chromic

chromic https://chromic.org/ Recent content on chromic en-ca chimo@chromic.org (chimo) chimo@chromic.org (chimo) gdk-pixbuf symbol not found https://chromic.org/blog/gdk-pixbuf-sumbol-not-found/ Mon, 28 Sep 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/gdk-pixbuf-sumbol-not-found/ <p>If you’re running Alpine Linux edge right now and some of your applications (for me, <a href="https://pkgs.alpinelinux.org/package/edge/testing/x86_64/signal-desktop">signal-desktop</a>. <a href="https://gitlab.alpinelinux.org/alpine/aports/-/work_items/18450">For others</a>, Firefox) throw an error similar to the following:</p> <blockquote> <p>g_module_open() failed for /usr/lib/gdk-pixbuf-2.0/2.10.0/loaders/libpixbufloader-xpm.so: Error relocating /usr/lib/libgdk_pixbuf-2.0.so.0: gly_pixel_density_new: symbol not found Executing gtk+3.0-3.24.52-r0.trigger</p> </blockquote> <p>A workaround is to <a href="https://gitlab.alpinelinux.org/alpine/aports/-/work_items/18450#note_645726">downgrade gdk-pixbuf</a>.</p> <p>I don’t downgrade/pin packages very often. As a result, I never remember how to do it. In this case, I followed the process described by the commenter:</p> <ol> <li>Add the 3.24 repositories as “<a href="https://wiki.alpinelinux.org/wiki/Repositories#Tagged_repository">tagged repositories</a>” in /etc/apk/repositories:</li> </ol> <blockquote> <p>@v3.24 <a href="http://dl-cdn.alpinelinux.org/alpine/v3.24/main">http://dl-cdn.alpinelinux.org/alpine/v3.24/main</a><br> @v3.24 <a href="http://dl-cdn.alpinelinux.org/alpine/v3.24/community">http://dl-cdn.alpinelinux.org/alpine/v3.24/community</a></p> </blockquote> <ol start="2"> <li>Install that specific version:</li> </ol> <blockquote> <p>apk add <a href="mailto:gdk-pixbuf@v3.24">gdk-pixbuf@v3.24</a></p> </blockquote> <p>You can also provide a specific version number to `apk add` as described in the “<a href="https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper#Package_pinning">Package pinning</a>” section of the Alpine wiki.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: gdk-pixbuf%20symbol%20not%20found">Reply by email</a> </p> Modern Motherfucking Website https://chromic.org/likes/modern-motherfucking-website/ Sun, 27 Sep 2026 12:10:21 +0000 chimo@chromic.org (chimo) https://chromic.org/likes/modern-motherfucking-website/ <div class="h-entry"> <div class="h-cite u-like-of"> Liked "<a class="u-url" href="https://modernmotherfuckingwebsite.dreamstation.systems/">Modern Motherfucking Website</a>" by <span class="p-author h-card"> <a class="u-url p-name" href="https://dreamstation.systems/personal/">robin</a>. </span> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Modern%20Motherfucking%20Website">Reply by email</a> </p> rbw missing field access_token https://chromic.org/blog/rbw-missing-field-access-token/ Sun, 27 Sep 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/rbw-missing-field-access-token/ <p>If `<a href="https://github.com/doy/rbw">rbw</a>` all of the sudden returns:</p> <blockquote> <p>failed to sync database from server: failed to parse JSON: missing field <code>access_token</code> at line 1 column 25</p> </blockquote> <p>A <a href="https://github.com/doy/rbw/issues/32#issuecomment-761613019">workaround</a> is to run `rbw purge`. Note that this deletes the local copy of the password database.</p> <p>It looks like there might be <a href="https://github.com/doy/rbw/pull/362">a fix in a pending PR</a> from August, but it hasn’t been merged.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: rbw%20missing%20field%20access_token">Reply by email</a> </p> Beaf Stu https://chromic.org/blog/beaf-stu/ Sat, 26 Sep 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/beaf-stu/ <div class="notice"> This is a follow-up to <a href="https://chromic.org/blog/advertising-causes/">Chic 'N' Stu</a>. </div> <p>My fallback “go-to” gas station (i.e.: the one that’s second-most conveniently located for me) started harassing<sup><a href="#f1">1</a></sup> <a name="r1" id="r1">people</a> to open a credit card while you’re trapped filling-up.</p> <p>It is also no longer my go-to gas station.</p> <p><sup><a id="f1" name="f1">1</a></sup> To be fair, the people have always been pleasant and not pushy while just doing their job. But I still don’t want to deal with this.<a href="#r1"><sup>^</sup></a></p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Beaf%20Stu">Reply by email</a> </p> Gatus https://chromic.org/blog/gatus/ Mon, 07 Sep 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/gatus/ <p>I’m testing out <a href="https://github.com/TwiN/gatus">Gatus</a> as a replacement to <a href="https://github.com/louislam/uptime-kuma">Uptime Kuma</a>.</p> <p>Since <a href="https://github.com/louislam/uptime-kuma/releases/tag/2.3.0">v2.3.0</a>, Uptime Kuma supports monitoring Oracle databases. That’s all well and good, but I’ve had trouble <a href="https://git.srht.chromic.org/~chimo/apkbuilds/tree/master/item/uptime-kuma">packaging it for Alpine Linux</a> due to <a href="https://www.npmjs.com/package/oracledb">NodeJS oracledb</a> including binaries that rely on libraries that aren’t available on Alpine Linux.</p> <p>Uptime Kuma uses “<a href="https://node-oracledb.readthedocs.io/en/latest/user_guide/introduction.html#architecture">thin mode</a>”, which doesn’t use those binaries, but the fact that they’re part of the package caused abuild to error-out at the “Tracing dependencies” step.</p> <p>I managed to get it packaged, and learned about APKBUILD “<a href="https://github.com/alpinelinux/abuild/blob/master/APKBUILD.5.scd?plain=1#L396">somask</a>” and “<a href="https://wiki.gentoo.org/wiki/Hardened/PaX_Utilities#The_scanelf_application">scanelf</a>” in the process, but I had been thinking about finding an alternative that provides an API for a while now.</p> <p>The packaging issues gave me the kick-in-the-pants I needed to actually look into it.</p> <p>Some of the things I like about Gatus so far:</p> <ul> <li>Has an API.</li> <li>Supports PostgreSQL.</li> <li>Configuration via text file.</li> <li>Clean web interface.</li> <li>Seems lighter on resources (CPU/RAM).</li> </ul> <p>I also had a look at <a href="https://github.com/statping-ng/statping-ng">statping-ng</a>, but it doesn’t have maintenance windows (and the project doesn’t look very active).</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Gatus">Reply by email</a> </p> Incus VM won't start https://chromic.org/blog/incus-vm-wont-start/ Sat, 05 Sep 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/incus-vm-wont-start/ <div class="notice"> Update: This <a href="https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/108599">has been fixed</a>. </div> <p>If your Incus VMs fail to start with a generic “Failed to run: forkqemu…” error and absolutely nothing in the logs anywhere (qemu.log, or otherwise), it might be because you need to install the “<a href="https://pkgs.alpinelinux.org/package/v3.24/community/x86_64/qemu-audio-spice">qemu-audio-spice</a>” package.