Adventurous Computing

Stopping systemd services under memory pressure

2025-12-15T00:00:00+00:00

Do you have your "favourite" server that is responsible for just a tiny number too many things ?</p>

I know I have one, it's my CI runner, but also serves as a platform to experiment with local LLMs, but also is a Nix builder for all other machines I have. Lots of responsiblities ! Luckily, I don't need it to do all of those things at the same time, it just didn't know that previously. I would run a local build, it would get sent to that machine, but that machine would have e.g. an LLM loaded in memory and the build would OOM. I would then need to go there, stop the LLM service and restart the build. What if a computer could do it all by itself ?!</p>

I present to you - a simple service that monitors available memory and based on that starts and stops other services dynamically. The code is in Nix, that helps with keeping the service code and definition in one place, making sure paths make sense etc. However, there is nothing fundamentally Nix specific to the service itself, you can just take the Bash code and use it anywhere.</p>

Speaking of usage (from Nix) you can use it like so; if you want to, say, stop LLM server if available mem is under 8GB and start it again when it's over whatever LLM needs (in my case it's 64GB):</p>

{</span>
</span>services</span>.custom</span>.memoryManager</span> =</span> {</span>
</span>  enable</span> =</span> true</span>;</span>
</span>  stopThresholdMB</span> =</span> 8192</span>;</span>
</span>  startThresholdMB</span> =</span> config</span>.</span>services</span>.</span>llm</span>.</span>memoryRequirementMB</span>;</span>
</span>  checkIntervalSeconds</span> =</span> 30</span>;</span>
</span>  managedServices</span> =</span> [</span> "</span>llama-cpp-fast"</span></span> ]</span>;</span>
</span>}</span>;</span>
</span>}</span>
</span>
</span></code></pre>
And this gives you</p>
Dec 15 12:05:22 memory-manager: Memory: 7543MB available | Decision: STOP (below 8192MB)
</span>Dec 15 12:05:22 memory-manager: LOW MEMORY: Stopping llama-cpp-fast
</span>Dec 15 12:05:23 memory-manager: llama-cpp-fast stopped successfully
</span>Dec 15 12:08:53 memory-manager: Memory: 88902MB available | Decision: START (above 65536MB)
</span>Dec 15 12:08:53 memory-manager: MEMORY OK: Starting llama-cpp-fast
</span>Dec 15 12:08:54 memory-manager: llama-cpp-fast started successfully
</span></code></pre>
And here is the module definition itself. The code is not the prettiest but it does help a lot already !</p>
{</span>
</span>  config</span>,</span>
</span>  lib</span>,</span>
</span>  pkgs</span>,</span>
</span>  ...
</span></span></span>}</span>:</span>
</span>
</span>with</span> lib</span>;
</span>
</span>let</span>
</span>  cfg</span> =</span> config</span>.</span>services</span>.</span>custom</span>.</span>memoryManager</span>;</span>
</span>
</span>  monitorScript</span> =</span> pkgs</span>.</span>writeShellScript</span> "</span>memory-manager"</span></span> ''</span>
</span></span>    set -euo pipefail
</span></span>
</span></span>    STOP_THRESHOLD_MB=${</span>toString</span> cfg</span>.</span>stopThresholdMB</span>}</span></span>
</span></span>    START_THRESHOLD_MB=${</span>toString</span> cfg</span>.</span>startThresholdMB</span>}</span></span>
</span></span>    CHECK_INTERVAL=${</span>toString</span> cfg</span>.</span>checkIntervalSeconds</span>}</span></span>
</span></span>    MANAGED_SERVICES="${</span>concatStringsSep</span> "</span> "</span></span> cfg</span>.</span>managedServices</span>}</span></span>"
</span></span>    STATE_DIR="/run/memory-manager"
</span></span>    DEBUG=${</span>if</span> cfg</span>.</span>debug</span> th</span>en</span> "</span>true"</span></span> el</span>se</span> "</span>false"</span></span>}</span></span>
</span></span>
</span></span>    mkdir -p "$STATE_DIR"
</span></span>
</span></span>    log() {
</span></span>      echo "$1"
</span></span>    }
</span></span>
</span></span>    get_available_memory_mb() {
</span></span>      ${</span>pkgs</span>.</span>gawk</span>}</span></span>/bin/awk '/^MemAvailable:/ { printf "%d", $2 / 1024 }' /proc/meminfo
</span></span>    }
</span></span>
</span></span>    is_service_stopped_by_us() {
</span></span>      local service="$1"
</span></span>      [ -f "$STATE_DIR/$service.stopped" ]
</span></span>    }
</span></span>
</span></span>    mark_service_stopped() {
</span></span>      local service="$1"
</span></span>      touch "$STATE_DIR/$service.stopped"
</span></span>    }
</span></span>
</span></span>    mark_service_started() {
</span></span>      local service="$1"
</span></span>      rm -f "$STATE_DIR/$service.stopped"
</span></span>    }
</span></span>
</span></span>    stop_service() {
</span></span>      local service="$1"
</span></span>      
</span></span>      if ${</span>pkgs</span>.</span>systemd</span>}</span></span>/bin/systemctl is-active --quiet "$service"; then
</span></span>        log "LOW MEMORY: Stopping $service"
</span></span>        
</span></span>        ${</span>pkgs</span>.</span>systemd</span>}</span></span>/bin/systemctl disable --runtime --now "$service" 2>/dev/null || true
</span></span>        
</span></span>        if ! ${</span>pkgs</span>.</span>systemd</span>}</span></span>/bin/systemctl is-active --quiet "$service"; then
</span></span>          mark_service_stopped "$service"
</span></span>          log "$service stopped successfully"
</span></span>        else
</span></span>          log "WARNING: Failed to stop $service"
</span></span>        fi
</span></span>      fi
</span></span>    }
</span></span>
</span></span>    start_service() {
</span></span>      local service="$1"
</span></span>      
</span></span>      if is_service_stopped_by_us "$service"; then
</span></span>        log "MEMORY OK: Starting $service"
</span></span>        ${</span>pkgs</span>.</span>systemd</span>}</span></span>/bin/systemctl enable --runtime "$service" 2>/dev/null || true
</span></span>        ${</span>pkgs</span>.</span>systemd</span>}</span></span>/bin/systemctl start "$service" 2>/dev/null || true
</span></span>        
</span></span>        if ${</span>pkgs</span>.</span>systemd</span>}</span></span>/bin/systemctl is-active --quiet "$service"; then
</span></span>          mark_service_started "$service"
</span></span>          log "$service started successfully"
</span></span>        else
</span></span>          log "WARNING: Failed to start $service"
</span></span>        fi
</span></span>      fi
</span></span>    }
</span></span>
</span></span>    log "Memory manager started"
</span></span>    log "Stop threshold: ''$</span>{STOP_THRESHOLD_MB}MB, Start threshold: ''$</span>{START_THRESHOLD_MB}MB"
</span></span>    log "Managed services: $MANAGED_SERVICES"
</span></span>    log "Check interval: ''$</span>{CHECK_INTERVAL}s, Debug: $DEBUG"
</span></span>
</span></span>    while true; do
</span></span>      AVAILABLE_MB=$(get_available_memory_mb)
</span></span>      
</span></span>      if [ "$AVAILABLE_MB" -lt "$STOP_THRESHOLD_MB" ]; then
</span></span>        DECISION="STOP (below ''$</span>{STOP_THRESHOLD_MB}MB)"
</span></span>        log "Memory: ''$</span>{AVAILABLE_MB}MB available | Decision: $DECISION"
</span></span>        for service in $MANAGED_SERVICES; do
</span></span>          stop_service "$service"
</span></span>        done
</span></span>      elif [ "$AVAILABLE_MB" -gt "$START_THRESHOLD_MB" ]; then
</span></span>        DECISION="START (above ''$</span>{START_THRESHOLD_MB}MB)"
</span></span>        NEEDS_START=false
</span></span>        for service in $MANAGED_SERVICES; do
</span></span>          if is_service_stopped_by_us "$service"; then
</span></span>            NEEDS_START=true
</span></span>            break
</span></span>          fi
</span></span>        done
</span></span>        if [ "$NEEDS_START" = "true" ]; then
</span></span>          log "Memory: ''$</span>{AVAILABLE_MB}MB available | Decision: $DECISION"
</span></span>          for service in $MANAGED_SERVICES; do
</span></span>            start_service "$service"
</span></span>          done
</span></span>        else
</span></span>          log "Memory: ''$</span>{AVAILABLE_MB}MB available | Decision: OK (services running)"
</span></span>        fi
</span></span>      else
</span></span>        if [ "$DEBUG" = "true" ]; then
</span></span>          DECISION="WAIT (between ''$</span>{STOP_THRESHOLD_MB}MB and ''$</span>{START_THRESHOLD_MB}MB)"
</span></span>          SERVICE_STATUS=""
</span></span>          for service in $MANAGED_SERVICES; do
</span></span>            if ${</span>pkgs</span>.</span>systemd</span>}</span></span>/bin/systemctl is-active --quiet "$service"; then
</span></span>              SERVICE_STATUS="$SERVICE_STATUS $service=running"
</span></span>            elif is_service_stopped_by_us "$service"; then
</span></span>              SERVICE_STATUS="$SERVICE_STATUS $service=stopped-by-us"
</span></span>            else
</span></span>              SERVICE_STATUS="$SERVICE_STATUS $service=stopped"
</span></span>            fi
</span></span>          done
</span></span>          log "Memory: ''$</span>{AVAILABLE_MB}MB available | Decision: $DECISION | Services:$SERVICE_STATUS"
</span></span>        fi
</span></span>      fi
</span></span>      
</span></span>      sleep "$CHECK_INTERVAL"
</span></span>    done
</span></span>  ''</span></span>;</span>
</span>in</span>
</span>{</span>
</span>  options</span>.services</span>.custom</span>.memoryManager</span> =</span> {</span>
</span>    enable</span> =</span> mkEnableOption</span> "</span>memory pressure manager that stops/starts services based on available memory"</span></span>;</span>
</span>
</span>    stopThresholdMB</span> =</span> mkOption</span> {</span>
</span>      type</span> =</span> types</span>.</span>int</span>;</span>
</span>      default</span> =</span> 8192</span>;</span>
</span>      description</span> =</span> "</span>Stop managed services when available memory drops below this threshold (in MB)"</span></span>;</span>
</span>      example</span> =</span> 4096</span>;</span>
</span>    }</span>;</span>
</span>
</span>    startThresholdMB</span> =</span> mkOption</span> {</span>
</span>      type</span> =</span> types</span>.</span>int</span>;</span>
</span>      default</span> =</span> 16384</span>;</span>
</span>      description</span> =</span> "</span>Restart managed services when available memory rises above this threshold (in MB). Should be higher than stopThresholdMB to prevent flapping."</span></span>;</span>
</span>      example</span> =</span> 32768</span>;</span>
</span>    }</span>;</span>
</span>
</span>    checkIntervalSeconds</span> =</span> mkOption</span> {</span>
</span>      type</span> =</span> types</span>.</span>int</span>;</span>
</span>      default</span> =</span> 10</span>;</span>
</span>      description</span> =</span> "</span>How often to check memory availability (in seconds)"</span></span>;</span>
</span>      example</span> =</span> 30</span>;</span>
</span>    }</span>;</span>
</span>
</span>    managedServices</span> =</span> mkOption</span> {</span>
</span>      type</span> =</span> types</span>.</span>listOf</span> types</span>.</span>str</span>;</span>
</span>      default</span> =</span> [</span> ]</span>;</span>
</span>      description</span> =</span> "</span>List of systemd service names to manage (without .service suffix)"</span></span>;</span>
</span>      example</span> =</span> [</span> "</span>llama-cpp-fast"</span></span> ]</span>;</span>
</span>    }</span>;</span>
</span>
</span>    debug</span> =</span> mkOption</span> {</span>
</span>      type</span> =</span> types</span>.</span>bool</span>;</span>
</span>      default</span> =</span> false</span>;</span>
</span>      description</span> =</span> "</span>Enable debug logging (logs WAIT states on every check interval)"</span></span>;</span>
</span>    }</span>;</span>
</span>  }</span>;</span>
</span>
</span>  config</span> =</span> mkIf</span> cfg</span>.</span>enable</span> {</span>
</span>    assertions</span> =</span> [</span>
</span>      {</span>
</span>        assertion</span> =</span> cfg</span>.</span>startThresholdMB</span> ></span> cfg</span>.</span>stopThresholdMB</span>;</span>
</span>        message</span> =</span> "</span>services.custom.memoryManager.startThresholdMB must be greater than stopThresholdMB to prevent flapping"</span></span>;</span>
</span>      }</span>
</span>      {</span>
</span>        assertion</span> =</span> cfg</span>.</span>managedServices</span> !=</span> [</span> ]</span>;</span>
</span>        message</span> =</span> "</span>services.custom.memoryManager.managedServices must not be empty when enabled"</span></span>;</span>
</span>      }</span>
</span>    ]</span>;</span>
</span>
</span>    systemd</span>.services</span>.memory-manager</span> =</span> {</span>
</span>      description</span> =</span> "</span>Memory Pressure Service Manager"</span></span>;</span>
</span>      wantedBy</span> =</span> [</span> "</span>multi-user.target"</span></span> ]</span>;</span>
</span>      after</span> =</span> [</span> "</span>multi-user.target"</span></span> ]</span>;</span>
</span>
</span>      serviceConfig</span> =</span> {</span>
</span>        Type</span> =</span> "</span>simple"</span></span>;</span>
</span>        Restart</span> =</span> "</span>always"</span></span>;</span>
</span>        RestartSec</span> =</span> "</span>10s"</span></span>;</span>
</span>        ExecStart</span> =</span> "</span>${</span>monitorScript</span>}</span></span>"</span></span>;</span>
</span>        RuntimeDirectory</span> =</span> "</span>memory-manager"</span></span>;</span>
</span>        RuntimeDirectoryMode</span> =</span> "</span>0755"</span></span>;</span>
</span>      }</span>;</span>
</span>    }</span>;</span>
</span>  }</span>;</span>
</span>}</span>
</span></code></pre>



Long posts in Mastodon, but now declarative !
2025-12-06T00:00:00+00:00
So much to say on the Fediverse and so little characters of text per post !
Well, at least on Mastodon.
But fear not - someone smarter than me figured out what files to patch to increase the limit</a>.</p>
As I am increasingly lazy, I dreaded the change as it would mean I would need to reapply the patch whenever I redeploy the server, which is more often than I want to admit.</p>
Nix to the rescue !</p>
{</span> pkgs</span>,</span> inputs</span>,</span> ... </span>}</span>:</span>
</span>let</span>
</span>  mastodonCharacterLimit</span> =</span> 5000</span>;</span>
</span>  mastodonBase</span> =</span> inputs</span>.</span>nixpkgs-nixos-unstable</span>.</span>legacyPackages</span>.</span>${</span>pkgs</span>.</span>system</span>}</span></span>.</span>mastodon</span>;</span>
</span>  package</span> =</span> mastodonBase</span>.</span>overrideAttrs</span> (</span>old</span>:</span> {</span>
</span>    postPatch</span> =</span> (</span>old</span>.</span>postPatch</span> or</span> "</span>"</span></span>)</span> +</span> ''</span>
</span></span>      substituteInPlace app/javascript/mastodon/features/compose/containers/compose_form_container.js \
</span></span>        --replace-fail "max_characters'], 500)" "max_characters'], ${</span>toString</span> mastodonCharacterLimit</span>}</span></span>)"
</span></span>
</span></span>      substituteInPlace app/validators/status_length_validator.rb \
</span></span>        --replace-fail "MAX_CHARS = 500" "MAX_CHARS = ${</span>toString</span> mastodonCharacterLimit</span>}</span></span>"
</span></span>    ''</span></span>;</span>
</span>  }</span>)</span>;</span>
</span>in</span>
</span>{</span>
</span>  services</span>.mastodon</span> =</span> {</span>
</span>    enable</span> =</span> true</span>;</span>
</span>    inherit</span> package</span>;</span>
</span>  }</span>;</span>
</span>}</span>
</span></code></pre>
And that is all ! Doctor Doom toots as he pleases !
See you on the fedi - ping me at @cyplo@peninsula.industries</code> !</p>


