a stay at a hotel with onsen. Taking it slowly, bathing multiple times a day, reading, enjoying good food and talking with others. Also the visitors I met at the hotel did not plan any trips while they stayed there, this was meant for recovery. With the cold snow outside, the hotel even had something of a refugium. Greatly recommended! Also the traditional ryokan, but consider the season of the stay.\n","date":"2020-12-07T13:12:56+01:00","image":"/posts/2020/12/onsen-hotel/20201204_020832139_rooms_outside.jpg","permalink":"/posts/2020/12/onsen-hotel/","title":"What are Japanese Onsen and Ryokan like? What's onsen travel?"},{"content":"There is great food in Japan, let me share a week of autumn food in Japan. I do not eat much for breakfast, but for lunch I visit various places.\nKeema-Nattou Curry Fridays Keema-Nattou curry with rice. Keema is minced meet, Nattou is a speciality from Japan often eaten for breakfast, these are fermented soybeans. Fermented means \u0026lsquo;has been changed\u0026rsquo;, like sauerkraut in Germany, or Kimuchi in Korea. Also in the curry: carrot, potato, pepper, okura (a vegetable not known in Germany), broccoli. 2 cooked eggs of unspecified origin, not from chicken.\nNabewariyamas Udon On Saturday I went hiking to Nabewariyama in the Tanzawa area, trip details. A hut is at the summit, and fantastic udon is available. Udon a kind of thick, wheat-flour noodle which is common in Japan. Here along with egg, pumpkin, mushroom and vegetable. Very tasty, and sold here at the top of a 1270m mountain, for 1500円/12€.\nSanma Sanma, a kind of makarel which is typical to be caught in autumn. I like fish, but I\u0026rsquo;m very bad as dissecting fish with bones just big enough so one can not eat them. Also on the plate: rice, Miso soup, Toufu, vegetable, radish and soy sauce. Miso is the base ingredient of Miso soup, it is made from fermented vegetables.\nAutumn Soba On Monday, we had a holiday and I went to a bath, an Onsen. In the attached restaurant, this audumn soba was available. Soba is another kind of noodle, made from buckwheat flour. Ontop shrimp and vegetables. Along with this I ordered this special kind of mochi, which are made from rice with miso based souce ontop.\nKajiki fish I work mostly from the apartment nowadays, but once a week I cycle to Ebisu where the office is, and enjoy the food at Totoya. They have Sashimi (raw fish), but also cooded fish like that day, or eel. Reading the menu is still challenging.\nKao rapu gai This is Kao-rapu-gai (カオ ラープ ガイ), at a Taiwanese restaurant in Shimokita: spicy minced chicken with salad, egg, roasted rice powder, and dessert.\nChirashi Sushi Once per week I go to this Sushi place, this time I enjoyed \u0026lsquo;chirashi sushi\u0026rsquo;, so \u0026lsquo;sushi things spread over rice\u0026rsquo;. Along with green tea, miso soup, vegetable. On the right, that is no pudding but a \u0026lsquo;royale\u0026rsquo; (German: Eierstich) made from egg and fish. A piece of honey melon as dessert.\nSummary I hope you agree now at least in that food in Japan looks delicious :)\nFood in Japan is just amazing. The most expensive of these meals were around 1700円/14€.\n","date":"2020-11-27T00:10:49+01:00","image":"/posts/2020/11/autumn-week-of-food-in-japan/20201120_025541883_kema_curry_coconut_soup.jpg","permalink":"/posts/2020/11/autumn-week-of-food-in-japan/","title":"A week of autumn food in Japan"},{"content":"Intro I get weekly Japanese lessons via Skype. At the moment we are reading Japanese news articles: discussing grammar and rare words, and the topic itself. After 5 years, my teacher knows me well enough to select topics which are interesting to me. Recently we read the article Is it hard to live in Japan?.\nSeems like my teacher shortened the article to 5 points - after reading the original I find the shorter version more to the point. I will quote the authors points, and compare with Germany and other countries.\n1: Society enforces conformity I can fully agree on this one. \u0026ldquo;The nail that sticks out is hammered down\u0026rdquo; - this is well known for Japan. If I do not properly separate my trash, chances are that a neighbor will notice and comment on that. Everything which sticks out is suspicious: also a child in school which turns out as a genius will have rough times, just for sticking out. Of course, this also has good sides: in Corona times, even if you question the usefulness of wearing a mask: you wear one to not stick out. For wearing a mask, this did not come with any down sides and turned out very useful in preventing the virus from spreading.\nYes, social pressure in Japan is stronger than in Germany.\n2: Seniors first In Japanese companies, the word of seniors has more weight than that of young employees. In Germany there is just a tendency to do so, but this is strong on traditional Japanese companies.\nI work at the Japanese dependence of an American company, and also here older employees can enter the company directly at a high job level, whereas in Europe they would start lower and first have to prove themself. This is in line with the general flow of things in Japan: women giving birth in general stop working, so the family relies on the income of the husband. When he is changing the employer, having to start at a lower wage level would be a severe hit to the family.\nFor young people, having to live with \u0026rsquo;the seniors are always right' can be tough, a valid point for \u0026lsquo;Japanese life is hard\u0026rsquo;.\n3: Decline of personal relations With this point, the author points at neighbors barely knowing each other, and Japanese entering their worklife basically just attending 2 communities: colleagues at the workplace and their family.\nThis is a valid point, quite similar to Germany. There is a tendency to more contact between people on the countryside, also in both countries. In Japan, with such a high density of people in the big cities, staying anonymous seems like \u0026lsquo;self protection\u0026rsquo;.