</p> <p>More info in the <a href="https://gitlab.alpinelinux.org/alpine/aports/-/work_items/18278">Alpine Linux Issue Tracker</a>.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Incus%20VM%20won%27t%20start">Reply by email</a> </p> Home Assistant https://chromic.org/blog/home-assistant/ Mon, 31 Aug 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/home-assistant/ <p><a href="https://chromic.org/blog/jellyfin-for-tizen/">Earlier this year</a>, I mentioned something along the lines of:</p> <blockquote> <p>If the newly released <a href="https://github.com/jellyfin/jellyfin-tizen">Jellyfin app</a> for my television works well, I’ll likely retire my raspberrypi as a media player and re-use it for a different purpose.</p> </blockquote> <p>About a month ago, I decided to finally give <a href="https://www.home-assistant.io/">Home Assistant</a> a try, and figured it would be a good use-case for the raspberrypi.</p> <p>I don’t really have any notes, except that I missed the mention of the “shell” subcommand in the console, because it’s not listed with the other commands in `ha help`. I just need to learn to read, I suppose.</p> <p>I needed the “shell” function because the default “.local” URL wasn’t magically accessible from other machines (which, given my network configurations, I expected and prefer).</p> <p>Other than that, it seems to be doing what it’s supposed to, and seems to be doing it well.</p> <p>Going forward, I want to be more conscious about which IoT devices enter my home to ensure they don’t <em>require</em> talking to third-party services and can work in an air-gapped environment.</p> <p>I tend to use my devices until they either die or stop getting security updates so it might take a little while before my current devices get replaced.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Home%20Assistant">Reply by email</a> </p> RecipeJSON https://chromic.org/bookmarks/recipe-json/ Sun, 30 Aug 2026 14:44:16 +0000 chimo@chromic.org (chimo) https://chromic.org/bookmarks/recipe-json/ <div class="h-entry"> <div class="h-cite u-bookmark-of"> Bookmarked "<a class="u-url" href="https://recipejson.github.io/format/">RecipeJSON</a>" by <span class="p-author h-card"> <a class="u-url p-name" href="https://github.com/leonkunert">Leon Kunert</a>. </span> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: RecipeJSON">Reply by email</a> </p> Unifi Network Application -> Unifi OS Server https://chromic.org/blog/unifi-network-application-unifi-os-server/ Sun, 30 Aug 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/unifi-network-application-unifi-os-server/ <p>Thanks to <a href="https://neilzone.co.uk/">Neil</a>’s excellent <a href="https://neilzone.co.uk/2026/08/migrating-my-self-hosted-unifi-controller-from-network-manager-to-unifi-os-server/">Migrating my self-hosted UniFi controller from Network Manager to UniFi OS Server</a> blog post showing up in my RSS Reader, it reminded me that I had tried installing Unifi OS Server on Incus a little while back but ran into issues.</p> <p>A web search later, I found <a href="https://discussion.scottibyte.com/u/scott/summary">Scott</a>’s <a href="https://discussion.scottibyte.com/t/self-hosted-unifi-os-in-an-incus-container/673">Self Hosted Unifi OS in an Incus Container</a> post.</p> <p>Armed with those two articles, the migration went fairly smoothly. The one thing I had to change from Scott’s post is:</p> <blockquote> <p>lxc.cgroup.devices.allow = a * rwm</p> </blockquote> <p>to:</p> <blockquote> <p>lxc.cgroup2.devices.allow = a * rwm</p> </blockquote> <p>Another note: when exporting the Network Manager settings to a backup file, the file had no extension. Unifi OS Server didn’t want to import it until I added the “.unf” extention to it.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Unifi%20Network%20Application%20-%3e%20Unifi%20OS%20Server">Reply by email</a> </p> They broke old.reddit https://chromic.org/blog/they-broke-old-reddit/ Sat, 29 Aug 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/they-broke-old-reddit/ <p>I don’t use reddit often, but when a search result would point me there, I would use the old.reddit.com interface. It seemed cleaner to me, and didn’t require JavaScript. To that effect, I had <a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/86a20b6750aa70702a99b6eb1247e5f612058a8d/item/.config/qutebrowser/pyconfig/redirectors.py#L9-10">qutebrowser interceptors</a> that would automatically redirect me from reddit.com links to old.reddit.com.</p> <p>At some point (in July, I think) Reddit <a href="https://redditinc.com/news/modernizing-reddits-infrastructure-and-moderation-tools">decided to require login</a>:</p> <blockquote> <p>In order to protect our platform, business, and users from abusive scraping and spamming, over time we plan to make changes involving Old Reddit, the legacy desktop web experience for Reddit that was built 21 years ago. In the short term, we will continue to require users to be logged in to access it and will work on migrating critical bots and moderator workflows to our modern stack. These efforts will help us mitigate the bad behavior and give us time to plan the complex follow-up steps.</p> </blockquote> <p>I <a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/120b4431c77f7c06ed411c997ccf0a0e4da78dc5/item/.config/qutebrowser/pyconfig/redirectors.py#L9-10">updated my interceptors</a> so that I’m redirected to a <a href="https://github.com/redlib-org/redlib">redlib</a> <a href="https://github.com/redlib-org/redlib-instances/blob/main/instances.md">instance</a> instead. /shrug</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: They%20broke%20old.reddit">Reply by email</a> </p> Show search count in vim https://chromic.org/blog/show-search-count-in-vim/ Sat, 15 Aug 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/show-search-count-in-vim/ <p>To show “search count” message in the status bar when searching (ex: “[13/37]”):</p> <blockquote> <p>set shortmess-=S</p> </blockquote> <p>Or to perform a count on-demand:</p> <blockquote> <p>:%s/some-word//gn</p> </blockquote> <p>Credits to <a href="https://stackoverflow.com/questions/49297579/how-to-count-search-results-in-vim">stackoverflow</a>.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Show%20search%20count%20in%20vim">Reply by email</a> </p> Create git branch from Jira issue summary https://chromic.org/blog/create-git-branch-from-jira-issue-summary/ Wed, 29 Jul 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/create-git-branch-from-jira-issue-summary/ <p>We use Jira to track issues/tasks at work. I got tired of switching to the browser to refer to (or copy, paste and edit) the Jira issue summary to give my git branches meaningful names.</p> <p>I wrote a quick script that:</p> <ol> <li>Takes a Jira issue key as an input<br> (ex: TARDIS-1234).</li> <li>Gets my Jira credentials from my password manager<br> (<a href="https://github.com/doy/rbw">rbw</a>).