Creating a base OCI image for Nix flake builds within Gitea/Forgejo
2023-09-18T00:00:00+00:00
I've been moving more and more of my infrastructure to be self-hosted recently.
Part of that involves setting up CI jobs for testing and publishing artifacts, mostly rust crates but also this very blog.</p>
I really wanted to re-use my existing Nix flakes for those projects, this way I know my local dev env</a> would be the same env then used on CI</a>.</p>
I am self-hosting</a> a Gitea instance (will probably be migrating to Forgejo) and it uses a CI system</a> built to resemble Github actions</a> - basically you run your jobs as containers and within those you can run arbitrary commands. You can also take advantage of the existing ecosystem of actions</code>.</p>
I wanted a base image that would have on one hand nix</code> with flakes</code> enabled but on the other hand would be compatible with running popular actions from other authors. This meant having nix</code>, git</code> but also nodejs</code> available amongts other things. I couldn't find one that would have both, so I built one !</p>
I'm building on top of the definitions</a> from docker-nixpkgs</a> and just tweaking them to add the things needed for actions</code> and also for the definition itself to be a flake, for an added flavour. This allows me to add multiple image definitions in the same repo and the build them independently when needed.</p>
Here's the whole definition in all of its glory, defining 2 images - hello</code> and flakes-action</code> - the hello</code> being a test image for testing the process itself and flakes-action</code> is the one I'm using on CI currently.</p>
{</span>
</span>  description</span> =</span> "</span>docker base images"</span></span>;</span>
</span>
</span>  inputs</span> =</span> {</span>
</span>    nixpkgs</span>.url</span> =</span> "</span>github:nixos/nixpkgs/nixos-unstable"</span></span>;</span>
</span>    flake-utils</span>.url</span> =</span> "</span>github:numtide/flake-utils"</span></span>;</span>
</span>  }</span>;</span>
</span>
</span>  outputs</span> =</span> {</span> self</span>,</span> nixpkgs</span>,</span> flake-utils</span> }</span>:</span>
</span>    flake-utils</span>.</span>lib</span>.</span>eachDefaultSystem</span>
</span>      (</span>system</span>:</span>
</span>        let</span>
</span>          pkgs</span> =</span> import</span> nixpkgs</span> {</span> inherit</span> system</span>;</span> }</span>;</span>
</span>          pkgsStatic</span> =</span> pkgs</span>.</span>pkgsStatic</span>;</span>
</span>          lib</span> =</span> pkgs</span>.</span>lib</span>;</span>
</span>
</span>        in</span>
</span>        {</span>
</span>          packages</span> =</span> {</span>
</span>            hello</span> =</span> pkgs</span>.</span>dockerTools</span>.</span>buildImage</span> {</span>
</span>              name</span> =</span> "</span>hello-docker"</span></span>;</span>
</span>              config</span> =</span> {</span>
</span>                Cmd</span> =</span> [</span> "</span>${</span>pkgs</span>.</span>hello</span>}</span></span>/bin/hello"</span></span> ]</span>;</span>
</span>              }</span>;</span>
</span>            }</span>;</span>
</span>            flakes-action</span> =</span> pkgs</span>.</span>dockerTools</span>.</span>buildImageWithNixDb</span> {</span>
</span>              name</span> =</span> "</span>flakes-action"</span></span>;</span>
</span>              contents</span> =</span> with</span> pkgs</span>; [</span>
</span>                ./root</span>
</span>                bash</span>
</span>                coreutils</span>
</span>                curl</span>
</span>                gawk</span>
</span>                gitFull</span>
</span>                git-lfs</span>
</span>                gnused</span>
</span>                nodejs</span>
</span>                wget</span>
</span>                sudo</span>
</span>                nixFlakes</span>
</span>                cacert</span>
</span>                gnutar</span>
</span>                gzip</span>
</span>                openssh</span>
</span>                xz</span>
</span>                (</span>pkgs</span>.</span>writeTextFile</span> {</span>
</span>                  name</span> =</span> "</span>nix.conf"</span></span>;</span>
</span>                  destination</span> =</span> "</span>/etc/nix/nix.conf"</span></span>;</span>
</span>                  text</span> =</span> ''</span>
</span></span>                    accept-flake-config = true
</span></span>                    experimental-features = nix-command flakes
</span></span>                  ''</span></span>;</span>
</span>                }</span>)</span>
</span>              ]</span>;</span>
</span>
</span>              extraCommands</span> =</span> ''</span>
</span></span>                # for /usr/bin/env
</span></span>                mkdir usr
</span></span>                ln -s ../bin usr/bin
</span></span>
</span></span>                # make sure /tmp exists
</span></span>                mkdir -m 1777 tmp
</span></span>
</span></span>                # need a HOME
</span></span>                mkdir -vp root
</span></span>              ''</span></span>;</span>
</span>              config</span> =</span> {</span>
</span>                Cmd</span> =</span> [</span> "</span>/bin/bash"</span></span> ]</span>;</span>
</span>                Env</span> =</span> [</span>
</span>                  "</span>LANG=en_GB.UTF-8"</span></span>
</span>                  "</span>ENV=/etc/profile.d/nix.sh"</span></span>
</span>                  "</span>BASH_ENV=/etc/profile.d/nix.sh"</span></span>
</span>                  "</span>NIX_BUILD_SHELL=/bin/bash"</span></span>
</span>                  "</span>NIX_PATH=nixpkgs=${</span>./fake_nixpkgs</span>}</span></span>"</span></span>
</span>                  "</span>PAGER=cat"</span></span>
</span>                  "</span>PATH=/usr/bin:/bin"</span></span>
</span>                  "</span>SSL_CERT_FILE=${</span>pkgs</span>.</span>cacert</span>}</span></span>/etc/ssl/certs/ca-bundle.crt"</span></span>
</span>                  "</span>USER=root"</span></span>
</span>                ]</span>;</span>
</span>              }</span>;</span>
</span>            }</span>;</span>
</span>          }</span>;</span>
</span>        }</span>)</span>;</span>
</span>}</span>
</span></code></pre>
If you want to build this yourself you can:</p>
git</span></span> clone https://git.cyplo.dev/cyplo/base-images.git</span>
</span>cd</span></span> base-images</span>
</span>nix</span></span> build '</span>.#flakes-action'</span></span></span>
</span>docker</span></span> load <</span> result #</span></span> this took me so much time, to realise I need `load` and not `import`...</span>
</span></span></span>docker tag [</span>image id]</span> yourimage.repo/base-images/flakes-action:latest</span>
</span>docker</span></span> push yourimage.repo/base-images/flakes-action:latest</span>
</span></code></pre>
Then to use on CI, an example of a Gitea CI config:</p>
on</span>:</span> push</span>
</span>jobs</span></span>:</span>
</span>  Publish</span></span>:</span>
</span>    runs-on</span></span>:</span> flakes-action</span>
</span>    steps</span></span>:</span>
</span>      -</span> uses</span></span>:</span> actions/checkout@v3</span>
</span>        name</span></span>:</span> Checkout</span>
</span>      -</span> name</span></span>:</span> Build</span> 
</span>        run</span></span>:</span> |</span>
</span>          nix develop -c hugo --gc --minify
</span></span></code></pre>
It uses the image pushed and both a custom build script but also a well-known checkout</code> action.</p>
You need to teach your Gitea runner about the image first btw; if you use NixOS for the runner definition, it could look like this:</p>
services</span>.</span>gitea-actions-runner</span> =</span> {</span>
</span>  instances</span>.boltyone</span> =</span> {</span>
</span>    enable</span> =</span> true</span>;</span>
</span>    url</span> =</span> "</span>https://yourgitea.domain"</span></span>;</span>
</span>    tokenFile</span> =</span> config</span>.</span>sops</span>.</span>secrets</span>.</span>"</span>gitea-runner-token"</span></span>.</span>path</span>;</span>
</span>    name</span> =</span> "</span>bolty one"</span></span>;</span>
</span>    labels</span> =</span> [</span>
</span>      "</span>flakes-action:docker://yourimage.repo/base-images/flakes-action:latest"</span></span>
</span>      "</span>ubuntu-kinetic:docker://ubuntu:kinetic"</span></span>
</span>      "</span>linux_amd64:host"</span></span>
</span>    ]</span>;</span>
</span>  }</span>;</span>
</span>}</span>;</span>
</span></code></pre>
P.S. shoutout to nixery</a> that I tried first and the resulting images were just a bit off as it was not easy to get them to support flakes. I think it's an amazing tool in its own right though and you should try it, you can do things like docker run -ti nixery.dev/shell/git/htop bash</code> and it will happily just give you an image with those arbitrary nixpkgs included !</p>
Happy hacking !</p>
Discuss this post on the Fediverse (original post no longer available: peninsula.industries/@cyplo/111087014413124274)</p>


This blog is now self-hosted !
2023-08-19T00:00:00+00:00
This is just a quick info that this blog is now self-hosted by yours truly. I'm using a small server at Mythic Beasts</a>.</p>
I have a small CI situation</a> going on via (self-hosted) Gitea Actions, it rsync</code>s the contents to the server. Good old nginx</code> + rsync</code> and that's it.</p>
I took this as an opportunity to hook up htmltest</a> and it highlighted that a lot of places I linked to no longer exist or have moved. I've updated</a> all 13 years worth of blog posts and I'm happy to report that there's no longer any dead link present !</p>


Containerised Mastodon server on NixOS
2023-01-15T00:00:00+00:00
If you're here to learn about how to set up a Mastodon server on NixOS the easy way - you can use services.mastodon.enable = true; and take it from there !
</span>This post does describe a bit more convoluted setup than that.
</span></code></pre>
I've been on the fediverse on and off since friendi.ca</a> started to be a thing.
I remember hosting an instance at ponk.pink that suddenly got popular and the server melted away while I was apologising to the users.
I can see someone bought the domain to host some psychedelic My Little Pony stuff, I'm all for that.</p>
I moved to todon.nl</a> when it got started and was quite happy there - check it out if you're looking for an instance !
The itch was there though, what if I host my own small cozy place on the fediverse myself.
I knew I didn't have as much time to do server admin, so I decided to start small, with a single user instance and use the power of NixOS to help with maintenance.</p>
Here is my full config - hosting Mastodon under the apex domain of peninsula.industries</a>, Mastodon is running inside a systemd container and the config is using nix-sops to store the secrets.
Few things that were unexpected/of note</p>

I needed to create the folder structure with correct permissions so that Mastodon starts normally. This in turn required setting up users and groups both on host and inside the container so that their uids and gids match.</li>
I'm decrypting secrets on the host and making them available read-only to the container. Not sure if this is better or worse than having sops inside of the container, but I was having some trouble using sops module from there, so left it as it is for now - something to look into in the future</li>
I needed to change the postgres port that is running inside of the Mastodon container because I had another postgres on this host already.</li>
If you're gonna be playing with this a lot, you might need to remove the container and its data and start from scratch - you can do so by doing:

remove the container definition and nixos-rebuild switch</code></li>
rm -fr /var/lib/nixos-containers/mastodon/</code></li>
</ul>
</li>
to use Mastodon CLI you need to be running it as a correct user and within the Nix-changed env:nixos-container</span></span> root-login mastodon</span>
</span>sudo</span></span> -</span>u</span> mastodon bash</span>
</span>cd</span></span>
</span>source</span></span> mastodon-env</span>
</span>RAILS_ENV</span>=</span>production</span> tootctl</span></span>
</span></code></pre>
</li>
The below can use some refactoring, there is some weird repetition in some places still</li>
You can find the newest version of this configuration at git.cyplo.dev/cyplo/dotfiles/src/branch/main/nixos/boxes/cupsnet/mastodon.nix</a>.</li>
Share your thoughts on the fedi thread for this article (original post no longer available: peninsula.industries/@cyplo/109692744740475900) !</li>
</ul>
{</span>
</span>  config</span>,</span>
</span>  pkgs</span>,</span>
</span>  inputs</span>,</span>
</span>  lib</span>,</span>
</span>  ...
</span></span></span>}</span>:</span> let</span>
</span>  domain</span> =</span> "</span>peninsula.industries"</span></span>;</span>
</span>  streamingPort</span> =</span> 55000</span>;</span>
</span>  webPort</span> =</span> 55001</span>;</span>
</span>  postgresPort</span> =</span> 5433</span>;</span>
</span>  path</span> =</span> "</span>/var/lib/mastodon/"</span></span>;</span>
</span>  mailgunSmtpSecretName</span> =</span> "</span>mastodon-mailgun-smtp-password"</span></span>;</span>
</span>  mailgunSmtpPasswordPath</span> =</span> "</span>/run/secrets/${</span>mailgunSmtpSecretName</span>}</span></span>"</span></span>;</span>
</span>  mastodonDbSecretName</span> =</span> "</span>mastodon-db"</span></span>;</span>
</span>  mastodonDbSecretPath</span> =</span> "</span>/run/secrets/${</span>mastodonDbSecretName</span>}</span></span>"</span></span>;</span>
</span>  uid</span> =</span> 2049</span>;</span>
</span>  gid</span> =</span> 3049</span>;</span>
</span>  systemUserName</span> =</span> "</span>mastodon"</span></span>;</span>
</span>  systemGroupName</span> =</span> "</span>mastodon"</span></span>;</span>
</span>  users</span> =</span> {</span>
</span>    users</span>."</span>${</span>systemUserName</span>}</span></span>"</span></span> =</span> {</span>
</span>      inherit</span> uid</span>;</span>
</span>      isSystemUser</span> =</span> true</span>;</span>
</span>      isNormalUser</span> =</span> false</span>;</span>
</span>      group</span> =</span> systemGroupName</span>;</span>
</span>    }</span>;</span>
</span>    groups</span>."</span>${</span>systemGroupName</span>}</span></span>"</span></span> =</span> {</span>
</span>      inherit</span> gid</span>;</span>
</span>      members</span> =</span> [</span>"</span>${</span>systemUserName</span>}</span></span>"</span></span> "</span>nginx"</span></span>]</span>;</span>
</span>    }</span>;</span>
</span>  }</span>;</span>
</span>  secretSettings</span> =</span> {</span>
</span>    owner</span> =</span> systemUserName</span>;</span>
</span>    group</span> =</span> systemGroupName</span>;</span>
</span>  }</span>;</span>
</span>  publicPath</span> =</span> "</span>${</span>path</span>}</span></span>/public-system/"</span></span>;</span>
</span>  package</span> =</span>
</span>    inputs</span>.</span>nixpkgs-nixos-unstable</span>.</span>legacyPackages</span>.</span>"</span>${</span>pkgs</span>.</span>system</span>}</span></span>"</span></span>.</span>mastodon</span>;</span>
</span>in</span> {</span>
</span>  imports</span> =</span> [</span>../nginx.nix</span>]</span>;</span>
</span>
</span>  services</span>.nginx</span> =</span> {</span>
</span>    virtualHosts</span> =</span> {</span>
</span>      "</span>${</span>domain</span>}</span></span>"</span></span> =</span> {</span>
</span>        forceSSL</span> =</span> true</span>;</span>
</span>        enableACME</span> =</span> true</span>;</span>
</span>        root</span> =</span> "</span>${</span>package</span>}</span></span>/public/"</span></span>;</span>
</span>
</span>        locations</span>."</span>/"</span></span> =</span> {</span>tryFiles</span> =</span> "</span>$uri @proxy"</span></span>;</span>}</span>;</span>
</span>        locations</span>."</span>/system/"</span></span>.alias</span> =</span> "</span>${</span>publicPath</span>}</span></span>"</span></span>;</span>
</span>
</span>        locations</span>."</span>@proxy"</span></span> =</span> {</span>
</span>          proxyPass</span> =</span> "</span>http://127.0.0.1:"</span></span> +</span> toString</span> webPort</span>;</span>
</span>          proxyWebsockets</span> =</span> true</span>;</span>
</span>        }</span>;</span>
</span>        locations</span>."</span>/api/v1/streaming/"</span></span> =</span> {</span>
</span>          proxyPass</span> =</span> "</span>http://127.0.0.1:"</span></span> +</span> toString</span> streamingPort</span>;</span>
</span>          proxyWebsockets</span> =</span> true</span>;</span>
</span>        }</span>;</span>
</span>      }</span>;</span>
</span>    }</span>;</span>
</span>  }</span>;</span>
</span>
</span>  sops</span>.secrets</span>."</span>${</span>mailgunSmtpSecretName</span>}</span></span>"</span></span> =</span>
</span>    {</span>
</span>      sopsFile</span> =</span> ./mailgun.sops.yaml</span>;</span>
</span>      path</span> =</span> mailgunSmtpPasswordPath</span>;</span>
</span>    }</span>
</span>    //</span> secretSettings</span>;</span>
</span>  sops</span>.secrets</span>."</span>${</span>mastodonDbSecretName</span>}</span></span>"</span></span> =</span>
</span>    {</span>
</span>      sopsFile</span> =</span> ./mastodon-db.sops.yaml</span>;</span>
</span>      path</span> =</span> mastodonDbSecretPath</span>;</span>
</span>    }</span>
</span>    //</span> secretSettings</span>;</span>
</span>
</span>  inherit</span> users</span>;</span>
</span>
</span>  systemd</span>.services</span>.mastodon-make-path</span> =</span> {</span>
</span>    script</span> =</span> ''</span>
</span></span>      mkdir -p ${</span>path</span>}</span></span>
</span></span>      chown -R ${</span>systemUserName</span>}</span></span>:${</span>systemGroupName</span>}</span></span> ${</span>path</span>}</span></span>
</span></span>      mkdir -p ${</span>publicPath</span>}</span></span>
</span></span>      chmod -R o-rwx ${</span>publicPath</span>}</span></span>
</span></span>      chmod -R g-rwx ${</span>publicPath</span>}</span></span>
</span></span>      chmod -R g+X ${</span>publicPath</span>}</span></span>
</span></span>      chmod -R g+r ${</span>publicPath</span>}</span></span>
</span></span>      chmod -R u+rwX ${</span>publicPath</span>}</span></span>
</span></span>    ''</span></span>;</span>
</span>    serviceConfig</span> =</span> {</span>
</span>      Type</span> =</span> "</span>oneshot"</span></span>;</span>
</span>    }</span>;</span>
</span>    before</span> =</span> [</span>"</span>container@mastodon.service"</span></span>]</span>;</span>
</span>  }</span>;</span>
</span>
</span>  containers</span>.mastodon</span> =</span> {</span>
</span>    autoStart</span> =</span> true</span>;</span>
</span>    forwardPorts</span> =</span> [</span>
</span>      {</span>
</span>        containerPort</span> =</span> streamingPort</span>;</span>
</span>        hostPort</span> =</span> streamingPort</span>;</span>
</span>      }</span>
</span>      {</span>
</span>        containerPort</span> =</span> webPort</span>;</span>
</span>        hostPort</span> =</span> webPort</span>;</span>
</span>      }</span>
</span>    ]</span>;</span>
</span>    bindMounts</span> =</span> {</span>
</span>      "</span>${</span>path</span>}</span></span>"</span></span> =</span> {</span>
</span>        hostPath</span> =</span> "</span>${</span>path</span>}</span></span>"</span></span>;</span>
</span>        isReadOnly</span> =</span> false</span>;</span>
</span>      }</span>;</span>
</span>      "</span>${</span>mailgunSmtpPasswordPath</span>}</span></span>"</span></span> =</span> {</span>
</span>        hostPath</span> =</span> "</span>${</span>mailgunSmtpPasswordPath</span>}</span></span>"</span></span>;</span>
</span>        isReadOnly</span> =</span> true</span>;</span>
</span>      }</span>;</span>
</span>      "</span>${</span>mastodonDbSecretPath</span>}</span></span>"</span></span> =</span> {</span>
</span>        hostPath</span> =</span> "</span>${</span>mastodonDbSecretPath</span>}</span></span>"</span></span>;</span>
</span>        isReadOnly</span> =</span> true</span>;</span>
</span>      }</span>;</span>
</span>    }</span>;</span>
</span>    config</span> =</span> {</span>
</span>      config</span>,</span>
</span>      pkgs</span>,</span>
</span>      lib</span>,</span>
</span>      ...
</span></span>    </span>}</span>:</span> {</span>
</span>      system</span>.stateVersion</span> =</span> "</span>22.05"</span></span>;</span>
</span>      services</span>.postgresql</span>.port</span> =</span> postgresPort</span>;</span>
</span>      users</span> =</span>
</span>        users</span>
</span>        //</span> {</span>
</span>          mutableUsers</span> =</span> false</span>;</span>
</span>          allowNoPasswordLogin</span> =</span> true</span>;</span>
</span>        }</span>;</span>
</span>      services</span>.mastodon</span> =</span> {</span>
</span>        enable</span> =</span> true</span>;</span>
</span>        inherit</span> package</span>;</span>
</span>        localDomain</span> =</span> "</span>${</span>domain</span>}</span></span>"</span></span>;</span>
</span>        user</span> =</span> systemUserName</span>;</span>
</span>        group</span> =</span> systemGroupName</span>;</span>
</span>        smtp</span> =</span> {</span>
</span>          host</span> =</span> "</span>smtp.eu.mailgun.org"</span></span>;</span>
</span>          port</span> =</span> 465</span>;</span>
</span>          authenticate</span> =</span> true</span>;</span>
</span>          user</span> =</span> "</span>postmaster@${</span>domain</span>}</span></span>"</span></span>;</span>
</span>          fromAddress</span> =</span> "</span>Peninsula Industries Mastodon <mastodon@${</span>domain</span>}</span></span>>"</span></span>;</span>
</span>          createLocally</span> =</span> false</span>;</span>
</span>          passwordFile</span> =</span> "</span>${</span>mailgunSmtpPasswordPath</span>}</span></span>"</span></span>;</span>
</span>        }</span>;</span>
</span>        extraConfig</span> =</span> {</span>
</span>          SMTP_TLS</span> =</span> "</span>true"</span></span>;</span>
</span>          SMTP_ENABLE_STARTTLS_AUTO</span> =</span> "</span>true"</span></span>;</span>
</span>          SINGLE_USER_MODE</span> =</span> "</span>true"</span></span>;</span>
</span>          RAILS_SERVE_STATIC_FILES</span> =</span> "</span>true"</span></span>;</span>
</span>        }</span>;</span>
</span>        inherit</span> streamingPort</span>;</span>
</span>        inherit</span> webPort</span>;</span>
</span>        configureNginx</span> =</span> false</span>;</span>
</span>        enableUnixSocket</span> =</span> false</span>;</span>
</span>        database</span> =</span> {</span>
</span>          port</span> =</span> postgresPort</span>;</span>
</span>          passwordFile</span> =</span> mastodonDbSecretPath</span>;</span>
</span>        }</span>;</span>
</span>      }</span>;</span>
</span>    }</span>;</span>
</span>  }</span>;</span>
</span>}</span>
</span></code></pre>


How to use a non-default ssh port for a Nix distributed build host
2022-11-01T00:00:00+00:00
I wanted to host my ssh server on a different port than the default 22</code>, this allows me to skip on some spam in the logs, as the default port gets scanned quite often.
By changing that on the server I broke distributing my nix builds, as they were using the default port as well.
It took me a while to figure out how to configure the port the builder would use so I thought I would share here.</p>
Here's an example of a client-side</em> configuration -  using a build server buildHostName</code> with a user named nix-builder</code>, connecting via ssh to port 1234</code>.</p>
programs</span>.</span>ssh</span>.</span>extraConfig</span> =</span> ''</span>
</span></span>    Host buildHostName
</span></span>	HostName buildHostName
</span></span>	Port 1234
</span></span>	StrictHostKeyChecking=accept-new
</span></span>''</span></span>;</span>
</span>
</span>nix</span>.</span>buildMachines</span> =</span> [</span>{</span>
</span>    hostName</span> =</span> "</span>buildHostName"</span></span>;</span>
</span>    sshUser</span> =</span> "</span>nix-builder"</span></span>;</span>
</span>    sshKey</span> =</span> "</span>/path/to/key"</span></span>;</span>
</span>    systems</span> =</span> [</span> "</span>x86_64-linux"</span></span> ]</span>;</span>
</span>    maxJobs</span> =</span> 2</span>;</span>
</span>    speedFactor</span> =</span> 2</span>;</span>
</span>    supportedFeatures</span> =</span> [</span> "</span>kvm"</span></span> ]</span>;</span>
</span>    mandatoryFeatures</span> =</span> [</span> ]</span>;</span>
</span>}</span>]</span>;</span>
</span></code></pre>


legdur - keep your legacy durable
2022-06-25T00:00:00+00:00
Hey, I wrote a thing. Thing being a piece of software.
I have a collection of photos & documents that I really care about.
I synch them between computers using syncthing and also run backups regularly.
What I didn't have was a way to quickly detect bitrot.</p>
Enter legdur</a></p>
legdur</code> is a simple CLI program to compute hashes of large sets of files in large directory structures and compare them with a previous snapshot.
Think having your photo collection you acquired over time and worrying about bitrot.</p>
Installation</h2>
cargo install legdur --force</code> should get you there on a system that has Rust installed already.</p>
Try it out without installation</h2>
If you have a nix-flakes enabled system:
nix run git+https://git.cyplo.dev/cyplo/legdur.git</code></p>
Usage</h2>
legdur path/to/a/directory/</code></p>
working:</p>
legdur</span></span> ~</span></span>/documents</span>
</span>2022-06-25T06:45:51.000214Z</span></span>  INFO legdur: scanning '</span>/home/cyryl/documents'</span></span></span>
</span>2022-06-25T06:45:51.044471Z</span></span>  INFO legdur: list of files acquired, calculating hashes...</span>
</span>█████████████████████████████████████████████████████████████░░░░░░░░░░░░░░░░</span></span> 2788/3190</span>
</span></code></pre>
finished:</p>
2022-06-25T06:49:23.776229Z</span></span>  INFO legdur: hash calculation complete</span>
</span>2022-06-25T06:49:23.784499Z</span></span>  INFO legdur: /home/cyryl/documents/legdur.db saved</span>
</span>2022-06-25T06:49:23.792585Z</span></span>  INFO legdur: comparing /home/cyryl/documents/legdur.db with /home/cyryl/documents/legdur.old</span>
</span>2022-06-25T06:49:23.826548Z</span></span>  WARN legdur: /home/cyryl/documents/legdur.db: 395f65f5727f946c0208f79cfe1f3de1bd81e491bb7631ba6f41fc578d3db368 -></span> 7</span>d1b9748ed291eb6874c91917b5619eb8e1410e7cbfd37a517c5bd4ddf8c7895</span>
</span>2022-06-25T06:49:23.826715Z</span></span>  WARN legdur: /home/cyryl/documents/legdur.old: e56cf4f6b7fdc2daa3ca3430e8e64bf9d042f1c0e465dcd452986709cb25f7d8 -></span> 395</span>f65f5727f946c0208f79cfe1f3de1bd81e491bb7631ba6f41fc578d3db368</span>
</span></code></pre>
How it works</h2>

it will compute a hash of each file present in the directory structure (it works recursively).</li>
if previously computed legdur.db</code> exists it will compare the current state of the world to the one represented by legdur.db</code> and output any differences. Only files that changed or got deleted get printed out, additions or file moves anywhere within the directory are not.</li>
it will move the current legdur.db</code> to legdur.old</code> and write the new state of the world to a new legdur.db</code></li>
</ul>
Contact & contributions</h2>

Let me know if you'd like to hack on this by contacting me on legdur@cyplo.dev</code> or via cyplo@peninsula.industries</code> on the Fediverse.</li>
</ul>
Sources</h2>
git clone https://git.cyplo.dev/cyplo/legdur.git</code></p>


Quick reproducible zig dev env using nix flakes
2022-01-01T00:00:00+00:00
Want this great feeling of entering a directory and having all your dev setup done but can never remember how to set up nix flakes ?</p>
What you'll need is, in addition to a working direnv</a> with nix</code> and nix flake</code> enabled, is:</p>
.envrc</code>:</p>
use flake
</span></code></pre>
.gitignore</code>:</p>
.direnv/
</span></code></pre>
flake.nix</code> (this nixpkgs</code> hash points to the first revision with zig 0.9 present):</p>
{</span>
</span>  inputs</span> =</span> {</span>
</span>    nixpkgs</span> =</span> {</span>
</span>      type</span> =</span> "</span>github"</span></span>;</span>
</span>      owner</span> =</span> "</span>nixos"</span></span>;</span>
</span>      repo</span> =</span> "</span>nixpkgs"</span></span>;</span>
</span>      ref</span> =</span> "</span>e1eeb53e64159fbb2610ba7810ed511e4d5c10ca"</span></span>;</span>
</span>    }</span>;</span>
</span>    flake-compat</span> =</span> {</span>
</span>      url</span> =</span> "</span>github:edolstra/flake-compat"</span></span>;</span>
</span>      flake</span> =</span> false</span>;</span>
</span>    }</span>;</span>
</span>  }</span>;</span>
</span>
</span>  outputs</span> =</span> {</span> self</span>,</span> nixpkgs</span>,</span> flake-compat</span> }</span>:</span>
</span>    let</span> pkgs</span> =</span> nixpkgs</span>.</span>legacyPackages</span>.</span>x86_64-linux</span>;</span>
</span>    in</span>
</span>    {</span>
</span>      devShell</span>.x86_64-linux</span> =</span>
</span>        pkgs</span>.</span>mkShell</span> {</span>
</span>          buildInputs</span> =</span> with</span> pkgs</span>;[</span>
</span>            nixpkgs-fmt</span>
</span>            zig</span>
</span>            zls</span>
</span>          ]</span>;</span>
</span>        }</span>;</span>
</span>    }</span>;</span>
</span>}</span>
</span></code></pre>
and some shell.nix</code> for this extra bit of nix-shell</code> compatibility, if you'd want it:</p>
(</span>import</span>
</span>  (</span>
</span>    let</span>
</span>      lock</span> =</span> builtins</span>.</span>fromJSON</span> (</span>builtins</span>.</span>readFile</span> ./flake.lock</span>)</span>;</span>
</span>    in</span>
</span>    fetchTarball</span> {</span>
</span>      url</span> =</span> "</span>https://github.com/edolstra/flake-compat/archive/${</span>lock</span>.</span>nodes</span>.</span>flake-compat</span>.</span>locked</span>.</span>rev</span>}</span></span>.tar.gz"</span></span>;</span>
</span>      sha256</span> =</span> lock</span>.</span>nodes</span>.</span>flake-compat</span>.</span>locked</span>.</span>narHash</span>;</span>
</span>    }</span>
</span>  )</span>
</span>  {</span>
</span>    src</span> =</span> ./.</span>;</span>
</span>  }</span>)</span>.</span>shellNix</span>
</span></code></pre>
Don't forget to:</p>

git add</code> all of the above, otherwise nix flake operations might not work</li>
direnv allow</code> this directory</li>
have fun !</li>
</ul>