\nThere might be more personal relations in countries like Italy or Korea, just judging from the personality of Italians and Koreans.\n4: The \u0026lsquo;work above all\u0026rsquo; principle Word has it that Japanese put work above everything else. I think this is a valid point, but is slowly changing - younger Japanese seem to put more emphasis on family.\nFor Germans, in general the family has top priority. I would probably not count this point as \u0026rsquo;living in Japan is hard\u0026rsquo;, after all the priorities are a personal matter and the preferences of others do not directly affect me, they do not make my life harder.\n5: Constantly thinking about money With this, the author of the article refers to the falling birthrates and overaging population in Japan, so private saving of money for retirement being required. It is said that one should aim at 200.000$ of savings for retirement. Having to constantly think of money after retirement is a huge burden according to the article author, making life harder.\nI agree with that point, Germany is in the same situation although not to the same extent because of immigrants - Japan has almost no immigrants.\nDeclining birth rate is seen in all developed countries, to escape this one could for example go to Africa. Birth rates are high there as people consider children as their \u0026lsquo;retirement strategy\u0026rsquo;.\nFurther relevant points The author of the article has apparently not lived abroad, that would allow more comparisons with other countries.\nFor me, comparing life between countries seems easier than saying \u0026rsquo;life in Japan is hard\u0026rsquo;. For that, I would consider additional criteria.\nFor example: how safe do I feel in a country? I can safely have a walk in the evening in all places in Germany and in Japan, in America I would feel unsafe in many places.\nConsidering military security: Germany as well as Japan have no attack capabilities, and protect themself with \u0026lsquo;military agreements\u0026rsquo; from perceived threats. Germany in the EU, Japan with Australia and the USA, effective safety seems to be similar.\nHow progressive is the country in fighting climate crisis? Germany is doing better than Japan there, but both are still better than America. For example Norway seems to be better than Germany.\nCan you agree with the countries laws? For example, the death penalty exists in Japan, that\u0026rsquo;s a point for Germany.\nIn terms of free speech, both countries fare well.\nSummary Everybody would call out different points in such comparisons, for me both Germany and Japan are good countries to live in. :)\n","date":"2020-11-24T14:26:28+01:00","image":"/posts/2020/11/japan-hard-to-live/20170628_183010_osaka_water_cityview.jpg","permalink":"/posts/2020/11/japan-hard-to-live/","title":"Is it harder to live in Japan than in Germany?"},{"content":"The previous article introduces into HDMI grabbers on Linux.\nWhen is latency relevant? My first use case was to play a game on a device like the Nintendo Switch, with the HDMI output only visible via grabber. While playing with various options and software, I needed a way to nicely measure the latency between the HDMI signal output, and the picture displayed on my screen. How to do that? Best is to show the current time, sufficiently accurate, on both\nshown by the HDMI source and shown by your displaying system. With latency in seconds range, you can then directly read the latency.\nWhen coming down to centiseconds, you need a screenshot: So with this, we can optimize for low latency when displaying locally.\nHow \u0026lsquo;smooth\u0026rsquo; is it? That is a different goal than \u0026rsquo;low latency\u0026rsquo; of course. For this, you can allocate bandwidth, user a buffer on the receiver side, and get nice results. For comparison I opened wipeout webgl on both the system providing the HDMI stream and the local system in parallel.\nEasy streaming The wiki pages have various approaches for streaming to single systems, but how to do that at scale, for multiple clients? owncast is really nice for this, OpenSource.\nOur streaming setup has these components:\na system on the internet, where we will run Owncast. Should not be behind a NAT, so incoming packets make it to the system. ffmpeg 4.1.5 or later is required. I use a KVM system running Debian 10, which comes with ffmpeg 4.1.6. a system where the usb-grabber is connected to, Fedora 33 for me.\nffmpeg needs to be installed. the HDMI source First, we setup our system with good internet connectivity, acting as the server.\nmkdir /opt/soft/owncast-0.0.2-linux-64bit cd /opt/soft/owncast-0.0.2-linux-64bit wget \\ https://github.com/owncast/owncast/releases/download/v0.0.2/owncast-0.0.2-linux-64bit.zip unzip owncast-0.0.2-linux-64bit.zip vi config.yaml # customize at least the streamingKey ./owncast Next, we configure our streaming system. It has the USB-grabber connected and will send the stream to the server we just configured. This assumes /dev/video2 as video device, and that hw:1,0 is the alsa device from the grabber.\nffmpeg -f alsa -ac 2 -i hw:1,0 -thread_queue_size 64 \\ -f v4l2 -framerate 60 -video_size 1280x720 -input_format yuyv422 -i /dev/video2 \\ -c:v libx264 -preset veryfast -b:v 1984k -maxrate 1984k -bufsize 3968k \\ -vf \u0026#34;format=yuv420p\u0026#34; -g 60 -c:a aac -b:a 128k -ar 44100 \\ -f flv rtmp://\u0026lt;ip-of-your-server\u0026gt;/live/\u0026lt;your-streaming-key\u0026gt; With that, we can have browsers on clients access the owncast instance:\n\u0026lt;server-ip\u0026gt;:8080 If that works, you should next read up on customizing the owncast instance and consider to tune the ffmpeg options of your grabber-system.