</li> <li>Calls the Jira API to retrieve the Jira issue “Summary”<br> (ex: “Implement Time-Traveling”).</li> <li>Turns the summary into a slug<br> (thanks, <a href="https://github.com/Mayeu/slugify/blob/master/slugify#L24">github.com/Mayeu/slugify</a>)</li> <li>Creates and switches to a new branch<br> (ex: `git checkout -b tardis-1234-implement-time-traveling`).</li> </ol> <figure> <figcaption>~/.local/bin/jira-branch.sh</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl"><span class="cp">#!/bin/sh -eu </span></span></span><span class="line"><span class="cl"><span class="cp"></span> </span></span><span class="line"><span class="cl"><span class="c1"># https://github.com/Mayeu/slugify/blob/master/slugify#L24</span> </span></span><span class="line"><span class="cl">to_slug<span class="o">()</span> <span class="o">{</span> </span></span><span class="line"><span class="cl"> <span class="c1"># Forcing the POSIX local so alnum is only 0-9A-Za-z</span> </span></span><span class="line"><span class="cl"> <span class="nb">export</span> <span class="nv">LANG</span><span class="o">=</span>POSIX </span></span><span class="line"><span class="cl"> <span class="nb">export</span> <span class="nv">LC_ALL</span><span class="o">=</span>POSIX </span></span><span class="line"><span class="cl"> <span class="c1"># Keep only alphanumeric value</span> </span></span><span class="line"><span class="cl"> sed -e <span class="s1">'s/[^[:alnum:]]/-/g'</span> <span class="p">|</span> </span></span><span class="line"><span class="cl"> <span class="c1"># Keep only one dash if there is multiple one consecutively</span> </span></span><span class="line"><span class="cl"> tr -s <span class="s1">'-'</span> <span class="p">|</span> </span></span><span class="line"><span class="cl"> <span class="c1"># Lowercase everything</span> </span></span><span class="line"><span class="cl"> tr A-Z a-z <span class="p">|</span> </span></span><span class="line"><span class="cl"> <span class="c1"># Remove last dash if there is nothing after</span> </span></span><span class="line"><span class="cl"> sed -e <span class="s1">'s/-$//'</span> </span></span><span class="line"><span class="cl"><span class="o">}</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="nv">issue_key</span><span class="o">=</span><span class="s2">"</span><span class="si">${</span><span class="nv">1</span><span class="si">}</span><span class="s2">"</span> </span></span><span class="line"><span class="cl"><span class="nv">issue_key</span><span class="o">=</span><span class="k">$(</span><span class="nb">echo</span> <span class="s2">"</span><span class="si">${</span><span class="nv">issue_key</span><span class="si">}</span><span class="s2">"</span> <span class="p">|</span> tr A-Z a-z<span class="k">)</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="nv">username</span><span class="o">=</span><span class="k">$(</span>rbw get -f username jira<span class="k">)</span> </span></span><span class="line"><span class="cl"><span class="nv">password</span><span class="o">=</span><span class="k">$(</span>rbw get -f password jira<span class="k">)</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="nv">jira_api_root</span><span class="o">=</span><span class="s2">"https://example.org/rest/api/2"</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="nv">issue_summary</span><span class="o">=</span><span class="k">$(</span> </span></span><span class="line"><span class="cl"> curl -s -u <span class="s2">"</span><span class="si">${</span><span class="nv">username</span><span class="si">}</span><span class="s2">:</span><span class="si">${</span><span class="nv">password</span><span class="si">}</span><span class="s2">"</span> --request GET <span class="se">\ </span></span></span><span class="line"><span class="cl"><span class="se"></span> --url <span class="s2">"</span><span class="si">${</span><span class="nv">jira_api_root</span><span class="si">}</span><span class="s2">/issue/</span><span class="si">${</span><span class="nv">issue_key</span><span class="si">}</span><span class="s2">?fields=summary"</span> <span class="se">\ </span></span></span><span class="line"><span class="cl"><span class="se"></span> --header <span class="s2">"Accept: application/json"</span> <span class="se">\ </span></span></span><span class="line"><span class="cl"><span class="se"></span> <span class="p">|</span> jq .fields.summary -r <span class="p">|</span> to_slug </span></span><span class="line"><span class="cl"><span class="k">)</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl">git checkout -b <span class="s2">"</span><span class="si">${</span><span class="nv">issue_key</span><span class="si">}</span><span class="s2">-</span><span class="si">${</span><span class="nv">issue_summary</span><span class="si">}</span><span class="s2">"</span></span></span></code></pre></div> </figure> <p>I might call this from a git-alias in the future. Right now it’s just a script in my $PATH.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Create%20git%20branch%20from%20Jira%20issue%20summary">Reply by email</a> </p> Logseq DB (v2) and git sync https://chromic.org/blog/logseq-db-git-sync/ Thu, 23 Jul 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/logseq-db-git-sync/ <p>Just jotting down notes on a naive way of syncing my <a href="https://logseq.com/">Logseq</a> (<a href="https://discuss.logseq.com/t/whats-new-with-logseq-db-may-16th-2026/35020">v2</a>) graph to a remote git repository using the <a href="https://discuss.logseq.com/t/whats-new-with-logseq-db-may-16th-2026/35020#p-70276-h-1-markdown-mirror-your-db-graph-also-on-disk-3">Markdown Mirror</a> Logseq feature.</p> <ol> <li>Initialize a git repository somewhere (ex: ~/logseq-git).</li> <li>Enable Markdown Mirror in your Logseq graph.</li> <li>Create a `logseq` alias/script which <ol> <li>Starts an auto-commit loop in the background.</li> <li>Launches Logseq.</li> <li>Auto-commits one last time once Logseq terminates (ex: when Quitting).</li> </ol> </li> </ol> <figure> <figcaption>~/.local/bin/logseq</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl"><span class="cp">#!/bin/sh -eu </span></span></span><span class="line"><span class="cl"><span class="cp"></span> </span></span><span class="line"><span class="cl"><span class="nv">git_dir</span><span class="o">=</span><span class="s2">"~/logseq-git/.git"</span> </span></span><span class="line"><span class="cl"><span class="nv">work_tree</span><span class="o">=</span><span class="s2">"~/logseq/graphs/Demo/mirror/markdown"</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl">auto_commit<span class="o">()</span> <span class="o">(</span> </span></span><span class="line"><span class="cl"> <span class="c1"># Add all files</span> </span></span><span class="line"><span class="cl"> git --git-dir<span class="o">=</span><span class="s2">"</span><span class="si">${</span><span class="nv">git_dir</span><span class="si">}</span><span class="s2">"</span> --work-tree<span class="o">=</span><span class="s2">"</span><span class="si">${</span><span class="nv">work_tree</span><span class="si">}</span><span class="s2">"</span> add -A </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"> <span class="c1"># Commit</span> </span></span><span class="line"><span class="cl"> git --git-dir<span class="o">=</span><span class="s2">"</span><span class="si">${</span><span class="nv">git_dir</span><span class="si">}</span><span class="s2">"</span> --work-tree<span class="o">=</span><span class="s2">"</span><span class="si">${</span><span class="nv">work_tree</span><span class="si">}</span><span class="s2">"</span> commit -m <span class="s1">'Auto-commit.'