Using own OpenWRT router with Vodafone/Cityfibre
2021-11-04T00:00:00+00:00
For something very different this time, a quick recipe for a very specific situation.</p>
One of the very few fiber-to-the-home solutions in the UK is through Cityfibre, sold as Vodafone "Gigafast" packages where I live. As with other ISPs they give you their own internet router. If you would like to use your own, here's how.</p>
TLDR; Obtain PPPoE credentials specific to your account from Vodafone, then use them to configure your router, you need to tag all your packets with VLAN 911.</code></p>

Obtain PPPoE credentials from Vodafone. You can do so by using the chat feature on their website. Here's an example chat transcript pasted here for your enjoyment and pleasure</li>
</ol>
 Tobi : I’m TOBi - Vodafone’s chatbot Ask me anything, I’m here to help.
</span>You : hi, I would like to swap the provider router for the own router, could you share the PPPoE credentials I would need to use ? thank you
</span>Tobi : Please try asking your question again using fewer words, making sure you only ask one question at a time. You’ll be able to ask other questions later if you need to.
</span>You : ppoe credentials
</span>Tobi : TOBi is thinking...
</span>TOBi : Sorry, I didn't understand the question. Can you please try rephrasing?
</span>You : own router ppoe credentials
</span>TOBi : Sorry, I didn't understand the question. Can you please try rephrasing?
</span>You : can I speak to a human please ?
</span>TOBi : Could you tell me your full name please?
</span>You : xxxxxxxxxxxxxxx
</span>TOBi : Could you please tell me your phone number?
</span>You : xxxxxxxxxxx
</span>TOBi : Please tell me what type of account you have with us:
</span>TOBi :
</span>    Pay Monthly
</span>    Pay As You Go
</span>    Small Business
</span>    Broadband
</span>You : Broadband
</span>Pragya : Hi , you're now connected to Broadband chat support. Please give me 1-2 minutes as I quickly go through your conversation so that you don't have to repeat yourself.
</span>You : Hi, perfect, thank you :)
</span>Pragya : Cyryl , Thank you for waiting
</span>Pragya : As per the pre chat , I understand that you wish to have PPOE details , right , Cyryl ?
</span>You : that's correct, I'm trying to establish connection using own hardware
</span>Pragya : Cyryl , Thank you lets do a quick security check ,
</span>Please share your broadband account number or home phone number once so I can locate your account and proceed further.
</span>You : account number xxxxxxxxxx
</span>Pragya : Thank you very much
</span>Pragya : I have your account ready. Let's complete a quick security check so that I can try to resolve this for you.
</span>Pragya : Confirm the 2nd and 3rd digit of the security PIn ?
</span>You : xxxxxxx
</span>Pragya : Thank you let me verify it
</span>Pragya : Brilliant that matches the record
</span>Pragya : Would you mind being online for few minutes, while I am checking the details ?
</span>You : no problem, thank you
</span>Pragya : Thank you , Moving away from the chat window or not responding for 8 mins may result in losing the chat. If it does, here's a link to get back to us faster than usual: https://www.vodafone.co.uk/bbservcd
</span>Pragya : Thank you for waiting
</span>Pragya : Here is the username - dsl123456789@broadband.vodafone.co.uk
</span>Pragya : Password is - XXXXXXXX
</span>
</span></code></pre>

armed with that knowledge you can configure your router, I'm gonna give an example of configuring an OpenWRT router via their GUI interface LuCI</li>
</ol>

backup your settings</li>
-> Network->Interfaces</li>
note down the physical interface of current WAN interface - (WAN</code>->Edit->Physical Settings), e.g. eth2</code></li>
remove existing WAN</code> interface, press save and apply</code></li>
add a new PPPoE interface:

Name: WAN</code></li>
Protocol: PPPoE</code></li>
interface - type in the custom</code> field - name of your physical interface followed by a dot and 911</code> indicating the VLAN number to tag packets with - e.g. eth2.911</code></li>
-> create interface</li>
PAP/CHAP username: (username from Vodafone)</li>
PAP/CHAP password: (password from Vodafone)</li>
-> Firewall settings</li>
Create / Assign firewall-zone: WAN</code></li>
-> save</li>
-> save and apply</li>
wait around a minute for the router to connect</li>
</ul>
</li>
</ul>


My 2020 setup
2020-12-29T00:00:00+00:00

Update 2023 - updated links to the most current code store - git.cyplo.dev instead of sr.ht</p>
</blockquote>
Hello and welcome to the second edition of me doing a summary of the year of using tech tools.</p>
Here's a list of software and some hardware I find useful, either things that I use daily or things that make an unusual task pleasant instead of incredibly difficult.
This is constantly evolving, so please mind the publish date of this post, please also check the last year's installment</a> as this article will make references to the previous setup.</p>
Laptop, OS and other work hardware</h2>

Thinkpad T480</a>, the device-specific config lives here</a> (foureighty has been migrated to foryog).</li>
CalDigit TS3 Plus usb-c hub</li>
Iiyama ProLite 27" 4K monitor</li>
IKEA BEKANT motorised standing desk</li>
</ul>
No dramatic changes here since the last year</a>, I'm still on NixOS</a> on the T480. Thinkpad got a RAM upgrade to 48GiB and is handling it well, despite it being above it stated memory limit. I've also managed to get my home-manager</code> config called from the main system configuration, so now I just need to do sudo nixos-rebuild switch --upgrade</code> once and it does my home-manager</code> setup as well. This allowed me to split the config into modules a bit better.</p>
I'm really happy with the standing desk as it allows me to switch from sitting to standing and back very quickly.</p>
For the USB-C hub, I switched from the StarTech one I had before and this one is much more stable now.</p>
TODOs for 2021:</h3>

to look into flakes</code></a> and see if I want to port my config over to that style.</li>
</ul>
Networking</h2>

Turris Omnia</a></li>
Devolo Magic 2 WiFi</a> powerline adapters</li>
</ul>
While I stopped having WiFi speed problems with the purchase of the Netgear R7800 router, it started dropping packets in the summer. I suspect it started overheating, whether it was solely a problem with the hardware itself or was also influenced by where the router was placed and the airflow avaiable I don't know. As I needed this solved quickly I bought Turris Omnia as a replacement and am quite happy with it.</p>
TODOs for 2021:</h3>

debug the overheating problem</li>
try to port Liminix</a> (evolution of NixWRT) to run on R7800</li>
</ul>
Graphical interface</h2>

i3</a></li>
grobi</a></li>
</ul>
Running grobi</code> now instead of autorandr</code> but otherwise the config seems to be stabilising.</p>
Secrets management</h2>

veracrypt</a> + syncthing</a></li>
password store</a> + passff</a></li>
bitwarden</a></li>
</ul>
I have a veracrypt</code> encrypted container, where my secrets reside, with a small set of scripts to mount</a> and unmount</a> it. The container is synced between different machines using syncthing</code>.
Inside the container, among other things, there is a password store</code> directory, which I use from either command line or from Firefox.</p>
For when I need to share a secret I use bitwarden</code> as it allows for that in a quite an easy way.</p>
Sync</h2>
Syncthing</a> just keeps working, no matter how many devices I attach and what is their configuration. I run it on all my machines, including mobile devices and it just works. This is how I keep all my documents, photos and other data always fresh between all devices. Just make sure to encrypt the data at rest when using it, you don't want to sync to a device which someone else can take from you and read all the data off of.
Sync is also not a replacement for backup, as file deletions and corruption can spread easily across your fleet.</p>
Backups</h2>

restic</a></li>
</ul>
Here is where I am not that happy with the overall setup, and not that much has changed from 2019.
Currently I use restic</code> to package and encrypt the backup and then ship it off of individual machines to my central NAS storage. Then from there it is being copied to Backblaze's b2</a> for off-site storage.</p>
I made some progress on bakare</a>, a small backup engine in Rust, but it's not ready for production use yet. Let me know</a> if you would be interested in collaborating with me on it.</p>
TODOs for 2021:</h3>

finish bakare ?</li>
</ul>
Editors</h2>

vim configured via nix</a></li>
VSCode (no longer configured via nix in this setup)</li>
</ul>
Not much of a change here.</p>
Fonts</h2>
I settled on Fira Code Retina</a> for most of my programming and terminal needs.</p>
Terminal</h2>

kitty</a> and my config</a> for it</li>
zsh</a> + config</a></li>
tmux</a> + config</a></li>
</ul>
While I still like termite</code> and alacritty</code>, I have switched to kitty</code> lately as it is still quite fast while providing for some fancy shenanigans like graphics in the terminal. I'm not fully happy with my setup yet, as, for example, kitty sometimes requires restart for the graphics feature to work.</p>
Here's a small collection of other tools I found help a lot when on the terminal:</p>

newsboat</a> - a new one on the list, I find it providing a cool and relaxing way of reading RSS in the terminal, see here</a> for my config, including my subscriptions list</li>
ripgrep</a> - it is just so much faster than grep</li>
fd</a> - same but for find</code></li>
bat</a> - a cooler cat</code></li>
genpass</a> for generating passwords</li>
z.lua</a> for faster navigation</li>
</ul>
TODOs for 2021:</h3>

make kitty</code>'s image mode always work out of the box</li>
</ul>
Code hosting</h2>
I'm trying to migrate off of GitHub, I'm trying out sr.ht</a> now. I like its simplicity and ease of setup, especially when it comes to the built-in CI.</p>
The end</h2>
And that's it !</p>
I hope you'll find this list useful and do not hesitate to contact me</a> on blog@cyplo.net</code> if you would have any questions or comments. Happy hacking !</p>


Don't let failures spread over your suite with process-based tests isolation
2020-12-28T00:00:00+00:00
Being able to precisely control what failures in underlying systems occur and at what time can be really useful in achieving a fast and stable test suite. While I am a big proponent of dependency inversion and being able to control dependencies via the explicit injection points in your API, sometimes it's impractical to do so. This is where fail</code></a> can help us immensely, providing an escape hatch for situations like those as it allows to inject failures into previously defined failure points.</p>
It comes at a price though. If you would mix your other unit tests and tests activating fail points you will notice some unexpected failures in the test suite. As cargo test</code> runs tests in parallel by default, the tests activating a fail point can interfere with another test that did not want that fail point active at all that is ran at the same time. The crate authors recommend</a> running all of the tests using fail points in a separate executable and using FailScenario</code> to serialise test execution.</p>
There is another way, that I found simpler for the way I write tests, if you allow for yet another helper crate. We can run each test in a separate process, effectively isolating it from the rest, stopping failures from spreading.</p>
Let's take a look at an example from bakare</code></a> - my experiment in writing a backup system.</p>
cargo.toml</code></p>
[</span>dependencies</span></span>]</span>
</span>fail</span></span> =</span> "</span>0.4"</span></span>
</span>
</span>[</span>dev-dependencies</span></span>]</span>
</span>two-rusty-forks</span></span> =</span> "</span>0.4.0"</span></span>
</span>
</span>[</span>features</span></span>]</span>
</span>failpoints</span></span> =</span> [</span> "</span>fail/failpoints"</span></span> ]</span>
</span></code></pre>
lock.rs</code></p>
///</span> this function is called from `Lock::lock()`
</span></span>fn</span> </span>create_lock_file</span></span>(</span>lock_id</span>:</span> Uuid, index_directory</span>:</span> &</span>VfsPath</span>)</span></span></span></span> -></span> Result</span><</span>(</span>)</span>></span></span></span> </span>{</span>
</span></span></span>    ...</span>
</span></span></span>    fail_point!</span>(</span>"</span>create-lock-file"</span></span>,</span> |</span></span></span>e</span>:</span> Option</span><</span>String</span>></span></span>|</span></span> </span>Err</span>(</span>anyhow!</span>(</span>e.</span>unwrap</span>(</span></span>)</span></span></span>)</span></span></span>)</span></span></span></span>)</span></span>;</span>
</span></span></span>    let</span> mut</span> file =</span> lock_file_path.</span>create_file</span>(</span></span>)</span></span>?</span>;</span>
</span></span></span>    ...</span>
</span></span></span></span>}</span></span></span>
</span>
</span>mod</span> must</span> {</span>
</span></span></span>    use</span> super</span>::</span></span>Lock;</span>
</span></span></span>    use</span> anyhow::</span></span>Result;</span>
</span></span></span>
</span></span></span>    ///</span> only import the macro when `failpoints` feature is enabled
</span></span></span></span>    #</span>[</span>cfg</span>(</span></span></span>feature =</span> "</span>failpoints"</span></span></span></span>)</span></span></span>]</span></span>
</span></span></span>    use</span> two_rusty_forks::</span></span>rusty_fork_test;</span>
</span></span></span>    use</span> vfs::</span></span>{</span>MemoryFS,</span> VfsPath</span>}</span></span>;</span>
</span></span></span>
</span></span></span>    #</span>[</span>test</span>]</span></span>
</span></span></span>    ///</span> this is a normal unit test
</span></span></span></span>    ///</span> we don't want for it to be affected by the fail points being active
</span></span></span></span>    fn</span> </span>be_released_when_dropped</span></span>(</span></span>)</span></span></span></span> -></span> Result</span><</span>(</span>)</span>></span></span></span> </span>{</span>
</span></span></span></span></span>        let</span> temp_dir:</span> VfsPath =</span> MemoryFS::</span></span>new(</span></span>)</span></span>.</span>into</span>(</span></span>)</span></span>;</span>
</span></span></span></span></span>        {</span>
</span></span></span></span></span></span>            let</span> _lock =</span> Lock::</span></span>lock(</span>&</span>temp_dir</span>)</span></span>;</span>
</span></span></span></span></span></span>        </span>}</span></span>
</span></span></span></span></span>        let</span> entries =</span> temp_dir.</span>read_dir</span>(</span></span>)</span></span>?</span>.</span>count</span>(</span></span>)</span></span>;</span>
</span></span></span></span></span>
</span></span></span></span></span>        assert_eq!</span>(</span>entries,</span> 0</span></span>)</span></span>;</span>
</span></span></span></span></span>        Ok</span>(</span>(</span></span>)</span></span></span>)</span></span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span>
</span></span></span>    #</span>[</span>cfg</span>(</span></span></span>feature =</span> "</span>failpoints"</span></span></span></span>)</span></span></span>]</span></span>
</span></span></span>    rusty_fork_test!</span> {</span> ///</span> use the macro to create a separate process for this test
</span></span></span></span></span>        #</span>[</span>test</span>]</span></span>
</span></span></span></span>        fn</span> </span>be_able_to_lock_when_creating_lock_file_fails_sometimes</span></span>(</span></span>)</span></span></span></span> </span>{</span>
</span></span></span></span></span></span>            ///</span> activate the fail point
</span></span></span></span></span></span></span>            fail::</span></span>cfg(</span>"</span>create-lock-file"</span></span>,</span> "</span>90%10*return(some lock file creation error)->off"</span></span></span>)</span></span>
</span></span></span></span></span></span>                .</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span></span></span></span>            let</span> path =</span> MemoryFS::</span></span>new(</span></span>)</span></span>.</span>into</span>(</span></span>)</span></span>;</span>
</span></span></span></span></span></span>
</span></span></span></span></span></span>            let</span> lock =</span> Lock::</span></span>lock(</span>&</span>path</span>)</span></span>.</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span></span></span></span>            lock.</span>release</span>(</span></span>)</span></span>.</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span></span></span></span>        </span>}</span></span></span>
</span></span></span></span>    </span>}</span></span>
</span></span></span>
</span></span></span>    ...</span>
</span></span></span></span>}</span></span></span>
</span>
</span></code></pre>