\nUse cases and good approaches Use case You might want to look into.. Low latency display locally mplayer Streaming over LAN, single viewer µStreamer Recording to file 6by9 yavta, µStreamer Stream over Internet, many clients Owncast, ffmpeg Remote hands solution TinyPilot ","date":"2020-11-13T13:05:47+01:00","image":"/posts/2020/11/linux-hdmi-grabber-2/hdmi_grab_wipeod.jpg","permalink":"/posts/2020/11/linux-hdmi-grabber-2/","title":"HDMI grabbing: latency and streaming"},{"content":"HDMI grabbing? These are devices you connect you a computer (might be a tiny one like a Raspberry Pi), and which convert HDMI input from a cable into something useful for further processing. I\u0026rsquo;m sharing here details from my experiments with 2 grab devices, and using Linux for processing.\nSo practically, what can I do with this? My usecase is that I have just a single monitor, with just a single HDMI connector. I saw that HDMI grab devices came down to 1200円(~10€), I wondered how useful they were to grab/display/record games running on the Nintendo Switch.\nAlong my journey, I met many more use cases:\nRemote hands: Using a Raspi with HDMI-grabber, to control a nearby system: you can connect the Raspi also via serial connection, and via USB you can have the Raspi emulate USB-keyboard and maybe mouse. TinyPilot is such a project. Streaming:: grabbing HDMI from game consoles, one can stream to the internet, for example to Twitch. Grabbing the HDMI signal of your reflex camera and using it for video conferencing, you can get better quality than from the buildin camera of your laptop. Recording/processing: Digitilize videotapes in playing then on something with HDMI output, and record that. Also, retro game emulators now got OCR features which read the screen and translate to different languages. You could do this also with you HDMI stream, so for movies and so on. The Hardware My 2 devices:\nThe H2C-RPI-B01, an addon board for the Raspi. With the Toshiba TC358743, it can directly encode to x264. I was not able to grab audio with this one. Price tag: ~1500円/12€. A USB-grabber, 1190円/10€ from Aliexpress. I have no more specific name for it, it\u0026rsquo;s available from multiple vendors. usb id 534d:2109. Setup and Limitations Both grabbers, with 2 HDMI sources, and a system receiving the grabbed data:\nMIPI CSI-2: protocol for point-to-point transmissions. Ribbon cable, for normal Raspi 2 data lanes, each up to 1Gbit/s. The H2C-RPI-B01 can encode to x264 directly, putting less traffic onto the CSI-2 cable - but driving latency up. Raspi4 connectors: Raspi4 has GBit ethernet (800mbit/sec at 0.3ms roundtrip time), and USB3 via usb-c connector (180mbit/sec at 0.3ms roundtrip time) details are here. H2C-RPI-B01 board experiences This was my first device, connected to the Raspi4 via ribbon cable, also Raspi cameras get connected like this. The device is cheap, but took many weeks to arrive from China. Getting started is complicated by the 2 modes of operation which are possible: firmware- and kernelbased. Kernel based operation is newer/more promising, it requires to manually load the EDID data with the modes which the board offers to the HDMI source.\nGrabbing is limited to 1920×1080@25Hz - which not all HDMI sources support. Also when using a Linux system as HDMI source, you might have to define an own display mode with \u0026lsquo;xrandr\u0026rsquo;, details to this and the other H2C-RPI-B01 details are in the wiki.\nThe Toshiba chip on the board can do x264 encoding, that keeps the load on the Raspi cpus down. In that combination, latency goes up though.\nSeveral combinations for encoding, streaming and encoding are possible. I played with ffmpeg/ffplayer, mencoder/mplayer, vlc and yavta. First one should be clear about the goal: streaming? recording? Is latency relevant? Is it ok to have artifacts in the picture? The limiting resources are bandwidth between the single components which are involved (see graphic above)\u0026rsquo;, resolution/screen frequency of the HDMI source. For me, biggest limitation of the H2C-RPI-B01 was missing audio: it seems like one has to solder extra pins on the board, and connect them to the Raspi. Me and others also then got no sound, with amplification or such it might be doable - details are in in the wiki.\nFor recording to file, yavta proved useful, for streaming ustreamer.\nffplay (from ffmpeg) or a webbrowser can be used as client.\nBe sure to have a proper power source, this setup seems to need more than you get via usb-c\nUSB-grabber experiences This was my second device - to also get audio. The software stacks and experiences from the H2C-RPI-B01 mostly apply again, but this device can not directly encode to x264. Thanks to a patch from @marcan42, also stereo works now. Same restriction to 1920x1080@25Hz here, and USB can here become bottleneck. usbtop can be used to monitor. This device is now so mainstream, also engadget reported - but for Wintendo only.\nLatency wise, ~120ms are doable with the device. Further details regarding config are in the wiki.\nSummary I got the setup I was aiming at implemented, but latency turned out to high for playing. For many other use cases, the grabber devices run quite well with Linux. The config details for the devices are here and here.\nMany HDMI sources like the Nintendo Switch can also provide 720p@60Hz, which both devices can deal with. If you need more than 1920x1080@25Hz, watch out for other devices.\nPotential alternatives I haven\u0026rsquo;t yet played:\n4k grabber device, @30Hz: aliexpress 1080p@60Hz: aliexpress, driver is now in upstream kernels USB3/4k/1080p@60Hz: aliexpress (via marcan) If you have a Raspi, do not need audio and want to encode to x264 then the H2C-RPI-B01 can do that directly. For pure encoding the NVidia Jetson Nano 2GB ($60) has some impressive specs, encoding 4kp@30 live.\nI keep a list of alternatives here. The next article has details regarding streaming and measuring latency.\n","date":"2020-11-07T02:57:52+01:00","image":"/posts/2020/11/linux-hdmi-grabber-1/raspi_switch2_zelda_menu.jpg","permalink":"/posts/2020/11/linux-hdmi-grabber-1/","title":"Linux and HDMI grabbing in 2020"},{"content":"Climbing Mount Fuji at Night: Ascend covers the ascend.