</span> <span class="o">||</span> <span class="nb">true</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"> <span class="c1"># Push</span> </span></span><span class="line"><span class="cl"> git --git-dir<span class="o">=</span><span class="s2">"</span><span class="si">${</span><span class="nv">git_dir</span><span class="si">}</span><span class="s2">"</span> --work-tree<span class="o">=</span><span class="s2">"</span><span class="si">${</span><span class="nv">work_tree</span><span class="si">}</span><span class="s2">"</span> push <span class="o">||</span> <span class="nb">true</span> </span></span><span class="line"><span class="cl"><span class="o">)</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl">main<span class="o">()</span> <span class="o">(</span> </span></span><span class="line"><span class="cl"> <span class="c1"># Auto-commit loop</span> </span></span><span class="line"><span class="cl"> <span class="o">(</span><span class="k">while</span> <span class="nb">true</span> </span></span><span class="line"><span class="cl"> <span class="k">do</span> </span></span><span class="line"><span class="cl"> auto_commit </span></span><span class="line"><span class="cl"> sleep <span class="m">60</span> </span></span><span class="line"><span class="cl"> <span class="k">done</span><span class="o">)</span><span class="p">&</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"> <span class="nv">loop_pid</span><span class="o">=</span><span class="si">${</span><span class="p">!</span><span class="si">}</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"> <span class="c1"># Launch logseq</span> </span></span><span class="line"><span class="cl"> logseq </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"> <span class="nb">kill</span> <span class="s2">"</span><span class="si">${</span><span class="nv">loop_pid</span><span class="si">}</span><span class="s2">"</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"> <span class="c1"># Commit on close</span> </span></span><span class="line"><span class="cl"> auto_commit </span></span><span class="line"><span class="cl"><span class="o">)</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl">main</span></span></code></pre></div> </figure> <p>Notes:</p> <ol> <li>This would probably be better as an actual Logseq plugin.</li> <li>The script above only auto-commits a single graph (I only have one).</li> <li>From limited testing, it seems to be working okay but time will tell.</li> <li>Some weird things may or may not happen if multiple Logseq instances are launched via this script (I only ever have one instance running).</li> </ol> <p> <a href="mailto:chimo@chromic.org?subject=Re: Logseq%20DB%20%28v2%29%20and%20git%20sync">Reply by email</a> </p> Keep Android Open https://chromic.org/blog/keep-android-open/ Fri, 10 Jul 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/keep-android-open/ <p>I use <a href="https://www.davx5.com/">DAVx5</a> (installed via <a href="https://f-droid.org">F-Droid</a>) to sync my contacts and calendars with my <a href="https://nextcloud.com/">Nextcloud</a> instance.</p> <p>I don’t open the app very often, and I mean that in the most positive way. I configured it a long time ago and it has been doing its thing in the background since then. For whatever reason, I opened the app recently and saw a small banner at the top saying:</p> <blockquote> <p>Android’s openness is under threat. Learn how you can help at <a href="https://keepandroidopen.org/">keepandroidopen.org</a>.</p> </blockquote> <p>Some random thoughts that crossed my mind while reading the page:</p> <ol> <li>The link above is the only thing I read about this topic (at least so far). As such, I don’t consider myself informed on the matter.</li> <li>My Pixel 6 Pro will stop receiving security updates on October, 1st 2026, and as such;</li> <li>I am already considering moving to an alternative operating system such as <a href="https://lineageos.org/">LineageOS</a> or <a href="https://grapheneos.org/">GrapheneOS</a> without <a href="https://en.wikipedia.org/wiki/Google_Mobile_Services">Google Mobile Services</a>.</li> <li>I do realize that switching to a different OS doesn’t really do anything to address the issue at hand.</li> <li>From a security/accountability perspective, I understand how a corporation would want more control over its ecosystem.</li> <li>I also understand that individual rights (e.g.: privacy) can be eroded in the name of “security” while being thinly veiled attempts at surveillance.</li> </ol> <p>Anyway, I don’t really have any insights on this it except a general “I don’t like this” feeling. I don’t publish anything on Google Play, and I’m planning on migrating away from the “Google Android” ecosystem, so it doesn’t directly affect me (a flawed point-of-view, I know). But I might migrate away from “Google Android” before the planned Oct. 1st deadline I had set for myself.</p> <p>I ran Cyanogenmod/LineageOS as my daily-driver for about three years (~2015-2018). I’m sure it’ll be nice to be back in a lot of ways.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Keep%20Android%20Open">Reply by email</a> </p> btdu https://chromic.org/blog/btdu/ Mon, 06 Jul 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/btdu/ <p>Today I learned about `<a href="https://github.com/CyberShadow/btdu">btdu</a>`. It’s like `<a href="https://dev.yorhel.nl/ncdu">ncdu</a>`, but for <a href="https://btrfs.readthedocs.io/en/latest/">btrfs</a>.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: btdu">Reply by email</a> </p> Jellyfin, YouTube and Twitch https://chromic.org/blog/jellyfin-youtube-twitch/ Sun, 05 Jul 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/jellyfin-youtube-twitch/ <p>I <a href="https://chromic.org/blog/jellyfin-for-tizen/">recently</a> started using <a href="https://jellyfin.org/">Jellyfin</a> a bit more and <a href="https://kodi.tv/">Kodi</a> a bit less. Not because Kodi isn’t great, but mainly because I like trying different things. The fact that my Tv has a native/official Jellyfin client is also a pretty good motivator.</p> <p>Coincidentally (or more likely due to Google snooping on my search history), I came across <a href="https://www.xda-developers.com/set-up-jellyfin-movies-youtube-videos-became-real-reason-use-it/">an article on XDA</a> talking about watching YouTube videos via Jellyfin (with the intermediary step of downloading them locally).