Buttermilk cornbread
2020-10-10T00:00:00+00:00
I like baking, as it can be something that does not involve any computers.
Please have this recipe for cornbread, based on buttermilk and baked in a skillet.
As for baking goes, this recipe is quite forgiving and you don't need to watch ratios as closely as with others.</p>
Ingredients</h2>
This is for a 20cm (8") skillet, which gives you enough bread for you and one other person to enjoy over a weekend. For a 25cm (10") skillet - double the amounts. It's best to use a cast iron skillet as it has enough thermal mass to take pouring cold batter into it and still be reasonably hot afterwards. Flour and cornmeal volume measurements are for flour poured into a cup.</p>
1/2 tablespoon vegetable oil
</span>300ml cornmeal
</span>150ml all-purpose flour
</span>1/2 teaspoons salt
</span>1 teaspoons baking powder
</span>1/2 teaspoon baking soda
</span>300ml buttermilk
</span>2 eggs
</span>2 tablespoons melted butter
</span></code></pre>
Steps</h2>

Put the veggie oil in the skillet, put the skillet in the oven, preheat to 200C. This takes a while and gives you time to prepare the batter. The oil shouldn't smoke.</li>
Combine all dry ingredients in a large bowl, mix well.</li>
In another bowl whisk eggs and then melted butter into the buttermilk. No need for any extensive whisking, just to combine well.</li>
Make sure your oven came to temp.</li>
Pour wet ingredients into the dry, mix well. Acid from the buttermilk will start to react with bases from the baking soda and powder leavening the mixture. Do not overmix and work reasonably quickly, to not lose the bubblage. The resulting batter should be thick, barely pourable.</li>
Take the hot skillet out of the oven, make sure the oil is spread evenly.</li>
Pour the batter onto the hot skillet, most of it should pour but the last of it you may need to transfer using a spoon. Even out the surface, but don't be too pedantic about it.</li>
Put the skillet back into the oven, reduce heat to 185C</li>
Bake for around 20-30 minutes, it does not burn easily, you can put it under a broiler for few minutes at the end if you want more brownage. The time needed does not change as much with the volume.</li>
After baking you can keep it in the skillet, or take it out. It does not need to release much moisture, but also should come out pretty easily when fully baked.</li>
</ul>


My 2019 setup
2019-11-16T00:00:00+00:00

Update 2023: my dotfiles are self-hosted now at git.cyplo.dev</a>;
They also have a different structure than described here; I've removed now-dead links to github.</p>
</blockquote>
Here's a list of software and some hardware I find useful, either things that I use daily or things that make an unusual task pleasant instead of incredibly difficult.
This is constantly evolving, so please mind the publish date of this post to gauge how dated it is. The best place to get most recent settings I currently use is my dotfiles</a> repository, especially the nixos</a> configuration.</p>
Laptop and OS</h2>
My daily driver is Thinkpad T480</a> running NixOS</a>. I like the hardware for its matte screen, nice keyboard and general sturdiness. It's also black and black is nice.
The device-specific config lives here</a>.</p>
I use home manager</a> to manage my dotfiles.
I still haven't found a good way of having all this setup ported exactly to non-NixOS operating systems.
Within the NixOS land I'm all set up now though, as an example -
I have an older laptop that I used previously, and installing NixOS there took total of around half an hour, giving me exactly</strong> the same look and feel to all the software I run as I'm used to.</p>
NixOS's ability to boot into a previous version of the whole system is another big plus to me, as I like to experiment with my settings and sometimes I break something important. It creates an environment that does not punish you for trying something new, it encourages that.</p>
Networking gear</h2>
I had a lot of problems with network speed over wifi previously, because of that I bought a Netgear Nighthawk X4S R7800</a> router and installed OpenWRT</a> on it. After all this change, I am able to run very frequent multi-gigabyte backup transfers over wifi and everything works nicely.</p>
Graphical interface</h2>
I started running i3</a> recently because of the strain Gnome3 was putting on my system, and I am liking it so far, mostly for the speed of navigation and little resource usage. There's polybar</a> on top and not much else in terms of bells and whistles present.
I use autorandr</a> to keep track of different display devices on different machines - it will automatically set the best resolution for whatever the screen combination I am currently using. See config</a> for more details.</p>
Firefox</a> remains my browser of choice, I highly recommend you try it, it is so much faster now than it used to be. Make sure to switch the tracking protection on.</p>
Secrets management</h2>
I have a veracrypt</a> encrypted container, where my secrets reside, with a small set of scripts to mount</a> and unmount</a> it. The container is synced between different machines using syncthing</a>.
Inside the container, among other things, there is a password store</a> directory, which I use from either command line or from Firefox, using this plugin</a></p>
Sync</h2>
Syncthing</a> just keeps working, no matter how many devices I attach and what is their configuration. I run it on all my machines, including mobile devices and it just works. This is how I keep all my documents, photos and other data always fresh between all devices. Just make sure to encrypt the data at rest when using it, you don't want to sync to a device which someone else can take from you and read all the data off of.
Sync is also not a replacement for backup, as file deletions and corruption can spread easily across your fleet.</p>
Backups</h2>
Here is where I am not that happy with the overall setup.
Currently I use restic</a> to encrypt the backup and then ship it off of individual machines to my central NAS storage. Then from there it is being shipped to Backblaze's b2</a> for off-site storage.</p>
In principle, the setup I would like to retain, where the encryption credentials are only on machines creating backups and everything else only sees already encrypted files. In practice, restic itself seems to have a lot of troubles with the source machines being laptops and being constantly opened and closed, caused the running backup process to go through hibernation cycles. This locks/damages the central backup repo quite frequently and I need to run restic rebuild-index</code> quite often to keep things working.</p>
For this reason I started working on bakare</a>, a small backup engine in Rust - let me know</a> if you would be interested in collaborating with me on it.</p>
Editors</h2>
I use a combination of vim</a>, VSCode</a> and different JetBrains'</a> IDEs for work. I like IDEs mostly for refactoring and debugging capabilities, while vim and VSCode for speed of editing individual files. I still use vim-mode in IDEs though.</p>
Fonts</h2>
I settled on Fira Code Retina</a> for most of my programming and terminal needs.</p>
Terminal</h2>
I find termite</a> quite fast, while supporting extended character and colour sets.
My shell is zsh</a> with minimal oh-my-zsh</a> config</a>. I always run it inside a tmux</a> session though, and no matter how many terminal windows I open, I am always greeted with the same state. All history and window state is shared between all terminal windows all tmux windows as well - it is always the same one tmux session. Because I am always running tmux, sometimes I end up in a situation when I ssh into some box and need to run tmux there - for that reason I have my main tmux session having different leader key than the default, this way I can choose which tmux session will receive my command - my machine or the one I'm connecting to. All of the above comes from a combination of shell</a> and tmux</a> settings.</p>
Here's a small collection of other tools I found help a lot when on the terminal:</p>

ripgrep</a> - it is just so much faster than grep</li>
fd</a> - same but for find</code></li>
bat</a> - a cooler cat</code></li>
genpass</a> for generating passwords</li>
z.lua</a> for faster navigation</li>
</ul>
And that's it !
I hope you'll find this list useful and do not hesitate to contact me</a> on blog@cyplo.net</code> if you would have any questions or comments. Happy hacking !</p>


Fuzzers and how to run them.
2019-04-17T00:00:00+00:00
I am fascinated by the concept of fuzzing. It fits well with my desire to test weird code paths by using more of computer's time and less that of a programmer.</p>
What is fuzzing ?</h2>
It's a type of automated testing, especially good with finding edge cases in your code. It runs totally outside of your code and knows nothing about it - it just throws random data at it. Modern fuzzers instrument your code to be able to tell if by changing input they change the code paths covered and by doing that they try to achieve maximum coverage. While this sounds like a very naive approach it can lead to finding incredibly interesting bugs. For that reason fuzzers are oftentimes used in the security community for finding vulnerabilities. Fuzzing is a type of a black box testing - but how you define that box is entirely up to you. It can be the whole program, it can be a single function. It just needs a clear entry point with ability to manipulate input data.</p>
An example may be in order. We'll look into one provided by Rust Fuzzing Authority</a> - people behind cargo fuzz</a> and afl.rs</a>.
Imagine you're trying to see if you've handled all cases in your url parser. You can just do</p>
fuzz_target!</span>(</span>|</span></span></span>data</span>:</span> &</span>[u8</span>]|</span></span> </span>{</span>
</span></span></span></span>    if</span> let</span> Ok</span>(</span>s</span>)</span></span> =</span> std::</span></span>str::</span></span>from_utf8(</span>data</span>)</span></span> {</span>
</span></span></span></span></span>        let</span> _</span> =</span> url::</span></span>Url::</span></span>parse(</span>s</span>)</span></span>;</span>
</span></span></span></span></span>    </span>}</span></span>
</span></span></span></span></span>}</span></span></span></span>)</span></span>;</span>
</span></code></pre>
And that's it ! You define a function that takes raw bytes, discards the ones that are outside of a UTF8 string space and then tries to invoke the function under test with the said string. We assume that if this function completes then the run is considered ok, if it panics then not.</p>
I wanted to use this example because it is very typical of how a fuzzer would be used.</p>

get random input from the engine</li>
discard some of the input</li>
drive the function under test</li>
check some simple property of the run - the faster the property to check the better. For this reason - the 'did not crash' property is used often, however you can use any condition you want, as long as fuzzer has a way of distinguishing between successful and failed runs.</li>
</ol>
How does it compare to property-based tests ?</h2>
When thinking about other types of tests that are driven by randomness and are generally black-box-ish - property-based tests come to mind. How does fuzzing compare ?</p>

Fuzzing is good with finding hidden properties of the existing code, it used after the code is written, while property-based tests can also be used in a TDD workflow</li>
In property-based tests you typically want to describe behaviour - they serve as good documentation. Fuzzing is not really as good with that. It can automatically collect the examples though, that then can be manually transformed into a property.</li>
As you want your program to always hold all properties true you run all property-based tests within your CI/CD pipeline. For this reason they tend to test only things that are relatively fast to check, execution-time wise. Fuzzers are typically ran over nights or weeks - finding those edge cases, providing more properties for you to know and care about.</li>
</ul>
Want help with fuzzing your open source project ?</h2>
If you're not sure what fuzzing is, or you want help with adding fuzzer support to your open source project - just tell me ! Email</a> works best ! I'm happy to jump on a call, pair program or whatever you fancy ! I know the most about the Rust fuzzing space, but happy to help with other languages as well - this allows me to learn :)</p>
I also run a fuzzing server where I can test your project overnight and see if we find anything.</p>
With that in mind - happy fuzzing !</p>


Website refresh
2019-04-09T00:00:00+00:00
Hello !</p>
As you may have noticed - this website looks different now ! Why is that ? I've not only changed the visual theme but also a lot of underlying infrastructure.
Let's start with describing the old setup and see where we can improve.
The site previously ran on Nikola</a>, was built on Travis</a> and then pushed to Netlify</a>, which I later changed to Github Pages</a>. While it worked it had some issues of its own;</p>

Image galleries were hard to navigate and looked a bit out of place.</li>
The site was not rendering the main menu correctly on mobile devices.</li>
It was slow to build the site, especially on CI - it took 15 to 30 minutes for the CI run on travis to get the site deployed - not ideal for fast feedback</li>
Lack of previews for the work in progress - it was hard for me to set up a preview per branch, updated automatically.</li>
</ol>
All of the above, mixed with me having a bit of a time off, resulted in getting this site migrated to some new and exciting stack !</p>
It now runs on Hugo</a> and uses Netlify as the deployment target. I fixed all of the problems above and added some more niceties !</p>

I'm using Beautiful Hugo</a>'s Gallery now - see this old post</a> for an example</li>
This theme also makes the site mobile-friendly out of the box</li>
It takes 1-3 minutes for the full site deploy</li>
Previews just work :)</li>
No need to use custom Netlify sync script I wrote</a> anymore, as Hugo is natively supported on Netlify</li>
Netlify supports TLS on multiple domains - you can now use cyplo.dev</a> as well as cyplo.net</a> to reach this site :)</li>
</ol>
As for how the actual migration was done - it was a bit of work, as you can see in this PR</a>. While I could migrate the content mostly automatically, I chose to actually use this opportunity to review all the posts manually !
This allowed me to spot and fix some additional issues, like broken links, editorial mistakes etc.</p>
Overall I think this was worth it - the site is now more modern, the sources are smaller and it is fast to build.
If you're curious about the details - all the sources are here</a>.</p>
Happy blogging !</p>


Generate Rust tests from data files
2018-12-25T00:00:00+00:00
Sometimes you just have a bunch of example data laying around and you want to make sure your code works with all of them.
Some of them are probably short and sweet and could live happily as doctests</a>, which are amazing btw. But some of them are more awkward to present in such form, because, for example, of their size or number. Typically when you have an example of how the program should behave you write an example-based unit test. Ideally, each of them would represent an isolated example and they should fail independently. But, converting your source data files into a unit test one by one, manually, can be a bit tedious.</p>
Rust build scripts</a> to the rescue !</p>
What if you could could just iterate over the data files you have already and then produce unit tests accordingly ?
What follows is an example of such, where we iterate over directories and generate one unit test per each, assuming all of them contain files named according to our convention.</p>
I chose to generate integration tests here, but you can generate pretty much any code using this technique.</p>
tests/test_loader.rs</code></p>
//</span> include tests generated by `build.rs`, one test per directory in tests/data
</span></span>include!</span>(</span>concat!</span>(</span>env!</span>(</span>"</span>OUT_DIR"</span></span></span>)</span></span>,</span> "</span>/tests.rs"</span></span></span>)</span></span></span>)</span></span>;</span>
</span></code></pre>
build.rs</code></p>
use</span> std::</span></span>env;</span>
</span>use</span> std::</span></span>fs::</span></span>read_dir;</span>
</span>use</span> std::</span></span>fs::</span></span>DirEntry;</span>
</span>use</span> std::</span></span>fs::</span></span>File;</span>
</span>use</span> std::</span></span>io::</span></span>Write;</span>
</span>use</span> std::</span></span>path::</span></span>Path;</span>
</span>
</span>//</span> build script's entry point
</span></span>fn</span> </span>main</span></span>(</span></span>)</span></span></span></span> </span>{</span>
</span></span></span>    let</span> out_dir =</span> env::</span></span>var(</span>"</span>OUT_DIR"</span></span></span>)</span></span>.</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span>    let</span> destination =</span> Path::</span></span>new(</span>&</span>out_dir</span>)</span></span>.</span>join</span>(</span>"</span>tests.rs"</span></span></span>)</span></span>;</span>
</span></span></span>    let</span> mut</span> test_file =</span> File::</span></span>create(</span>&</span>destination</span>)</span></span>.</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span>
</span></span></span>    //</span> write test file header, put `use`, `const` etc there
</span></span></span></span>    write_header</span>(</span>&</span>mut</span> test_file</span>)</span></span>;</span>
</span></span></span>
</span></span></span>    let</span> test_data_directories =</span> read_dir</span>(</span>"</span>./tests/data/"</span></span></span>)</span></span>.</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span>
</span></span></span>    for</span> directory in</span> test_data_directories {</span>
</span></span></span></span>        write_test</span>(</span>&</span>mut</span> test_file,</span> &</span>directory.</span>unwrap</span>(</span></span>)</span></span></span>)</span></span>;</span>
</span></span></span></span>    </span>}</span></span>
</span></span></span></span>}</span></span></span>
</span>
</span>fn</span> </span>write_test</span></span>(</span>test_file</span>:</span> &</span>mut</span> File, directory</span>:</span> &</span>DirEntry</span>)</span></span></span></span> </span>{</span>
</span></span></span>    let</span> directory =</span> directory.</span>path</span>(</span></span>)</span></span>.</span>canonicalize</span>(</span></span>)</span></span>.</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span>    let</span> path =</span> directory.</span>display</span>(</span></span>)</span></span>;</span>
</span></span></span>    let</span> test_name =</span> format!</span>(</span></span>
</span></span></span></span>            "</span>prefix_if_needed_{}</span>"</span></span></span>,</span>
</span></span></span></span>            directory.</span>file_name</span>(</span></span>)</span></span>.</span>unwrap</span>(</span></span>)</span></span>.</span>to_string_lossy</span>(</span></span>)</span></span>
</span></span></span></span>        )</span></span>;</span>
</span></span></span>
</span></span></span>    write!</span>(</span></span>
</span></span></span></span>        test_file,</span>
</span></span></span></span>        </span>include_str!</span>(</span>"</span>./tests/test_template"</span></span></span>)</span></span>,</span>
</span></span></span></span>        name =</span> test_name,</span>
</span></span></span></span>        path =</span> path
</span></span></span></span>    )</span></span>
</span></span></span>    .</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span></span>}</span></span></span>
</span>
</span>fn</span> </span>write_header</span></span>(</span>test_file</span>:</span> &</span>mut</span> File</span>)</span></span></span></span> </span>{</span>
</span></span></span>    write!</span>(</span></span>
</span></span></span></span>        test_file,</span>
</span></span></span></span>        r</span>#</span>"
</span></span></span></span></span>use crate_under_test::functionality_under_test;
</span></span></span></span></span>"#</span></span>
</span></span></span></span>    </span>)</span></span>
</span></span></span>    .</span>unwrap</span>(</span></span>)</span></span>;</span>
</span></span></span></span>}</span></span></span>
</span></code></pre>
tests/test-template</code></p>