\nFor 2 hours I was waiting in the cold at the summit, drinking green tea from the vending machines. Due to the cold, I could not sleep or stay at one place without moving - and that was even with proper jacket, I had seen people climbing up in short trousers, probably not waiting for the daylight but directly descending.\nEventually a hut opened, people streamed in and ate udon. At 4:15, the first light could be seen, people started to stare towards east. Some wrapped the body in tinfoil, curious how well that works. Looking downwards, I saw the clouds, this time not seen from a plane but the Fujisan. At 5:02, the sun finally came up, and the Japanese national anthem was played.\nEverybody was happy to have made it so far. Shops opened and sold souvenirs. Pilgrims were around with cone shaped hats, for some it is considered good to have climbed Fujisan at least once in a lifetime. It got more and more crowded now, climbers arrived.\nTurning around, a crater was seen, rocks and snow. There was a shrine which I did not visit.\nEventually I started to descend. I had not slept and figured I should pay special attention to not slip, so taking time would be good. Some people used oxygen masks, as the air at this height gets thin.\nThe views while descending were amazing. First pure rocks. Then red soil, like on mars! All of that in front of the clouds from the top. The Fujisan is just open for climbing for 2 months in the summer, at the other times the stations are closed and hikers bring in tents and camp.\nI considered to take the long way down, but decided to just go with the shorter one to not block ascending people. Eventually the scenery got greener, first plants were seen. The wooden pole was really helpful, without that I would have needed to focus much more on each single step. After all of these hours, focusing got harder and harder.\nAt 8:00 in the morning, I reached the 5th station again. Buses were not yet running, so I took trains back to Tokyo. I stayed up that day until 18:00, the next day I went to the Odeo onsen, and spent there all day bathing and sorting the pictures.\nMore pictures\n","date":"2020-11-01T00:06:17+01:00","image":"/posts/2020/11/fujisan-daylight/20140806_221903_fujisan_torii_morning.jpg","permalink":"/posts/2020/11/fujisan-daylight/","title":"Climbing Mount Fuji: The Daylight"},{"content":"In many countries, climbing the highest mountain is just for professionals, or for expeditions. Mount Fuji in Japan has with it\u0026rsquo;s 3776m just the right height to be a bit of a challenge, but being doable for an average trained person. One starts with looking up reports from others and wonders: what should I take with me? Is my condition good enough to do it? Do I need maps and so on?\nIn August, with 39°C in Tokyo, I planned to catch an early bus to Mount Fuji, go up, descend and go back to Tokyo. The early buses for the next day were already gone, so I settled with a ticket for the last bus of that day, and climbing up at night. I got myself a headlight, and got onto the bus. I wore long trousers, in my backpack I had a winter jacket, winter hat, gloves which I use for cycling in winter, something to drink and a few onigiri - Japanese rice balls.\nThere are 4 main pathes to the Fujisan summit, I used the Yoshida trail. The bus brought me to the \u0026lsquo;Fuji Subaru Line 5th Station\u0026rsquo;, ~2300m above sea level. First time I saw something like this, it has a bit of a feeling of a base camp for expeditions. One can rest here, buy equipment or supplies - I bought a wooden hiking pole, recommended for stabilizing while walking.\nAt 19:22, I started to climb up, feeling a strange mixture of excitement and not knowing how things would go. From 5th station to the peak it\u0026rsquo;s just 6km distance wise, but 1400m height difference.\nClimbers continued to come down from the Fuji. Someone said \u0026lsquo;諦めた方がいいよ！\u0026rsquo; (\u0026lsquo;It is wise to give up!\u0026rsquo;). Made me think for a moment.. but I got no facts to decide on. I can anyway also turn around later when I think it makes sense. Around 8:30pm it gets cold, I put on the winter jacket.\nLooking towards the peak, lights could be seen. The stations are numbered, so after 5th station I expected to reach the 6th station. I did not use a map - the trail is marked really great. Android app OsmAnd is using OpenStreetMap data and can work offline, but I did not need it.\nWith the night starting, it also got silent - no traffic or industry noise, just silence. The Japanese word for peace, 平和/へいわ/heiwa came to mind. I used the headlight for some time, but turned it off as the moonlight was sufficient. The silence got interrupted by either fireworks or an army targeting exercise, very nice to see from here. At 21:16 I met the last descending climber, a great opportunity to rest and have a chat. I\u0026rsquo;m still full of this uncertainty, \u0026ldquo;will I safely make it to the top?\u0026rdquo;.\nI reached a point with a map, and a riddle got solved. I had not reached any stations for some time - had I missed them? Or was I slower than I had expected? The Yoshida-path has actually 2 pathes: one with better views and mostly used to climbing up, the other a bit shorter and used for descending - I had walked up that one. A bit later I arrived at the 8th station, 3400m. Climbing time was 3 hours.\nMy mood improved vastly: now the peak was not far any more. It was to near actually: of course I could go up, and then immediately descend again without staying. But seeing the first light of the day is an event one wants to catch, and in my situation it became clear that there were many hours left until the morning. It had become very cold at the 8th station, and the summit was even colder. I talked with a few people, many stayed over night in the station, planning to climb up in time to see the first light of the day. Many of them were so excited they could not sleep. One climber was above 70, and wearing a red hat!