</p> <p>The way I currently interface with YouTube (as little as possible) is as follows:</p> <ol> <li>Subscribe to the channel’s RSS feed to get notified of new videos.</li> <li>Once a new video shows up in my RSS reader, copy the video URL from the feed.</li> <li>Run `mpv [video-url]` to watch.</li> </ol> <p>It works well, I don’t need to be logged-in to third-party services, and I don’t have to visit the YouTube website (and therefore turn on JavaScript). Although whenever I need to browse YouTube I use <a href="https://docs.invidious.io/instances/">invidious</a>.</p> <p>I don’t think Twitch has RSS feeds however, so I don’t have a similar process for Twitch channels. I just use the website to go check once in a while whether new videos have been added and watch them via the same `mpv` method.</p> <p>The XDA article linked above mentioned a few different tools to download YouTube videos automatically. The one that grabbed my attention is “<a href="https://github.com/jmbannon/ytdl-sub">ytdl-sub</a>”. I don’t need any kind of UI if I’m going to end up watching using Jellyfin. Configuration is file-driven, no database needed, works with Twitch, Alpine has <a href="https://pkgs.alpinelinux.org/package/edge/community/x86_64/ytdl-sub">a package for it</a>.</p> <p>So this gives me a nicer process for Twitch, and a consistent process for both YouTube and Twitch:</p> <ol> <li>Videos from both sources end up in a Jellyfin library (courtesy of a daily cronjob running `ytdl-sub`).</li> <li>Watch video via Jellyfin.</li> </ol> <p>The “watching” step can be done with any <a href="https://jellyfin.org/downloads/clients/all/">Jellyfin client</a>. Since I like to stay in the terminal as much as possible, I opted for <a href="https://github.com/Aanok/jftui">jftui</a>.</p> <p>The last thing I want to do is build a custom notification system using the <a href="https://api.jellyfin.org/">Jellyfin API</a> to add a block in my <a href="https://git.srht.chromic.org/~chimo/shuibar">shuibar</a> whenever I have unwatched videos coming from YouTube/Twitch.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Jellyfin%2c%20YouTube%20and%20Twitch">Reply by email</a> </p> Jellyfin doesn't scan new folders https://chromic.org/blog/jellyfin-doesnt-scan-new-folders/ Thu, 02 Jul 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/jellyfin-doesnt-scan-new-folders/ <p>I use <a href="https://jellyfin.org/">Jellyfin</a> to manage my media. I have two libraries setup. I added a new folder to one of them, but Jellyfin didn’t show me the media items in that new folder. Running “Scan library” didn’t do anything either.</p> <p>Turns out I didn’t misconfigure anything (always my first assumption) but instead it’s <a href="https://github.com/jellyfin/jellyfin-web/issues/7801">a bug</a>.</p> <p>It has been fixed in the master branch and presumably will be included in the next release.</p> <p>In the meantime, one <a href="https://github.com/jellyfin/jellyfin-web/issues/7801#issuecomment-4190123026">work-around is to “Scan <strong>All</strong> Libraries</a>” after adding the new folder, apparently.</p> <p>If you have decent-size libraries and puny hardware (like I do), re-scanning all libraries might seem daunting. Instead, I opted to create a new library (I named it “tmp-bug-7801”). Jellyfin picked-up the media in the new library just fine.</p> <p>Once the bugfix has been shipped as part of a release, the plan is to delete this library and re-add the folder to the main library.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Jellyfin%20doesn%27t%20scan%20new%20folders">Reply by email</a> </p> stunnel 5.76 segfault https://chromic.org/blog/stunnel-576-segfault/ Wed, 01 Jul 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/stunnel-576-segfault/ <p>It turns out that <a href="https://stunnel.org">stunnel</a> 5.76 and 5.77, “<a href="https://www.stunnel.org/mailman3/hyperkitty/list/stunnel-users@lists.stunnel.org/thread/EVIIZ3GTPWFHINQKQOGKPJGP67ZRFFVL/?noscript">crash when both global (default) and service-level lists of values are configured for an option</a>”.</p> <p>For example, this crashes because “cert” and “key” are both defined in the global section, and the “example_net” section:</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl"><span class="nv">cert</span> <span class="o">=</span> /etc/ssl/private/example.org/fullchain.cer </span></span><span class="line"><span class="cl"><span class="nv">key</span> <span class="o">=</span> /etc/ssl/private/example.org/example.org.key </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="o">[</span>default<span class="o">]</span> </span></span><span class="line"><span class="cl"><span class="nv">accept</span> <span class="o">=</span> <span class="m">443</span> </span></span><span class="line"><span class="cl"><span class="nv">connect</span> <span class="o">=</span> 127.0.0.1:80 </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="o">[</span>example_net<span class="o">]</span> </span></span><span class="line"><span class="cl"><span class="nv">sni</span> <span class="o">=</span> default:example.net </span></span><span class="line"><span class="cl"><span class="nv">cert</span> <span class="o">=</span> /etc/ssl/private/example.net/fullchain.cer </span></span><span class="line"><span class="cl"><span class="nv">key</span> <span class="o">=</span> /etc/ssl/private/example.net/example.net.key </span></span><span class="line"><span class="cl"><span class="nv">connect</span> <span class="o">=</span> 127.0.0.1:80</span></span></code></pre></div> <p>This has been fixed in stunnel <a href="https://www.stunnel.org/NEWS.html">5.78</a> but since Alpine Linux still ships 5.76 in both <a href="https://pkgs.alpinelinux.org/packages?name=stunnel&branch=edge&repo=&arch=x86_64&origin=&flagged=&maintainer=">edge</a> and <a href="https://pkgs.alpinelinux.org/packages?name=stunnel&branch=v3.24&repo=&arch=x86_64&origin=&flagged=&maintainer=">3.24</a>, it is something I ran into.</p> <p>A work-around in my case is to move the “cert” and “key” properties in the global section to the “default” section.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: stunnel%205.76%20segfault">Reply by email</a> </p> haproxy 3.3 sets SNI by default https://chromic.org/blog/haproxy-33-sni/ Mon, 29 Jun 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/haproxy-33-sni/ <p>The last sentence of my <a href="https://chromic.org/blog/acme-alpinelinux-324/">previous post</a> was:</p> <blockquote> <p>Now I need to figure out why all my <a href="https://pkgs.alpinelinux.org/package/v3.24/community/x86_64/stunnel">stunnel</a>s are broken.</p> </blockquote> <p>Upgrading from Alpine 3.23 to 3.24 upgraded my haproxy from 3.2.x to 3.4.x.</p> <p>It turns out that starting with 3.3, haproxy <a href="https://www.haproxy.com/blog/announcing-haproxy-3-3#sni">automatically sets an SNI value from the “host” HTTP header and forward it to the backend server</a>.