</span>#</span>[</span>test</span>]</span></span>
</span>fn</span> {</span>name</span>}</span></span>(</span></span>)</span></span> {</span>{</span>
</span></span></span>    let</span> input =</span> include_str!</span>(</span>"</span>{path}/input-data"</span></span></span>)</span></span>;</span>
</span></span></span>    let</span> expected_output =</span> include_str!</span>(</span>"</span>{path}/output-data"</span></span></span>)</span></span>;</span>
</span></span></span>
</span></span></span>    let</span> actual_output =</span> functionality_under_test</span>(</span>input</span>)</span></span>;</span>
</span></span></span>
</span></span></span>    assert_eq!</span>(</span>expected_output,</span> actual_output</span>)</span></span>;</span>
</span></span></span></span>}</span></span></span>}</span></span>
</span>
</span></code></pre>
So to recap - first the build.rs</code> script creates $OUT_DIR/tests.rs</code> file containing all the generated tests code.
The compiler does not know there are tests to launch using normal integration tests procedure there though,
so then we use tests/test_loader.rs</code> to tell it so, basically including the generated Rust code into that file.
After the compilation proceeds normally, giving us one unit test per directory, giving us ability to pinpoint test cases that are problematic more precisely.</p>
You can then further improve on that, e.g. add more directory structure, split tests into modules etc - you can generate any Rust code this way.</p>
Happy hacking !</p>
p.s.
there are more Rust testing tricks</a> and let me know if you'd like to pair program with me on anything !</a></p>


Testing tricks in Rust
2018-09-25T00:00:00+00:00
Use verbs as test module names</h1>
Who said that the test module needs to be named test</code> ?
Experiment with different module names, pay attention to how the test runner displays the results.</p>
A structure that I like, an example:</p>
worker.rs:</code></p>
//</span> some production code here
</span></span>
</span>mod</span> should</span> {</span>
</span></span></span>
</span></span></span>    #</span>[</span>test</span>]</span></span>
</span></span></span>    fn</span> </span>consume_message_from_queue</span></span>(</span></span>)</span></span></span></span> </span>{</span>
</span></span></span></span></span>        //</span> mock queue, create worker with that queue injected
</span></span></span></span></span></span>        //</span> start worker
</span></span></span></span></span></span>        //</span> check if queue's 'get_message' was invoked
</span></span></span></span></span></span>    </span>}</span></span></span>
</span></span></span>
</span></span></span></span>}</span></span></span>
</span></code></pre>
Cargo prints worker::should::consume_message_from_queue</code> when running this test, which reads nicely and exposes the requirement.</p>
Interior mutability for controlling state of variables injected from the test</h1>
Use e.g. the atomic</code> types family or RefCell</code> itself to get an immutable handle to a internally mutable data.
Useful when you don't want your production code to accept anything that can mutate but you still want to control that value from the test.</p>
See injecting the system clock example in my previous blog post</a>.</p>
Write the test first</h1>
Not really a Rust trick, but hey.
Try writing your test first, before production code.
If you're building a feature or fixing a bug that will affect external behaviour - start with an integration test at the crate level.</p>
Try thinking what would be the ideal code you would like to interact with, what would be the types, what would be the functions ? A broad-strokes scenario, not caring much about implementation details, not caring much about covering all edge cases. Write that code. It does not compile. But it looks nice, you're pleased.</p>
Read through again, add assertions. Add the types. For each missing feature or a bug that is present in this high level scenario - write a unit test. Satisfy that test with changes to production code. Maybe refactor a bit in between. Once the big test is green - you're done !</p>
There is no Rust-focused TDD book just yet for me to recommend, but here, have some for other languages:</p>

Kent Beck - Test Driven Development: By Example</li>
Steve Freeman, Nat Pryce - Growing Object-Oriented Software, Guided by Tests</li>
</ul>
Rust allows for more cool tricks and generally writing less test code than mentioned in these books, so please use your judgment - and the tricks from this article !</p>
Let's talk !</h1>
Have any questions ? Would like to pair on Rust ? Curious about TDD ? Ping me !
Email is good - hello@cyplo.net</a>.</p>
thanks !</p>


Rust - controlling side effects from the test.
2018-07-30T00:00:00+00:00
Rust: controlling side effects from the test.</h1>
Hello and welcome to the newest episode on testing in Rust.
Imagine you want to write a timestamping repository of some sorts, that will associate the timestamp of when the storage operation was invoked with the stored value.
How to write it in Rust ? And more importantly - how to test it ?
I would like to share a solution I found and talk a bit about how it works.</p>
Please note that this solution can be used anywhere where you need to pass a handle that is remembered by the production code, and that thing it points to - you then want to change from the test.</p>
trait</span> Clock</span> {</span>
</span></span></span>    fn</span> </span>now</span></span>(</span>&</span>self</span></span>)</span></span></span></span> -></span> Instant</span></span>;</span>
</span></span></span></span>}</span></span></span>
</span>
</span>struct</span> </span>SystemClock</span></span>;</span>
</span>
</span>impl</span> </span>SystemClock</span> </span>{</span>
</span></span></span>    fn</span> </span>new</span></span>(</span></span>)</span></span></span></span> -></span> Self</span></span> </span>{</span>
</span></span></span></span></span>        SystemClock {</span></span>}</span></span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span></span>}</span></span></span>
</span>
</span>impl</span> </span>Clock for</span></span> SystemClock</span> </span>{</span>
</span></span></span>    fn</span> </span>now</span></span>(</span>&</span>self</span></span>)</span></span></span></span> -></span> Instant</span> </span>{</span>
</span></span></span></span></span>        Instant::</span></span>now(</span></span>)</span></span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span></span>}</span></span></span>
</span>
</span>struct</span> </span>TimestampingRepository</span><</span>'a</span>, ClockType></span></span></span></span>
</span></span>where
</span></span>    ClockType: Clock + 'a,
</span></span></span>{</span>
</span></span></span>    clock</span>:</span> &</span>'a</span> ClockType,
</span></span></span>    storage</span>:</span> Vec</span><</span>(</span>Instant, u32</span>)</span>></span></span>, //</span> (timestamp, value)
</span></span></span></span></span>}</span></span></span>
</span>
</span>impl</span></span><</span>'a</span>, ClockType></span></span></span> TimestampingRepository</span><</span>'a</span>, ClockType></span></span>
</span></span></span>where</span>
</span></span>    ClockType:</span> Clock + 'a</span>,
</span></span></span>{</span>
</span></span></span>    fn</span> </span>with_clock</span></span>(</span>clock</span>:</span> &</span>'a</span> ClockType</span>)</span></span></span></span> -></span> Self</span></span> </span>{</span>
</span></span></span></span></span>        TimestampingRepository {</span>
</span></span></span></span></span></span>            clock,</span>
</span></span></span></span></span></span>            storage:</span> vec!</span>[</span>]</span></span>,</span>
</span></span></span></span></span></span>        </span>}</span></span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span>
</span></span></span>    fn</span> </span>store</span></span>(</span>&</span>mut</span> self</span>, value</span>:</span> u32</span></span>)</span></span></span></span> </span>{</span>
</span></span></span></span></span>        self</span>.</span>storage.</span>push</span>(</span>(</span>self</span>.</span>clock.</span>now</span>(</span></span>)</span></span>,</span> value</span>)</span></span></span>)</span></span>;</span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span>
</span></span></span>    fn</span> </span>all_stored</span></span>(</span>&</span>self</span></span>)</span></span></span></span> -></span> Vec</span><</span>(</span>Instant, u32</span>)</span>></span></span></span> </span>{</span>
</span></span></span></span></span>        self</span>.</span>storage.</span>clone</span>(</span></span>)</span></span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span></span>}</span></span></span>
</span>
</span>#</span>[</span>cfg</span>(</span></span></span>test</span></span>)</span></span></span>]</span></span>
</span>mod</span> should</span> {</span>
</span></span></span>
</span></span></span>    #</span>[</span>test</span>]</span></span>
</span></span></span>    fn</span> </span>handle_seconds</span></span>(</span></span>)</span></span></span></span> </span>{</span>
</span></span></span></span></span>        let</span> clock =</span> FakeClock::</span></span>with_time(</span>Instant::</span></span>now(</span></span>)</span></span></span>)</span></span>;</span>
</span></span></span></span></span>        let</span> mut</span> repository =</span> TimestampingRepository::</span></span>with_clock(</span>&</span>clock</span>)</span></span>;</span>
</span></span></span></span></span>
</span></span></span></span></span>        repository.</span>store</span>(</span>1</span></span>)</span></span>;</span>
</span></span></span></span></span>        clock.</span>move_by</span>(</span>Duration::</span></span>from_secs(</span>32</span></span>)</span></span></span>)</span></span>;</span>
</span></span></span></span></span>        repository.</span>store</span>(</span>2</span></span>)</span></span>;</span>
</span></span></span></span></span>
</span></span></span></span></span>        let</span> time_difference =</span> time_difference_between_two_stored</span>(</span>repository</span>)</span></span>;</span>
</span></span></span></span></span>
</span></span></span></span></span>        assert_eq!</span>(</span>32</span>,</span> time_difference.</span>as_secs</span>(</span></span>)</span></span></span>)</span></span>;</span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span>
</span></span></span>    struct</span> </span>FakeClock</span> </span>{</span>
</span></span></span></span></span>        now</span>:</span> Instant,
</span></span></span></span></span>        move_by_secs</span>:</span> AtomicUsize,
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span>
</span></span></span>    impl</span> </span>FakeClock</span> </span>{</span>
</span></span></span></span></span>        fn</span> </span>with_time</span></span>(</span>now</span>:</span> Instant</span>)</span></span></span></span> -></span> Self</span></span> </span>{</span>
</span></span></span></span></span></span></span>            FakeClock {</span>
</span></span></span></span></span></span></span></span>                now,</span>
</span></span></span></span></span></span></span></span>                move_by_secs:</span> AtomicUsize::</span></span>new(</span>0</span></span>)</span></span>,</span>
</span></span></span></span></span></span></span></span>            </span>}</span></span>
</span></span></span></span></span></span></span>        </span>}</span></span></span>
</span></span></span></span></span>
</span></span></span></span></span>        //</span> WAT no `mut`
</span></span></span></span></span></span>        fn</span> </span>move_by</span></span>(</span>&</span>self</span>, duration</span>:</span> Duration</span>)</span></span></span></span> </span>{</span>
</span></span></span></span></span></span></span>            self</span>.</span>move_by_secs
</span></span></span></span></span></span></span>                .</span>store</span>(</span>duration.</span>as_secs</span>(</span></span>)</span></span> as</span> usize</span>,</span> Ordering::</span></span>SeqCst</span>)</span></span>;</span>
</span></span></span></span></span></span></span>        </span>}</span></span></span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span>
</span></span></span>    impl</span> </span>Clock for</span></span> FakeClock</span> </span>{</span>
</span></span></span></span></span>        fn</span> </span>now</span></span>(</span>&</span>self</span></span>)</span></span></span></span> -></span> Instant</span> </span>{</span>
</span></span></span></span></span></span></span>            let</span> move_by_secs =</span> self</span>.</span>move_by_secs.</span>load</span>(</span>Ordering::</span></span>SeqCst</span>)</span></span> as</span> u64</span>;</span>
</span></span></span></span></span></span></span>            self</span>.</span>now +</span> Duration::</span></span>from_secs(</span>move_by_secs</span>)</span></span>
</span></span></span></span></span></span></span>        </span>}</span></span></span>
</span></span></span></span></span>    </span>}</span></span></span>
</span></span></span>
</span></span></span></span>}</span></span></span>
</span></code></pre>
That's a lot of code. And I already skipped use</code>s and some definitions to make it less.
If you want to get the full source code that to follow along - try this playground</a> or this repo</a> for the full project including production code usage.</p>
Let's start with the test itself.
The clock appears to be immutable (immovable) in the test, yet we call move_by</code> on it and the whole thing appears to be working somehow.
First question: can't we just make the clock mutable and skip all this ?
It appears that sadly (but fortunately) Rust prevents us from doing so.
We cannot</a> both have a immutable and mutable borrow of the clock in the same scope.
For the full example with an error go here</a>.</p>
What is this sorcery then ?
We use a type that provides Interior Mutability</code></a>, namely AtomicUsize</code></a>.
On the outside - it look immutable, yet it provides a thread-safe and very narrow method of mutating the underlying state.
As we trust AtomicUsize</code> to be written correctly, we can then proceed and write our Rust code as usual, relying fully on the borrow checker. Rust compiler is happy and our test code is happy.</p>
I wouldn't use this as a pattern in production code - the borrow checker rules are there for a reason.
Please treat it as an escape hatch to be used in specific situations, situations like this.</p>
Happy Rusting !</p>
p.s. if you'd like to chat about Rust - ping me an email</a> !</p>


Resources for starting your adventure with Rust
2018-05-22T00:00:00+00:00
As I've been running several intro to Rust sessions throughout the last year, I've assembled a set of resources that help people ease into the language.</p>
Depending on your learning style you might like:</p>