\nEventually I just could not wait any more and climbed up the last part. Along the way, just before reaching the summit, the same figures which are seen at Japanese temples were setup, and Torii.\nAt the summit, already many people had gathered. Some in sleeping bags, it was very cold, especially when not moving around. To the rescue, there were vending machines with hot tea: 600ml for cold tea for 500円 (3.8€).\nWill do a second article with the sunlight and descend. More pictures are here.\nSecond part, The Daylight\n","date":"2020-10-31T13:50:26+01:00","image":"/posts/2020/10/fujisan-ascend/20140806_120229_fujisan_5th_station.jpg","permalink":"/posts/2020/10/fujisan-ascend/","title":"Climbing Mount Fuji at Night: Ascend"},{"content":"First things first: there will not be a clear winner, but let\u0026rsquo;s try to look at this from various angles. As a German living in Japan, I am tied into people and media in both countries, let\u0026rsquo;s try a comparison.\nInfection numbers Numbers of daily new infected people is considered one of the most interesting criteria. Being near to China, Japan had first infections quite early, but Germany was catching up in March. As of October, Germany has four times as many infections as Japan. In Germany, 6 times as many people are reported \u0026lsquo;currently sick\u0026rsquo; as in Japan.\nI think the biggest factor to consider here is the number of Corona tests: someone is just confirmed \u0026lsquo;infected\u0026rsquo; after a positive test. As of October, Germany has ramped up to ~2 tests per 1000 people, while Japan is at 0.17 tests, so Germany is doing 10 times the testing which is done in Japan. I think with that amount of testing, the number of infected people does not look high any more.\nVarious factors influence these numbers. Japan is a group of islands, and was quick in setting up immigration barriers. Germany is tied up in the EU, some people cross borders daily for work, restricting is harder.\nAs for citizen numbers: Japans 126 million are 1.5 times of Germany\u0026rsquo;s 83 million, so the fewer infected people in Japan are distributed over bigger population.\nHow prepared are both health systems? Both call themself 1st world countries, and have health insurance. While I consider the German 2 class health insurance system as not social and gives rich people an advantage, basic care is provided.\nAs of 2019, Japan had 2.5 doctors per 1000 citizens, while Germany had 4.3 doctors. Japan had 11.8 nurses per 1000 citizens, Germany had 13.2. Health spending points at the same direction, and all of that with Japan having more aged people than Germany, so Germany seems medically better prepared.\nHow did the countries help the citizens? Both countries have put up big programs to stimulate the economy.\nIn Japan, much was also done on more personal level: this started with each household receiving 2 facemasks, nicknamed \u0026lsquo;Abe no masuku\u0026rsquo;. People living in Japan were also entitled to apply for 100.000円 (800€) of \u0026lsquo;support money\u0026rsquo;, also I as a foreigner living in Japan would have been eligible. Economists also call that concept \u0026lsquo;helicopter money\u0026rsquo;.\nA further campaign called \u0026lsquo;Goto travel\u0026rsquo; provides money when visiting locations in Japan.\nHow safe/threatened feel the citizens? This is hard to measure objectively, I did not see comparable opinion polls being done in both countries. From consuming media in both countries, and talking to people, it seems like Germans and Japanese feel safe/unsafe in the situation to a comparable level.\nRegarding the \u0026rsquo;toiletpaper supply issue\u0026rsquo;, both countries felt similarly threatened: there was shortage in both countries. This might be surprising for Japan: typhoons and so on are not rare, most people have important items on stock at home.\nSafe/threatened feeling is probably heavily influenced by the media, and how the surrounding countries are doing. Germany looks quite good compared to France and Italy. Compared to Japan, South Korea and Taiwan are doing very well. China has more infections, but also more citizens.\nHow many restrictions have citizens to endure? How restricted feel people from the steps to counter Corona? Germany entered it\u0026rsquo;s lockdown in March, police checked that citizens stay at home, wearing masks was at least encouraged, and in many places also enforced. Opposing movements have formed, demos were on the streets. In Japan, by current law no lockdown is possible. Citizens were informed about the medical background of recommendations, and were called to \u0026lsquo;voluntarily self restrict\u0026rsquo;. Wearing of masks is already established, especially in influenza and hay fever time, so this was widely accepted. In the time of \u0026lsquo;self restrict\u0026rsquo;, also bars and bathes closed down, hiking was discouraged - but doing sport was ok. While restriction was \u0026lsquo;voluntarily\u0026rsquo;, the society was having an eye, of course. Cars on a beach, with the number plate showing that they were from far away, were not unlikely to receive a note reminding of the \u0026lsquo;self restrict\u0026rsquo;.\nNew guidelines were announced in Japan, like \u0026rsquo;exchange business cards online\u0026rsquo;.\nSubjectively, I feel like Germans felt more restricted by the Corona counter measures, than Japanese.\nSummary Putting numbers into relation, Germany looks Corona wise not necessarily \u0026rsquo;less safe\u0026rsquo; than Japan. Japan is having many immigration restrictions in place based on the \u0026lsquo;infection numbers\u0026rsquo;.\nDaily infections Daily corona tests OECD health care comparison ","date":"2020-10-11T10:04:03+02:00","image":"/posts/2020/10/japan_germany_corona/20200429_133340_abe_no_masuku_small.jpg","permalink":"/posts/2020/10/japan_germany_corona/","title":"Who deals better with Corona, Japan or Germany?"