</p> <p>All my stunnel instances only serve a single application, and use a certificate for an internal domain. For example, this blog (chromic.org on haproxy) maps to “blog.lxc.chromic.org” internally. My stunnel configuration looks like this:</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl"><span class="c1"># ...</span> </span></span><span class="line"><span class="cl"><span class="nv">cert</span> <span class="o">=</span> /etc/ssl/private/lxc.chromic.org/fullchain.cer </span></span><span class="line"><span class="cl"><span class="nv">key</span> <span class="o">=</span> /etc/ssl/private/lxc.chromic.org/lxc.chromic.org.key </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="nv">client</span> <span class="o">=</span> no </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="o">[</span>https<span class="o">]</span> </span></span><span class="line"><span class="cl"><span class="nv">accept</span> <span class="o">=</span> <span class="m">443</span> </span></span><span class="line"><span class="cl"><span class="nv">connect</span> <span class="o">=</span> 127.0.0.1:80 </span></span><span class="line"><span class="cl"><span class="c1"># ...</span></span></span></code></pre></div> <p>Now that haproxy sets SNI to chromic.org, the certificate configured in stunnel (*.lxc.chromic.org) doesn’t match.</p> <p>I could work-around the issue by disabling this new haproxy behaviour by adding “no-sni-auto” to every affected “server” block.</p> <p>In this case I simply opted to reconfigure stunnel to use my external cert instead of the internal one, so the SNI matches the cert.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: haproxy%203.3%20sets%20SNI%20by%20default">Reply by email</a> </p> Broken acme.sh with Alpine Linux 3.24 https://chromic.org/blog/acme-alpinelinux-324/ Sat, 27 Jun 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/acme-alpinelinux-324/ <p>If your acme.sh client randomly started throwing errors after upgrading to Alpine Linux 3.24, it might be due to issue <a href="https://gitlab.alpinelinux.org/alpine/aports/-/work_items/18285">aports#18285</a>.</p> <p>The fix is to install the real <a href="https://pkgs.alpinelinux.org/package/v3.24/main/x86_64/openssl">openssl</a> package. It should purge the <a href="https://pkgs.alpinelinux.org/package/v3.24/community/x86_64/aws-lc-tools">aws-lc-tools</a> package that got installed:</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">Running <span class="sb">`</span>apk add openssl<span class="sb">`</span> at 2026-06-28 01:02:30 </span></span><span class="line"><span class="cl">apk-tools 3.0.6-r0, compiled <span class="k">for</span> x86_64. </span></span><span class="line"><span class="cl"><span class="o">(</span>1/5<span class="o">)</span> Purging aws-lc-tools <span class="o">(</span>5.0.0-r0<span class="o">)</span> </span></span><span class="line"><span class="cl"><span class="o">(</span>2/5<span class="o">)</span> Installing openssl <span class="o">(</span>3.5.7-r0<span class="o">)</span> </span></span><span class="line"><span class="cl"><span class="o">(</span>3/5<span class="o">)</span> Purging aws-lc <span class="o">(</span>5.0.0-r0<span class="o">)</span> </span></span><span class="line"><span class="cl"><span class="o">(</span>4/5<span class="o">)</span> Purging libstdc++ <span class="o">(</span>15.2.0-r5<span class="o">)</span> </span></span><span class="line"><span class="cl"><span class="o">(</span>5/5<span class="o">)</span> Purging libgcc <span class="o">(</span>15.2.0-r5<span class="o">)</span> </span></span><span class="line"><span class="cl">Executing busybox-1.37.0-r31.trigger </span></span><span class="line"><span class="cl">OK: 20.0 MiB in <span class="m">53</span> packages</span></span></code></pre></div> <p>For me the error was:</p> <blockquote> <p>Register account Error: {“type”:“urn:ietf:params:acme:error:malformed”,“status”:400,“detail”:"[External Account Binding] Invalid MAC on JWS request"}</p> </blockquote> <p>Now I need to figure out why all my <a href="https://pkgs.alpinelinux.org/package/v3.24/community/x86_64/stunnel">stunnel</a>s are broken.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Broken%20acme.sh%20with%20Alpine%20Linux%203.24">Reply by email</a> </p> Jellyfin for Tizen https://chromic.org/blog/jellyfin-for-tizen/ Thu, 28 May 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/jellyfin-for-tizen/ <p>Now that <a href="https://jellyfin.org/posts/state-of-the-fin-2026-05-24/#jellyfin-for-tizen">Jellyfin is officially available on the Samsung Tizen store</a>, I decided to install it on my Tv. If everything goes well (the first tests were promising), that might mean I’ll be able to repurpose my RaspberryPi (running <a href="https://kodi.tv/">Kodi</a>) for other uses than a media player.</p> <p>Also, fewer wires and cables in the living room!</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Jellyfin%20for%20Tizen">Reply by email</a> </p> nftables Logs on Alpine Linux https://chromic.org/blog/nftables-logs-on-alpinelinux/ Wed, 27 May 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/nftables-logs-on-alpinelinux/ <p>If your <a href="https://nftables.org/projects/nftables/index.html">nftables</a> “log” action isn’t <a href="https://wiki.nftables.org/wiki-nftables/index.php/Logging_traffic">logging traffic</a> on Alpine Linux, it may be because sysklogd isn’t running:</p> <pre class="repl"><samp><span class="comment"># Install the init script</span><br /><span class="prompt">root@logic-b0mb:~# </span><kbd class="input">apk add sysklogd-openrc</kbd> <span class="comment"># Make it start at boot</span><br /><span class="prompt">root@logic-b0mb:~# </span><kbd class="input">rc-update add klogd boot</kbd> <span class="comment"># Start it now</span><br /><span class="prompt">root@logic-b0mb:~# </span><kbd class="input">rc-service klogd start</kbd></samp></pre> <p>Woo!</p> <pre class="repl"><samp><span class="prompt">root@logic-b0mb:~# </span><kbd class="input">tail -n 3 /var/log/messages</kbd> May 27 17:24:14 logic-b0mb kern.warn kernel: [ 6072.309129] rej-in: IN=eth0 OUT= MAC=xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx SRC=192.168.xx.xx DST=192.168.xx.xx LEN=160 TOS=0x00 PREC=0x00 TTL=64 ID=22361 DF PROTO=UDP SPT=36150 DPT=514 LEN=140 May 27 17:24:30 logic-b0mb kern.warn kernel: [ 6088.276676] rej-in: IN=eth0 OUT= MAC=xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx SRC=192.168.xx.xx DST=192.168.xx.xx LEN=56 TOS=0x00 PREC=0x00 TTL=64 ID=23582 DF PROTO=UDP SPT=33654 DPT=3478 LEN=36 May 27 17:24:37 logic-b0mb kern.warn kernel: [ 6095.330465] rej-in: IN=eth0 OUT= MAC=xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx SRC=192.168.xx.xx DST=192.168.xx.xx LEN=160 TOS=0x00 PREC=0x00 TTL=64 ID=23656 DF PROTO=UDP SPT=44940 DPT=514 LEN=140 </samp></pre> <p> <a href="mailto:chimo@chromic.org?subject=Re: nftables%20Logs%20on%20Alpine%20Linux">Reply by email</a> </p> Containers are great https://chromic.org/blog/containers-are-great/ Fri, 01 May 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/containers-are-great/ <p>I ran into an issue on one of my system-containers earlier today:</p> <blockquote> <p>apk Unable to lock database: Resource temporarily unavailable</p> </blockquote> <p>I thought about troubleshooting the issue but decided this time around to take advantage of the disposable nature of containers and delete the problematic container and rebuild it.