Rustlings</a> - This is a good set of starter exercises if you want to have a feeling for the language - have links to relevant book sections for each exercises so you can either start with the book or trying to figure it out yourself first. Ah, and it uses the Playground, which means you don't need to install anything on your machine to start.</li>
The book itself</a> - Second edition. Good when you want a solid baseline understanding of the language first.</li>
Rust by example</a> - An set of examples that are runnable within a browser, intertwined with explanatory prose.</li>
Exercism</a>’s Rust exercises - a CLI app that guides you through exercises of increasing difficulty.</li>
IntoRust</a> - A set of short screencasts for the foundational topics.</li>
</ul>
Make sure to stay up to date with:</p>

This week in Rust</a></li>
Awesome Rust</a></li>
</ul>
And finally:</p>

contribute back</a> !</li>
don't forget to join the user forums</a> for the warm welcome.</li>
</ul>
If you'd like someone to ask questions to or pair program with, ping me an email</a>.</p>


I’m running Rust pair programming sessions !
2018-05-20T00:00:00+00:00
Why ? Rust has such a wonderful community and I want to give back as much as I can.
I am not an expert in Rust but I am not a beginner either. In addition to that I love pair programming !
The result is always much better than I could produce myself. I am happy to both share the knowledge and learn.</p>
I would love to pair with you !
If you’re a new Rustacean, fresh to the language - come on in ! If you’re an expert - welcome !</p>
We can work on any of the following:</p>

Any project of yours !</li>
Contribute back to a larger open source project (I am a contributor to e.g. cargo</a>, rustc</a> and rustup</a>)</li>
A project of mine - e.g. genpass</a></li>
</ul>
Ping me an email</a> to schedule a session - can be a remote one or in person somewhere in London.</p>
Thank you !</p>


Configure AWS Elastic Beanstalk Docker environment variables
2018-04-23T12:00:00+00:00
AWS Beanstalk is a good 'intermediate' level hosting for Docker containers. It gives you load balancing and scalability pretty much out of the box in exchange for being a bit more opaque to configure. The Docker bits are a bit more hidden away there.
In a typical production setup you would want to have Docker images not containing anything environment related, e.g. to be able to run them both in production and locally. An easy way to achieve that with Docker is via environment variables. On the local environment it's docker run --env NAME=VALUE</code> - what would be a Beanstalk equivalent though ?</p>
It turns out that Beanstalk has a magical configuration directory structure that you can pass to an environment.
It goes like this:</p>
configuration.zip
</span>   Dockerrun.aws.json
</span>   .ebextensions/
</span>       environmentvariables.config
</span></code></pre>
Where Dockerrun.aws.json</code> is your regular Docker definition file for Beanstalk, can look like this:</p>
{
</span>    "AWSEBDockerrunVersion": "1",
</span>    "Image": {
</span>        "Name": "image:latest",
</span>        "Update": "true"
</span>    },
</span>    "Ports": [
</span>        {
</span>        "ContainerPort": "1234"
</span>        }
</span>    ]
</span>}
</span></code></pre>
While .ebextensions/environmentvariables.config</code> is where, well, you set the environment variables that will be defined in the container. Example:</p>
option_settings:
</span>  - option_name: ENV_VAR1
</span>    value: "some value"
</span>  - option_name: ENV_VAR2
</span>    value: "some other value"
</span></code></pre>
But wait, there's more ! Get the zip file and upload it to some S3 bucket, I'm going to assume that the file is at BUCKET_NAME/CONFIG_PATH</code> in the example below.
Then you need to tell Beanstalk where the file is located. This can be achieved by creating a new application version:</p>
aws elasticbeanstalk create-application-version --application-name APPLICATION_NAME --version-label VERSION --source-bundle S3Bucket=BUCKET_NAME,S3Key=CONFIG_PATH
</span>aws elasticbeanstalk update-environment --environment-name ENVIRONMENT_NAME --version-label VERSION
</span></code></pre>


Waiting for AWS Elastic Beanstalk environment to become ready
2018-04-23T00:00:00+00:00
Elastic Beanstalk on AWS seems to be one of those services that are pretty cool but it's hard to get to know them.
One of the tasks you may encounter while working with it is that after making some change to its configuration you would like to wait for it to be finished before proceeding further. The change may be setting an environment variable but can also be deploying a new version of the application. I created a small bash script to help with that, can be useful when you try to run this process unattended, e.g. from CI.</p>
#</span></span>!/bin/bash</span>
</span></span>set</span></span> -</span>e</span></span>
</span>set</span></span> -</span>o</span> pipefail</span>
</span>
</span>application_name</span>=</span>$</span>1</span></span></span>
</span>environment_name</span>=</span>$</span>2</span></span></span>
</span>timeout_seconds</span>=</span>$</span>3</span></span></span>
</span>
</span>function</span> getStatus</span>(</span>)</span> {</span>
</span></span>echo</span></span> `</span>aws</span></span> elasticbeanstalk describe-environments \
</span></span></span></span></span></span>    --</span>application-name</span> $</span>application_name</span></span> --</span>environment-name</span> $</span>environment_name</span></span></span> |</span>\
</span></span></span></span></span>    jq</span></span> -</span>r</span> '</span>.Environments | .[]?'</span></span></span> |</span> jq</span></span> -</span>r</span> '</span>.Status'</span></span></span>`</span></span></span>
</span></span>}</span></span>
</span>
</span>sleep_time_seconds</span>=</span>5</span>
</span>max_iterations_count</span>=</span>$</span>((</span>$</span>timeout_seconds</span></span>/</span>$</span>sleep_time_seconds</span></span>))</span></span></span>
</span>iterations</span>=</span>0</span>
</span>
</span>echo</span></span> "</span>Waiting for a maximum of $</span>timeout_seconds</span></span> seconds for $</span>environment_name</span></span> to become ready"</span></span></span>
</span>status</span>=</span>$</span>(</span>getStatus</span></span>)</span></span></span>
</span>while</span> [[</span> ( $</span>status</span></span> !=</span> "</span>Ready"</span></span> ) &&</span> ( $</span>iterations</span></span> -</span>lt</span> $</span>max_iterations_count</span></span> ) ]]</span></span>;</span> do</span>
</span>    status</span>=</span>$</span>(</span>getStatus</span></span>)</span></span></span>
</span>    echo</span></span> $</span>status</span></span></span>
</span>    sleep</span></span> $</span>sleep_time_seconds</span></span></span>
</span>    ((</span>iterations+=</span>1</span>))</span></span>
</span>done</span>
</span></code></pre>
Happy coding !</p>


Setting up Rust development environment using VSCode on a Mac
2017-11-25T13:50:27+00:00
Completion and highlighting</h2>
While on Linux VSCode with the Rust plugin seems to work more or less out of the box, on a Mac I needed to spend some time configuring it.</p>
First things first though, let's start by installing Rust version manager, rustup.</p>
curl https://sh.rustup.rs -sSf | sh
</span></code></pre>
We will be using nightly version of rust as to have one version that can compile all of our tools. This is mostly due to clippy</code></a> requiring a nightly compiler.</p>
rustup install nightly
</span>rustup default nightly
</span></code></pre>
We will need Rust Language Server to provide the code completion.</p>
rustup component add rls-preview --toolchain nightly
</span>rustup component add rust-analysis --toolchain nightly
</span>rustup component add rust-src --toolchain nightly
</span></code></pre>
For a more wholesome experience, please have some tools as well:</p>
cargo install clippy rustfmt rustsym
</span></code></pre>
Now finally, for the VSCode itself, press cmd-p</code> and ext install vscode-rust</code>. I'm using the new Rust</code> extension as Rusty Code</code> has been discontinued.</p>
If you're lucky - that's it, you should have working completion and highlighting in Rust files. Check this by opening any Rust source code file.
If you're greeted by this message: You have chosen RLS mode but neither RLS executable path is specified nor rustup is installed</code> - then we need to get the extension to get to know your setup a bit:</p>
In VSCode go to Settings</code> using cmd-,</code> and put the following config elements there:</p>
{</span>
</span></span>  </span>"</span>rust.cargoPath"</span></span></span>:</span> </span>"</span>/Users/yourusername/.cargo/bin/cargo"</span></span>,</span>
</span></span>  </span>"</span>rust.cargoHomePath"</span></span></span>:</span> </span>"</span>/Users/yourusername/.cargo"</span></span>,</span>
</span></span>  </span>"</span>rust.rustfmtPath"</span></span></span>:</span> </span>"</span>/Users/yourusername/.cargo/bin/rustfmt"</span></span>,</span>
</span></span>  </span>"</span>rust.rustsymPath"</span></span></span>:</span> </span>"</span>/Users/yourusername/.cargo/bin/rustsym"</span></span>,</span>
</span></span>  </span>"</span>rust.rustLangSrcPath"</span></span></span>:</span> </span>"</span>/Users/yourusername/.rustup/toolchains/nightly-x86_64-apple-darwin/lib/rustlib/src/rust/src"</span></span>,</span>
</span></span>  </span>"</span>rust.mode"</span></span></span>:</span> </span>"</span>rls"</span></span>,</span>
</span></span>  </span>"</span>rust.rls"</span></span></span>:</span> </span>{</span>
</span></span></span>    </span>"</span>executable"</span></span></span>:</span> </span>"</span>/Users/yourusername/.cargo/bin/rls"</span></span>,</span>
</span></span></span>    </span>"</span>useRustfmt"</span></span></span>:</span> </span>true</span>
</span></span></span>  }</span></span>
</span></span>}</span></span>
</span></code></pre>
As the paths in the config need to be absolute, remember to adjust to your situation (system username) accordingly.</p>
Now when you reload and start editing a Rust file you should see RLS: Analysis finished</code> on the bottom bar and the completion and highlighting should all work. Yay !</p>
Building and testing</h2>
VSCode has a system of tasks that we can leverage to run the build and test from within VSCode.
If you go to Tasks->Configure</code> tasks it will create an empty tasks.json</code> file in your repository.
Change it to the following to allow for cargo</code> to be hooked up as your build tool and test runner.</p>
{</span>
</span></span>  </span>"</span>version"</span></span></span>:</span> </span>"</span>2.0.0"</span></span>,</span>
</span></span>  </span>"</span>tasks"</span></span></span>:</span> </span>[</span>
</span></span></span>    {</span>
</span></span></span></span>      </span>"</span>label"</span></span></span>:</span> </span>"</span>build"</span></span>,</span>
</span></span></span></span>      </span>"</span>type"</span></span></span>:</span> </span>"</span>shell"</span></span>,</span>
</span></span></span></span>      </span>"</span>command"</span></span></span>:</span> </span>"</span>cargo build"</span></span>,</span>
</span></span></span></span>      </span>"</span>group"</span></span></span>:</span> </span>{</span>
</span></span></span></span></span>        </span>"</span>kind"</span></span></span>:</span> </span>"</span>build"</span></span>,</span>
</span></span></span></span></span>        </span>"</span>isDefault"</span></span></span>:</span> </span>true</span>
</span></span></span></span></span>      }</span></span>,</span>
</span></span></span></span>      </span>"</span>problemMatcher"</span></span></span>:</span> </span>[</span>]</span></span>
</span></span></span></span>    }</span></span>,</span>
</span></span></span>    {</span>
</span></span></span></span>      </span>"</span>label"</span></span></span>:</span> </span>"</span>test"</span></span>,</span>
</span></span></span></span>      </span>"</span>type"</span></span></span>:</span> </span>"</span>shell"</span></span>,</span>
</span></span></span></span>      </span>"</span>command"</span></span></span>:</span> </span>"</span>cargo test"</span></span>,</span>
</span></span></span></span>      </span>"</span>group"</span></span></span>:</span> </span>{</span>
</span></span></span></span></span>        </span>"</span>kind"</span></span></span>:</span> </span>"</span>test"</span></span>,</span>
</span></span></span></span></span>        </span>"</span>isDefault"</span></span></span>:</span> </span>true</span>
</span></span></span></span></span>      }</span></span>
</span></span></span></span>    }</span></span>
</span></span></span>  ]</span></span>
</span></span>}</span></span>
</span></code></pre>
You can use cmd-shift-b</code> to run the build now.</p>
Debugging</h2>
For the native debugger to work we need to install another extension to VSCode called 'LLDB Debugger'</a>. That would be cmd-p</code> and ext install vadimcn.vscode-lldb</code>.</p>
After reloading VSCode you should be able to set breakpoints on the side gutter and run the program using debugger by pressing F5</code>. First time doing this will result in the debugger choice window. Choose LLDB Debugger</code> as your debugger and you will be greeted with a JSON configuration file in which you need to tell the debugger a few details on your project. It may look like this:</p>
{</span>
</span></span>  </span>"</span>version"</span></span></span>:</span> </span>"</span>0.2.0"</span></span>,</span>
</span></span>  </span>"</span>configurations"</span></span></span>:</span> </span>[</span>
</span></span></span>    {</span>
</span></span></span></span>      </span>"</span>type"</span></span></span>:</span> </span>"</span>lldb"</span></span>,</span>
</span></span></span></span>      </span>"</span>request"</span></span></span>:</span> </span>"</span>launch"</span></span>,</span>
</span></span></span></span>      </span>"</span>name"</span></span></span>:</span> </span>"</span>Debug"</span></span>,</span>
</span></span></span></span>      </span>"</span>program"</span></span></span>:</span> </span>"</span>${workspaceRoot}/target/debug/name_of_your_executable"</span></span>,</span>
</span></span></span></span>      </span>"</span>args"</span></span></span>:</span> </span>[</span>]</span></span>,</span>
</span></span></span></span>      </span>"</span>cwd"</span></span></span>:</span> </span>"</span>${workspaceRoot}"</span></span>,</span>
</span></span></span></span>      </span>"</span>preLaunchTask"</span></span></span>:</span> </span>"</span>build"</span></span>
</span></span></span></span>    }</span></span>
</span></span></span>  ]</span></span>
</span></span>}</span></span>
</span></code></pre>
And that should be it !</p>
Now you should be able to set breakpoints and debug through the code.</p>
Start the debugging session by pressing F5</code> again - this should result in the build proceeding and then the debugger launching.</p>
Questions ?</h2>
Any questions ? Ask on https://users.rust-lang.org/</a> and ping me the link via email to me at blog@cyplo.net</a>. This way the answer will be visible to everyone in the community.</p>
Keep on Rusting !</p>


Adding graphs to posts in Nikola
2017-07-15T20:13:50+00:00
Update as of April 2019</strong></p>
This blog no longer runs on Nikola so the embedded examples do not work anymore.
The method described in the post is still valid though.</p>