},{"content":"Background Everybody knows about the Japanese \u0026ldquo;san\u0026rdquo;, which you attach when communicating with Japanese colleagues. So, when you also refer to non-Japanese colleagues in the same email, do you then also use ~san? If not, do you think they read the email as \u0026lsquo;he does not respect us equally\u0026rsquo;? Just asking to point out one of the small oddities about this piece of etiquette.\nFor those who do not know me, I am a German living in Japan, speaking to some level English and Japanese.\nThe setting Last week, I had a conference call with a customer, and was once more irritated whether I should use \u0026ldquo;san\u0026rdquo; or not. It\u0026rsquo;s simple, you think? Well.. maybe not.\nTo decide whether one should address a customer with lastname-san, the language is an important criteria. When speaking Japanese, one should of course address the customer with ~san, or even more honorific extensions like ~sama. What about conversations in English then? I think these depend on the occasion. A German talking with someone in France, talking in English, is unlikely to use ~san. But a German talking with a customer in Japan in English is likely to use ~san.\nIn my case, things were more complicated in that conference call: I talk regularly with half of the customers on the call, and only in English - it\u0026rsquo;s the smallest common denominator among us from Germany, India and Australia. Addressing is then also simply with \u0026lsquo;you\u0026rsquo; and the first name. I don\u0026rsquo;t have a sales job, I work together with customers on technical issues. We mostly see each other as \u0026lsquo;just people on the same level, trying to solve something\u0026rsquo;, so we just use first names.\nThen, there was the other half of the customer contacts on the call, Japanese, with whom I rarely communicate. Theoretically I could speak with them in Japanese, but considering also the other people, we all communicate in English. Now, when I address the Japanese customers, even though I speak English, I have somewhere in my mind the urge to \u0026rsquo;express respect\u0026rsquo; - so I use ~san with them. Also, despite us speaking English, I just feel like it\u0026rsquo;s more appropriate to stick with their last name.\nAlso colleagues were on the call. In conversations with English speaking colleagues, \u0026lsquo;you\u0026rsquo; and first name is used. In conversations with colleagues in Japanese, last name is used, and ~san. I even to not know the first name of some Japanese colleagues, this is how rarely we use the first name.\nNow, communication in Japanese with colleagues and customers is different again: of course, you address the customer with the last name, and ~san. As for the colleagues: we from the same company now more count as \u0026lsquo;one family/group\u0026rsquo;, and while we still refer to each other with our last name, we do not use ~san.\nThe twist So there we were, people from 2 companies on the call. Many had not yet been in contact, so a round of self introductions was in order. I was reading the list with the confcall attendees, and one by one asking the person to introduce. Easy enough, right? But we had 2 groups from the customer: the native Japanese, and the ones from other countries. As per above background, I was reading the name of the next attendee on the list, and then asking to self introduce. For the ones from Japanese, I should have used the last name and ~san, I mixed that up for one person, noticed, and corrected myself. Then, for one of the customers with whom I speak only English, I plainly used the first name for addressing - and he came right back on me asking \u0026ldquo;uh, not \u0026rsquo;lastname\u0026rsquo;?\u0026rdquo;. :) Also, for half of my Japanese colleagues I ended up using ~san, when as per \u0026ldquo;let\u0026rsquo;s use English addressing\u0026rdquo;, I should have used the first name. Just, also for by boss that would have felt awkward.\nMy conclusion for now When communicating in languages, you might for learning trying to compute and construct single sentences, but in a working environment the language becomes just a tool - you need to concentrate on the contents and rely to some degree on your instincts for addressing people with the right level.\nIf you have a better recipe for dealing with these situations, please share :)\nHow I get addressed in Japan Friends mostly address me as クリス, which is the phonetic expression of \u0026ldquo;Chris\u0026rdquo;.\nJapanese speaking colleagues mostly use クリス-san, sometimes also ホルン-san (ホルン is the phonetic expression of my family name), that variant is then the same as Japanese coworkers use to address each other in business setting: familyname-san. I do not even know the first name of many colleagues.\nAs for Japanese speaking customers, there is a wide variety: from Horn-san, over ホルン-san, and クリス-san. Funny detail: most japanese surnames are expressed with 2 full-width characters, for example 山田. My name with 4 Roman-style letters is mostly expressed in half-width characters, \u0026ldquo;Horn\u0026rdquo;. So when my name appears in lists of meeting attendees, it\u0026rsquo;s width matches exactly up to the width of Japanese names.\nIs it easier in German? In German, we have basically 2 layers to \u0026ldquo;express honorific layers\u0026rdquo; to the other party when speaking. When meeting a customer for the first time, as a service provider one uses the higher level. Again, me working as a specialist together with the customer to solve issues, we quickly dropped to the more familiar layer. Nothing wrong with that - just when talking directly with them, and having for example their boss entering the conversation, we have the situation of staying at the familiar level with the long-known-person, and having to talk high level with the boss - that is awkward.\n","date":"2020-10-06T14:52:19+02:00","image":"/posts/2020/09/japan-to-san-or-not-to-san/20191128_meeting_attendees.jpg","permalink":"/posts/2020/09/japan-to-san-or-not-to-san/","title":"To San or Not to San in Japan"},{"content":"German/Japanese/English versions of this text are here: https://fluxcoil.net/ .