</p> <p>The problem went away.</p> <p>Great.</p> <p>Except there’s a part of me that really doesn’t like the fact that I didn’t learn anything from the situation.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Containers%20are%20great">Reply by email</a> </p> Liked: Markdown Ate the World https://chromic.org/likes/markdown-ate-the-world/ Mon, 23 Mar 2026 19:11:41 +0000 chimo@chromic.org (chimo) https://chromic.org/likes/markdown-ate-the-world/ <div class="h-entry"> <div class="h-cite u-like-of"> Liked "<a class="u-url" href="https://matduggan.com/markdown-ate-the-world/">Markdown Ate the World</a>" by <span class="p-author h-card"> <a class="u-url p-name" href="https://www.matduggan.com/">Mathew Duggan</a>. </span> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Liked%3a%20Markdown%20Ate%20the%20World">Reply by email</a> </p> Organic Maps https://chromic.org/bookmarks/organic-maps/ Tue, 10 Mar 2026 21:08:58 +0000 chimo@chromic.org (chimo) https://chromic.org/bookmarks/organic-maps/ <div class="h-entry"> <div class="h-cite u-bookmark-of"> Bookmarked "<a class="u-url" href="https://organicmaps.app/">Organic Maps</a>": <blockquote>Organic Maps is a privacy-focused offline maps & GPS app for hiking, cycling, biking, and driving.</blockquote> </div> <p>Discovered via <a href="https://perrotta.dev/2026/03/google-translate-kagi-translate-deepl/">Google Translate → Kagi Translate, DeepL</a> by <a href="https://perrotta.dev/">Thiago Perrotta</a>. </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Organic%20Maps">Reply by email</a> </p> hledger https://chromic.org/bookmarks/hledger/ Tue, 10 Mar 2026 21:01:47 +0000 chimo@chromic.org (chimo) https://chromic.org/bookmarks/hledger/ <div class="h-entry"> <div class="h-cite u-bookmark-of"> Bookmarked "<a class="u-url" href="https://hledger.org/">hledger</a>": <blockquote>Robust, friendly, fast, plain text accounting.</blockquote> </div> <p>Discovered via <a href="https://sufyaan.me/hledger/">hledger for managing money</a> by <a href="https://sufyaan.me/">Sufyaan</a>. </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: hledger">Reply by email</a> </p> Liked: Nobody Gets Promoted for Simplicity https://chromic.org/likes/nobody-gets-promoted-for-simplicity/ Tue, 03 Mar 2026 22:19:21 +0000 chimo@chromic.org (chimo) https://chromic.org/likes/nobody-gets-promoted-for-simplicity/ <div class="h-entry"> <div class="h-cite u-like-of"> Liked "<a class="u-url" href="https://terriblesoftware.org/2026/03/03/nobody-gets-promoted-for-simplicity/">Nobody Gets Promoted for Simplicity</a>" by <span class="p-author h-card"> <a class="u-url p-name" href="https://terriblesoftware.org/about/">Matheus Lima</a>. </span> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Liked%3a%20Nobody%20Gets%20Promoted%20for%20Simplicity">Reply by email</a> </p> Liked: 406 Fail https://chromic.org/likes/406-fail/ Wed, 25 Feb 2026 22:47:25 +0000 chimo@chromic.org (chimo) https://chromic.org/likes/406-fail/ <div class="h-entry"> <div class="h-cite u-like-of"> Liked "<a class="u-url" href="https://406.fail/">406.fail</a>". </div> </div> <p>Found while lurking on IRC. Gave me a few chuckles.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Liked%3a%20406%20Fail">Reply by email</a> </p> oembed-proxy https://chromic.org/bookmarks/oembed-proxy/ Mon, 23 Feb 2026 21:21:37 +0000 chimo@chromic.org (chimo) https://chromic.org/bookmarks/oembed-proxy/ <div class="h-entry"> <div class="h-cite u-bookmark-of"> Bookmarked "<a class="u-url" href="https://codeberg.org/emersion/oembed-proxy">oembed-proxy</a>" by <span class="p-author h-card"> <a class="u-url p-name" href="https://emersion.fr/about/">Simon Ser</a>. </span> </div> <blockquote>A proxy that serves oEmbed replies for arbitrary URLs.</blockquote> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: oembed-proxy">Reply by email</a> </p> logbookd https://chromic.org/blog/logbookd/ Sat, 21 Feb 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/logbookd/ <p>Today I learned about <a href="https://sr.ht/~martijnbraam/logbookd/">logbookd</a> while lurking in the <a href="https://wiki.postmarketos.org/wiki/Matrix_and_IRC">#postmarketos</a> IRC channel.</p> <p>To quote the <a href="https://wiki.alpinelinux.org/wiki/Logbookd">Alpine Linux wiki</a>:</p> <blockquote> <p>logbookd is a syslogd implementation that uses an sqlite database as a backend.</p> </blockquote> <p>The wiki also links to an <a href="https://blog.brixit.nl/looking-closer-at-the-syslog/">interesting article</a> about its background.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: logbookd">Reply by email</a> </p> tldr pages https://chromic.org/bookmarks/tldr-pages/ Wed, 18 Feb 2026 23:36:04 +0000 chimo@chromic.org (chimo) https://chromic.org/bookmarks/tldr-pages/ <div class="h-entry"> <div class="h-cite u-bookmark-of"> Bookmarked "<a class="u-url" href="https://tldr.sh/">tldr pages</a>" by <span class="p-author h-card"> <a class="u-url p-name" href="https://github.com/tldr-pages/tldr">tldr-pages</a>. </span> </div> <blockquote> Collaborative cheatsheets for console commands </blockquote> <div>via <a href="https://jvns.ca/blog/2026/02/18/man-pages/">Notes on clarifying man pages</a> by <a href="https://jvns.ca/">Julia Evans</a></div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: tldr%20pages">Reply by email</a> </p> prek https://chromic.org/bookmarks/prek/ Wed, 18 Feb 2026 23:32:03 +0000 chimo@chromic.org (chimo) https://chromic.org/bookmarks/prek/ <div class="h-entry"> <div class="h-cite u-bookmark-of"> Bookmarked "<a class="u-url" href="https://github.com/j178/prek">prek</a>" by <span class="p-author h-card"> <a class="u-url p-name" href="https://github.com/j178">j178</a>. </span> </div> <blockquote> Better `pre-commit`, re-engineered in Rust </blockquote> <div>via <a href="https://perrotta.dev/2026/02/new-apkbuild-llm-edition/">New APKBUILD: LLM edition</a> by <a href="https://perrotta.dev/">Thiago Perrotta</a></div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: prek">Reply by email</a> </p> Random Search Engine in qutebrowser https://chromic.org/blog/random-search-engine-in-qutebrowser/ Sun, 15 Feb 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/random-search-engine-in-qutebrowser/ <p>I usually post things when I’m done, or reached a decent point, with whatever I’m doing. This time though I’m only halfway there but figure I’d post anyway.</p> <p>Over time, I’ve been stumbling across different search engines. I take note of them, intend to try them out, but inevitably end up using the same one out of muscle memory.