I really like to teach, try to explain things in a simple manner. There is often no better way of making an explanation than visualizing it.
The problem is that I really can't draw, especially on a computer.
Wouldn't it be awesome if I could make the computer draw for me ?
I found out that, unsurprisingly, there is a software for that already. The one I like is called mermaid - it renders a simple text description of a graph or diagram into an html representation. Can look something like this.</p>
<</span>div</span> class</span>=</span>"</span></span>mermaid</span>"</span></span></span>></span></span>
</span>  graph TB subgraph one a1-->a2 end subgraph two b1-->b2 end subgraph three
</span>  c1-->c2 end c1-->a2
</span></</span>div</span>></span></span>
</span></code></pre>
This blog is rendered by Nikola hence I would like to show you how I've added mermaid support to my Nikola installation.
I use USE_BUNDLES = False</code> in conf.py</code> as for it gives me more control and is more HTTP/2 friendly.
With that disabled I can include mermaid's style and js files like so (also in conf.py</code>):</p>
EXTRA_HEAD_DATA</span></span> =</span> """</span></span></span>
</span></span></span></span></span></span><link rel="stylesheet" type="text/css" href="/assets/css/fontawesome.css">
</span></span></span><link rel="stylesheet" type="text/css" href="/assets/css/titillium.css">
</span></span></span><link rel="stylesheet" type="text/css" href="/assets/css/mermaid.forest.css">
</span></span></span>"""</span></span></span>
</span>
</span>BODY_END</span></span> =</span> """</span></span></span>
</span></span></span></span></span></span><script src="/assets/js/mermaid.js"></script>
</span></span></span><script>mermaid.initialize({startOnLoad:true, cloneCssStyles: false});</script>
</span></span></span>"""</span></span></span>
</span></code></pre>
Where do all these files come from though ? In my case, I have a custom theme, based on zen</code> called zen-cyplo</code>. The assets in the sources are located under themes/zen-cyplo/assets/</code>. Oh, and cloneCssStyles: false</code> is there as the default of true</code> made the different css styles on my blog clash.
Finally, to use mermaid in the post do (for reStructured Text):</p>
<</span>div</span> class</span>=</span>"</span></span>mermaid</span>"</span></span></span>></span></span>
</span>  graph TB subgraph one a1-->a2 end subgraph two b1-->b2 end subgraph three
</span>  c1-->c2 end c1-->a2
</span></</span>div</span>></span></span>
</span></code></pre>
You can click on source</code> button located below the title of this post to see it in action. If you are interested in the build process and how all these come together - the complete sources for this blog are hosted under https://github.com/cyplo/blog</del></p>


Upload your site to Netlify using their incremental deployment API
2017-03-20T19:25:00+00:00
I've recently switched to a setup where I do all my builds for this blog on Travis</a>. While doing so I needed to migrate away from using Netlify's internal build infrastructure. This resulted in a quick python script</a> that allows you to upload arbitrary directory tree to Netlify and does so using their incremental deployment API</a>. All that means that while this site is quite big in size the deployments go rather quickly ! There are some known issues</a> but apart from them the script should just work for any custom Netlify deployment you would like to have. I use it on this very site, to have a preview of any PR before merging it as well as for deploying the main site after the PR is merged. I hope you will find it useful and please do not hesitate if you want to post an issue or a PR !</p>


Running host programs in isolation inside one-off Docker containers
2017-02-17T22:08:46+00:00
I am quite bad at remembering how to launch docker to have everything set up correctly. Hence the following - a script that launches any commandline specified in its arguments inside a new docker container. Current directory is mounted inside the container automatically, so the thing you are executing can have its local dependencies satisfied.</p>
#</span></span>!/bin/bash</span>
</span></span>USERNAME</span>=</span></span>`</span>whoami</span></span>`</span></span></span>
</span>MOUNT_PATH</span>=</span>"</span>/mnt"</span></span></span>
</span>CURRENT_DIRECTORY</span>=</span></span>`</span>pwd</span></span> -</span>P</span></span>`</span></span> #</span></span> untangle symbolic links if needed - SELinux needs the real path</span>
</span></span></span>IMAGE</span>=</span>"</span>debian:jessie"</span></span></span>
</span>
</span>if</span> [[</span> -</span>z</span> $</span>1</span></span> ]]</span></span>;</span> then</span>
</span>    echo</span></span> "</span>usage: `</span>basename</span></span> $</span>0</span></span></span>`</span></span> command_to_run_inside_a_container"</span></span></span>
</span>    exit</span></span> 1</span>
</span>fi</span>
</span>
</span>RESOLVED_ARGUMENTS</span>=</span>"</span>$</span>@</span></span>"</span></span></span>
</span>docker</span></span> run -</span>i</span> -</span>t</span> -</span>v</span> "</span>$</span>CURRENT_DIRECTORY</span></span>"</span></span>:"</span>$</span>MOUNT_PATH</span></span>"</span></span>:Z $</span>IMAGE</span></span> bash -</span>c</span> "</span>useradd -M -d '$</span>MOUNT_PATH</span></span>' $</span>USERNAME</span></span> && cd '$</span>MOUNT_PATH</span></span>' && bash -c '$</span>RESOLVED_ARGUMENTS</span></span>'"</span></span></span>
</span>
</span>#</span></span> restore SELinux context for the current directory</span>
</span></span>restorecon_path</span>=</span></span>`</span>which</span></span> restorecon</span>`</span></span></span>
</span>if</span> [[</span> -</span>x</span> "</span>$</span>restorecon_path</span></span>"</span></span> ]]</span></span>;</span> then</span>
</span>    restorecon</span></span> -</span>R</span> "</span>$</span>CURRENT_DIRECTORY</span></span>"</span></span></span>
</span>fi</span>
</span></code></pre>
I use vanilla Debian Jessie as a run platform there, mostly because this is what most of my servers run.
The script covers setting up SELinux and mounting the directory from which it is run as /mnt</code> inside the container while also having the default non-root user added.</p>
Run Jessie, run !</p>


Using ad hoc Docker volumes on SELinux systems
2016-12-30T18:28:50+00:00
I've recently tried running some quick Docker commands using host's directory as a volume:</p>
docker run -i -t -v `pwd`:/mnt debian:jessie bash -c "ls -hal /mnt"
</span>ls: cannot open directory /mnt: Permission denied
</span></code></pre>
I use Fedora as my main OS, which, it turns out, has some pretty nice SELinux settings. These deny access from inside the container to the outside.
Said Fedora consists mostly of almost-newest-but-stable everything though, which makes Docker to be in a fairly recent version.
A version that understands how to change a SELinux context for the directory we're mounting, by itself !
You need at least Docker v1.7 for this.</p>
docker run -i -t -v `pwd`:/mnt:Z debian:jessie bash -c "ls -hal /mnt"
</span>total 8.0K
</span>drwxrwxr-x.  2 1000 1000 4.0K Dec 30 18:34 .
</span>drwxr-xr-x. 21 root root  242 Dec 30 19:07 ..
</span></code></pre>
Please note the capital Z</code> as a mount parameter.
And that is it. Mostly. Some cleanup remains, as docker leaves the host's directory with a changed SELinux context.
To restore it you need to</p>
restorecon -R `pwd`
</span></code></pre>
Or use any other path you'd like instead of pwd</code> in the examples above.</p>
Happy dockering !</p>


RustFest - organization was the best. Also rhymes.
2016-09-18T18:43:47+00:00
I went to RustFest and it was amazing !</h1>
It was clearly the best conference organizational-wise I've been to so far.
It made me think of what I really liked there. What made it so awesome and welcomy ?
To me this was a large number of small things just coming together. I listing them here for me to remember and for others to use.
Let's make conferences at least this friendly, together !</p>
Before the event:</h2>

very clear emails, repeating messages couple of times in different emails for the important stuff</li>
maps of the popular routes provided, e.g. from the airport and bus station</li>
supporter ticket sale - two times the price - you pay for another person that wouldn't be able to get a ticket otherwise</li>
survey on dietary requirements/allergies</li>
survey on childcare needs</li>
clear statement of the code of conduct for the conference</li>
</ul>
During the event:</h2>


very good MC person</p>

keeping people entertained through the tech breaks</li>
keeping tabs on the talk length, allowing appropriate amount of questions if the time allowed</li>
</ul>
</li>

live captioning of all talks - small screen outside of the view of the main screen with the text live</p>
</li>

getting the next speaker prepared and mic-tested before the end of the current talk</p>
</li>

quiet room to rest with clear rules on no talking and interrupting there</p>
</li>

clear signage for the different parts of the venue (washrooms, quiet room, party space etc)</p>
</li>

washrooms</p>


all same, ungendered</p>
</li>

basic items for free in said washrooms</p>

chewing gum</li>
tooth paste</li>
tampons, pads</li>
baby wipes</li>
</ul>
</li>
</ul>
</li>

info desk/registration</p>

clear info during registration, handing out programme</li>
asking people if they want to be on the photos - giving out lanyards accordingly - flashy red lanyards for people who do not want photos of them taken</li>
</ul>
</li>

emergency number to call with stated purpose</p>

code of conduct violation</li>
if you are lost in the city</li>
</ul>
</li>
</ul>


The best terminal emulator for Windows 10's Bash or how to run X applications
2016-07-06T19:21:13+00:00
update as of 03/2017</strong></p>
As this article is by far the most popular on my site right now I feel that an update is necessary, as the landscape evolved a bit since the original publish date.
Conemu</a> started to support Bash on Windows properly now and this is what I settled on in the end.
The article below is still relevant for the running X applications on Windows</em> part though.</p>

I've been playing a bit with
GNU/kWindows</a> a.k.a.
Bash on Windows a.k.a Windows Subsystem for Linux (Beta). I was
especially interested whether I can use my regular Linux
dotfiles</a> to recreate my working
environment of zsh + tmux + vim. The biggest troubles I had were with
the terminal emulator. While the default one, invoked by saying bash</strong>
is much better than powershell or cmd.exe already - it lacks some things
I've learned to rely on, like 256-colour palette support. Thus the
search for the ultimate terminal emulator begun. I tried ConEmu</strong>,
cmder</strong> and their spinoffs to no avail. Either the colours were
lacking, or the emulator would swallow up certain strokes like the arrow
keys. Then I thought - hey, I use terminator</strong> on Linux, maybe it
would be possible to use it here as well. To my surprise the answer was - yes !</p>

install Windows Subsystem for Linux</li>
restart Windows</li>
install VcXsrv</a></li>
run VcXsrv</li>
invoke the following from the bash console:</li>
</ul>
sudo apt-get install terminator
</span>DISPLAY=:0 terminator -u
</span></code></pre>

profit !</li>
</ul>
{{< figure src="/images/terminal-emulator-windows-10/capture.png" postition="center" >}}</p>
The font rendering is not ideal and the borderless mode does
not work, but hey, it is quite usable nonetheless ! It even has
bidirectional clipboard sharing with Windows' windows, which is good.
P.S. konsole and gnome-terminal do seem to have troubles launching
(crashy-crashes there)</p>


Don't throw away. Fix.
2016-06-12T12:25:34+00:00
Try fixing your things. It's already broken - no need to sweat.
You might learn something.
Level 2 - push for things to be easily fixable.
It's good for you. It's good for the planet.</p>









  
  
    
  </a>
  
  
    
  </a>
  
  
    
  </a>
  
  
    
  </a>
  
  
    
  </a>
  
  
    
  </a>
  

</div>

Contact & contributions</h2>

Let me know if you'd like to hack on this by contacting me on `legdur@cyplo.dev</code> or via cyplo@peninsula.industries</code> on the Fediverse.</li> </ul> Sources</h2> git clone https://git.cyplo.dev/cyplo/legdur.git</code></p>`

Sources</h2>
`git clone https://git.cyplo.dev/cyplo/legdur.git</code></p>`

Fonts</h2>
I settled on Fira Code Retina</a> for most of my programming and terminal needs.</p>

Code hosting</h2>
I'm trying to migrate off of GitHub, I'm trying out sr.ht</a> now. I like its simplicity and ease of setup, especially when it comes to the built-in CI.</p>

The end</h2>
And that's it !</p>
I hope you'll find this list useful and do not hesitate to contact me</a> on `blog@cyplo.net</code> if you would have any questions or comments. Happy hacking !</p>`

Fonts</h2>
I settled on Fira Code Retina</a> for most of my programming and terminal needs.</p>

Let's talk !</h1>
Have any questions ? Would like to pair on Rust ? Curious about TDD ? Ping me ! Email is good - hello@cyplo.net</a>.</p>
thanks !</p>

`Questions ?</h2>Any questions ? Ask on https://users.rust-lang.org/</a> and ping me the link via email to me at blog@cyplo.net</a>. This way the answer will be visible to everyone in the community.</p> Keep on Rusting !</p>`

Adventurous Computing

Stopping systemd services under memory pressure

Long posts in Mastodon, but now declarative !

Creating a base OCI image for Nix flake builds within Gitea/Forgejo

This blog is now self-hosted !

Containerised Mastodon server on NixOS

How to use a non-default ssh port for a Nix distributed build host

legdur - keep your legacy durable

Quick reproducible zig dev env using nix flakes

Using own OpenWRT router with Vodafone/Cityfibre

My 2020 setup

Don't let failures spread over your suite with process-based tests isolation

Buttermilk cornbread

My 2019 setup

Fuzzers and how to run them.

Website refresh

Generate Rust tests from data files

Testing tricks in Rust

Rust - controlling side effects from the test.

Resources for starting your adventure with Rust

I’m running Rust pair programming sessions !

Configure AWS Elastic Beanstalk Docker environment variables

Waiting for AWS Elastic Beanstalk environment to become ready

Setting up Rust development environment using VSCode on a Mac

Adding graphs to posts in Nikola

Upload your site to Netlify using their incremental deployment API

Running host programs in isolation inside one-off Docker containers

Using ad hoc Docker volumes on SELinux systems

RustFest - organization was the best. Also rhymes.

The best terminal emulator for Windows 10's Bash or how to run X applications

Don't throw away. Fix.

Adventurous Computing

Stopping systemd services under memory pressure

Long posts in Mastodon, but now declarative !

Creating a base OCI image for Nix flake builds within Gitea/Forgejo

This blog is now self-hosted !

Containerised Mastodon server on NixOS

How to use a non-default ssh port for a Nix distributed build host

legdur - keep your legacy durable

Contact & contributions</h2> Let me know if you'd like to hack on this by contacting me on legdur@cyplo.dev</code> or via cyplo@peninsula.industries</code> on the Fediverse.</li> </ul> Sources</h2> git clone https://git.cyplo.dev/cyplo/legdur.git</code></p>

Sources</h2> git clone https://git.cyplo.dev/cyplo/legdur.git</code></p>

Quick reproducible zig dev env using nix flakes

Using own OpenWRT router with Vodafone/Cityfibre

My 2020 setup

TODOs for 2021:</h3> debug the overheating problem</li>

Don't let failures spread over your suite with process-based tests isolation

Buttermilk cornbread

My 2019 setup

Fuzzers and how to run them.

How does it compare to property-based tests ?</h2> When thinking about other types of tests that are driven by randomness and are generally black-box-ish - property-based tests come to mind. How does fuzzing compare ?</p>

Website refresh

Generate Rust tests from data files

Testing tricks in Rust

Use verbs as test module names</h1> Who said that the test module needs to be named test</code> ? Experiment with different module names, pay attention to how the test runner displays the results.</p> A structure that I like, an example:</p> worker.rs:</code></p>

Rust - controlling side effects from the test.

Resources for starting your adventure with Rust

I’m running Rust pair programming sessions !

Configure AWS Elastic Beanstalk Docker environment variables

Waiting for AWS Elastic Beanstalk environment to become ready

Setting up Rust development environment using VSCode on a Mac

Completion and highlighting</h2> While on Linux VSCode with the Rust plugin seems to work more or less out of the box, on a Mac I needed to spend some time configuring it.</p> First things first though, let's start by installing Rust version manager, rustup.</p>

Adding graphs to posts in Nikola

Upload your site to Netlify using their incremental deployment API

Running host programs in isolation inside one-off Docker containers

Using ad hoc Docker volumes on SELinux systems

RustFest - organization was the best. Also rhymes.

The best terminal emulator for Windows 10's Bash or how to run X applications

Don't throw away. Fix.

Contact & contributions</h2>

Let me know if you'd like to hack on this by contacting me on `legdur@cyplo.dev</code> or via cyplo@peninsula.industries</code> on the Fediverse.</li> </ul> Sources</h2> git clone https://git.cyplo.dev/cyplo/legdur.git</code></p>`

Sources</h2>
`git clone https://git.cyplo.dev/cyplo/legdur.git</code></p>`