\nEnglish: Interactions with technology in Japan When I was little and asked my grandmother where in the world the technology was most advanced, she said: \u0026ldquo;In Japan!\u0026rdquo;. I have been living here for a few years now, and I am still fascinated by the culture of the country, the mentality of the people and especially how the Japanese deal with technology. This fascination has led me to question my own behaviour and culture.\nJapanese and Germans are said to have a tendency towards perfectionism, but I have made the experience that the Japanese go much further. When you tackle something, you try to make it perfect. I see this every day in my work in IT: I used to deal with German customers, now I deal mainly with Japanese customers, who strive to pin everything down to the smallest detail. This desire to understand techniques and technological systems down to their roots reminds me of the typical characteristics of hackers and is part of my fascination with Japan.\nEspecially the combination of tradition and technology in Japan is fascinating for me, for example seen with this robot which can use a Katana, a Japanese sword, with great precision: But Japan\u0026rsquo;s technological history has also already created its own new traditions: In the land of Nintendo and Sega, computer games enjoy such a high reputation that the Japanese Prime Minister even appeared in an Olympic commercial as Mario the Plumber.\nMany technical solutions in Japan look like superfluous gimmicks from the outside, but with knowledge of the closer circumstances and typical social characteristics suddenly make sense. For example the heated toilet seat. Most Europeans hear about it and ask themselves \u0026ldquo;Does that make sense?!\u0026rdquo;.\nAs I learned: Yes, it makes a lot of sense :) I\u0026rsquo;m writing this article in January in a flat in Tokyo, wearing a sweater, 2 jackets and a cap. Most parts of Japan are in average warmer than areas in Germany. For buildings, earthquake resistance is more important than good insulation. Residential buildings are designed to last 30 years. Central heating is not very common, heating is mostly done with electric air conditioners. Bathrooms and toilets are usually unheated, and everybody who has ever sat down on a toilet at 5°C will for sure appreciate heated toilet seats. These conditions also led to the invention of the Kotatsu: a low table at which you sit and under which you put your legs. An electric heater under the table plate radiates heat downwards, a blanket keeps the heat under the table.\nThere have always been and still are technological inventions in Japan that have only been successful in Japan - perhaps it even makes sense to speak of \u0026ldquo;island solutions\u0026rdquo; here: Massage chairs to relax after work, extremely small notebooks, solar-powered gadgets at road junctions to warn vehicles, or cars announcing with a loud voice in which direction they are turning. One sees already more robots in Japan than in other parts of the world.\nI have also observed a particularly striking difference between Japan and Germany in the amount of confidence, with which technological systems are used. The crime rate in Japan is quite low. In a larger restaurant, you typically reserve a seat by putting your bag down and then going to the counter to order and pay. This trust can be seen in various places: In Europe, when someone enters a password and you sit together in front of a screen, the other person explicitly looks away from the screen and keyboard: Just don\u0026rsquo;t come under suspicion that you want to spy on the password. In Japan there is no such suspicion, people tend not to look away in this situation. In an open-plan office, Japanese people rarely lock their screens when they leave their workplace.\nAt tiny train stations, passengers have to take a ticket when entering the train. Later, when leaving the train, distance and fare are calculated. If you forget the ticket when entering the train, then later when leaving the train you will be asked at which station you boarded - and on that base the price will be determined. In Germany, there is no doubt that you would be charged with the maximum fare. The Japanese way to do it: trust by default.\nBeing born in Germany, I sometimes have problems with granting such a trust. For example, in the first flat I lived in, all the residents were able to open all the mailboxes: there were no locks or anything like that. In some areas, only the number of the flat is written on the mailbox, no names. That\u0026rsquo;s enough to address mail. This means that sales persons appearing at your door do not know your name when addressing you - very pleasant.\n)\nRegarding money, Japanese and Germans have much in common. Japanese are very attached to cash. NFC cards, the \u0026ldquo;Suica\u0026rdquo;, have become widespread in local transport. By default, these are not linked to a name: at machines you buy a card anonymously and can charge it with money. Before entering and after leaving the train, the card touches a reader and money is debited. If necessary, you can also register the Suica card with your name and get it activated for flatrate travel between 2 stations. Many commuters do this. Almost all employers reimburse the costs of commuting between work and home.\nThe Japanese government and companies are trying to push through more cashless payments. Last year, the value-added tax was raised from 8% to 10%. However, part of this increase can be reclaimed in exchange for personal data if you register for payback systems in combination with cashless cards.\nWhile many technological systems in Japan are based on trust, this principle does not apply to all areas and all applications of cashless payment. Mid-2019, one of the 3 major companies running convenience stores introduced a cashless payment system and terminated it 3 months later due to too much abuse. 2-factor authentication and other approaches had not been considered in the concept, but should have been..\nWhen applying for a bank account or credit card, you get to choose your PIN. With online banking, TAN lists are widely used. In Japan, many banks do not accept transactions to a different Japanese bank. This can mean that when you start a new job, the new employer may not be able to pay your salary to your existing account - you will have to open a new account. Also the landlord of your flat tells you at which bank his bank account is, and if your bank can not transfer the monthly rent, you might have to open a further bank account. In my first flat, no bank transfers were possible from my bank and I paid cash monthly. Transfers from other countries are rather rare. When transferring money from a German bank account, I received a letter from my bank with the transfer details.\nPeople living in Japan are assigned a National ID number. This so-called \u0026lsquo;my number\u0026rsquo; is used, for example, to map the identity to the person when paying taxes. Also banks ask for the \u0026lsquo;my number\u0026rsquo; when doing money transfers from/to foreign countries.\nInstead of signatures, \u0026lsquo;Hanko\u0026rsquo; are common in Japan, a kind of stamp. Hanko are registered in the city hall and can then be used in the way as signatures are used in Europe. For Japanese people, the imprint of the hanko contains the own names Kanji, my Hanko contains my name in the normal Latin alphabet. In order not to acknowledge a delivered parcel with the same Hanko as for example a rental contract, most Japanese use several Hankos, each one registered separately and used for different occasions.\nWhile in Germany the electronic patient file is still being worked on, in Japan the paper medicine book has been around for a long time, which you can take with you to the pharmacy or doctor. The medication is then recorded in the book, making it possible to check for intolerance. So this is a nice offline solution, with the data staying directly with the patient.\nSIM cards for mobile phones are almost never secured with a PIN in Japan, so convenience is more important than protection against theft. Moreover, mobile phones are usually locked onto one provider, and this is hardly seen as a restriction.\nIn contrast to Germany, there are many extreme situations such as earthquakes or typhoons. In these situations, mobile phones are used to send emergency alerts as SMS-broadcasts, e.g. when rivers have high water levels as a result of a typhoon or when an evacuation gets announced via SMS - as was recently the case in the district of Tokyo where I live..\nJapan is known for its well-developed Internet infrastructure. For my internet connection, physical medium (fiber optic cable) and content provider are two different companies, each with whom I have separate contracts. My content provider also offers IPv6 via opt-in, which I did activate online using a website.\nThe mobile network coverage in Japan is very good, even when using the Shinkansen, the Japanese high speed train. Interestingly enough, the much-cited punctuality of Japanese trains only became apparent after the privatization of the railways.\nIt seems like Yahoo Japan is used here more often than Google: Yahoo Maps was for a long time coping better with the peculiarities of Japanese house addressing. Yahoo Japan offers many country-specific overlay maps: predictions on the speed of typhoons, maps with the best places to see the cherry blossom, the current occupancy of train station toilets or trains. Almost all Japanese travel to work by train, usually with a smartphone in their hand. Many websites are optimized for smartphones, e.g. trailers of new movies are separated into single screenshots and described on a long website. For example, Gigazine has in this style prepared 36c3 video content.\nEvery day I make new discoveries that will challenge the previous explanations. What has remained, however, is my fascination how different people can develop and use different technological solutions for sometimes the same purposes. This article concentrates on technology-related topics, but there is still a lot to learn about society and life here - details are at https://fluxcoil.net/jpblog .\nAdditions After writing the original article, a number of points have emerged from conversations with friends and through research.\nPunctuality of the Japanese Railway increased with privatization: Along with the privatization, some other obstacles for the railway were removed: before privatization, the railway had to employ soldiers returning from the second world war. After privatization, many people lost their job. There were also debt cancellations with the privatization. Several areas also deteriorated with privatization: lines with fewer passengers, e.g. in the countryside, were no longer cross-financed from the other lines, they got labeled \u0026lsquo;uneconomical\u0026rsquo; and closed down. Tidbit regarding the kanji \u0026ldquo;鉄\u0026rdquo;: 鉄 is the first part of 鉄道, which means \u0026ldquo;railway company\u0026rdquo;. 鉄 is composed of 2 radicals, so \u0026ldquo;parts of the kanji\u0026rdquo;: 金 means money/metal, 失 means \u0026ldquo;to loose\u0026rdquo;. So canonically, the whole Kanji would be interpreted as \u0026ldquo;losing money\u0026rdquo;, but nowadays \u0026ldquo;金に矢\u0026rdquo; as interpretation is preferred: with 矢 as an arrow this means \u0026ldquo;as fast as an arrow to the money\u0026rdquo;. Because of the deficit (金を失う= loss of money) in the past, privatisation was started, and the Kanji later reinterpreted to \u0026ldquo;金に矢\u0026rdquo; to have a good omen. ","date":"2020-10-05T13:56:15+02:00","image":"/posts/2020/10/japan_tech_interactions/20200222_100931_dragonquest_store.jpg","permalink":"/posts/2020/10/japan_tech_interactions/","title":"Interactions with technology in Japan"}]