</p> <p>I thought it might be interesting if I could have my browser choose a random search engine out of a list for every web search I do.</p> <p>My attempt at implementing this consists of pointing my `<a href="https://qutebrowser.org/">qutebrowser</a>`’s <a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/main/item/.config/qutebrowser/config.py#L15-19">default search engine</a> at a <a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/main/item/.config/qutebrowser/random-search.html">local HTML file</a>, which includes JavaScript to randomly pick from a <a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/main/item/.config/qutebrowser/random-search.html#L10-14">predefined list</a> of search engines.</p> <p>As of now, I’m unable to have the browser automatically redirect me to the target search engine. Doing `window.location = search_engine;` gives me an error in the lines of “Network unavailable”.</p> <p>Dynamically inserting a link on the page and calling `.click()` on it has a similar behaviour.</p> <p>I’m wondering if this is a limitation of the file:// protocol, but I’ve not looked it up yet, and I’ve been out of the Web game for a few years now.</p> <p>As such, I’ve resorted to generating the link, and I have to click on it to get to the search results. Not the end of the world, but it would be nice if I didn’t have to.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Random%20Search%20Engine%20in%20qutebrowser">Reply by email</a> </p> newsboat bookmarks integration with qutebrowser https://chromic.org/blog/newsboat-bookmarks-qutebrowser/ Sat, 14 Feb 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/newsboat-bookmarks-qutebrowser/ <p>I’m playing with the idea of having my `<a href="https://newsboat.org/">newsboat</a>` bookmarks automatically added to `<a href="https://qutebrowser.org/">qutebrowser</a>`.</p> <p>Maybe “integration” is a bit of a strong word for what it does, but time will tell where this goes.</p> <p>Basically I added the following as my newsboat `<a href="https://newsboat.org/releases/2.42/docs/newsboat.html#bookmark-cmd">bookmark-cmd</a>`:</p> <figure class="figure-highlight"> <figcaption>~/.config/newsboat/config</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">bookmark-cmd <span class="s2">"sh -c 'echo </span><span class="si">${</span><span class="p">@</span><span class="si">}</span><span class="s2"> >> /opt/bookmarks/urls' -s"</span></span></span></code></pre></div> </figure> <p>It simply adds the bookmarked article to qutebrowser’s “bookmarks/urls” file. Normally you’d want to write to “~/.config/qutebrowser/bookmarks/urls”, but I have this at another location for <span style="white-space: nowrap">…reasons</span>.</p> <p>At the moment, to view the bookmark via qutebrowser’s `<a href="ihttps://qutebrowser.org/doc/help/commands.html#bookmark-list">:bookmark-list</a>` command, qutebrowser needs to be restarted so it loads the new file contents. In an upcoming qutebrowser release a new `<a href="https://github.com/qutebrowser/qutebrowser/blob/5d14f90ded9377b857dbe1147297d5e33535494a/doc/changelog.asciidoc#added">:bookmarks-reload</a>` command will be available so that shouldn’t be necessary.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: newsboat%20bookmarks%20integration%20with%20qutebrowser">Reply by email</a> </p> git pull --autostash https://chromic.org/blog/git-pull-autostash/ Fri, 13 Feb 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/git-pull-autostash/ <p>Today I learned about the <a href="https://git-scm.com/docs/git-pull#Documentation/git-pull.txt---autostash">--autostash</a> switch for git-pull:</p> <blockquote> <p>Automatically create a temporary stash entry before the operation begins, record it in the ref MERGE_AUTOSTASH and apply it after the operation ends. This means that you can run the operation on a dirty worktree. However, use with care: the final stash application after a successful merge might result in non-trivial conflicts.</p> </blockquote> <p> <a href="mailto:chimo@chromic.org?subject=Re: git%20pull%20--autostash">Reply by email</a> </p> EdgeRouter ER-X Firmware v3.0.1 https://chromic.org/blog/erx-firmware-v301/ Mon, 09 Feb 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/erx-firmware-v301/ <p>The following firmware versions for the <a href="https://store.ui.com/us/en/products/er-x">Ubiquiti ER-X</a> (and others) were released a while ago:</p> <ul> <li><a href="https://community.ui.com/releases/EdgeRouter-3-0-0/33ee3852-b5db-453e-ad14-430bc218c02f">EdgeRouter ER-X/ER-X-SFP/EP-R6/ER-10X: Firmware v3.0.0</a> (2025-07-29)</li> <li><a href="https://community.ui.com/releases/EdgeRouter-3-0-1/7fe6b39d-baea-4ce6-87a0-5dcdc9538c3a">EdgeRouter ER-X/ER-X-SFP/EP-R6/ER-10X: Firmware v3.0.1</a> (2025-10-30)</li> </ul> <p>For whatever reason the Web interface of my ER-X running v2.0.9-hotfix.7 (2023-08-01) did not notify me and kept showing as “Up to date”.</p> <p>In any case, I’ve upgraded today and so far things seem to have gone smoothly. I’m particularly interested in possibly replacing my IPSec site-to-site tunnel with WireGuard site-to-site instead.</p> <p> <a href="mailto:chimo@chromic.org?subject=Re: EdgeRouter%20ER-X%20Firmware%20v3.0.1">Reply by email</a> </p> Fallout S02E08: "The Strip" https://chromic.org/watching/tv/2026-02-06-gieagf/ Fri, 06 Feb 2026 00:04:39 +0000 chimo@chromic.org (chimo) https://chromic.org/watching/tv/2026-02-06-gieagf/ <div class="tv h-product h-tv-episode"> <div> <img class="u-photo" src="https://chromic.org/watching/tv/2026-02-06-gieagf/w185.jpg" /> </div> <div> <dl> <dt>Series</dt> <dd class="p-series-name">Fallout</dd> <dt>Season</dt> <dd class="p-season">2</dd> <dt>Episode</dt> <dd class="p-episode">8</dd> <dt>Episode Title</dt> <dd class="p-name">The Strip</dd> <dt>Episode Overview</dt> <dd class="e-description">Winning might be overrated...</dd> </dl> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Fallout%20S02E08%3a%20%22The%20Strip%22">Reply by email</a> </p> Fallout S02E07: "The Handoff" https://chromic.org/watching/tv/2026-02-05-bnlkpo/ Thu, 05 Feb 2026 00:12:06 +0000 chimo@chromic.org (chimo) https://chromic.org/watching/tv/2026-02-05-bnlkpo/ <div class="tv h-product h-tv-episode"> <div> <img class="u-photo" src="https://chromic.org/watching/tv/2026-02-05-bnlkpo/w185.jpg" /> </div> <div> <dl> <dt>Series</dt> <dd class="p-series-name">Fallout</dd> <dt>Season</dt> <dd class="p-season">2</dd> <dt>Episode</dt> <dd class="p-episode">7</dd> <dt>Episode Title</dt> <dd class="p-name">The Handoff</dd> <dt>Episode Overview</dt> <dd class="e-description">You always end up back where you started.</dd> </dl> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Fallout%20S02E07%3a%20%22The%20Handoff%22">Reply by email</a> </p> Clear /tmp at reboot on Alpine Linux https://chromic.org/blog/alpinelinux-clear-tmp-on-reboot/ Mon, 02 Feb 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/alpinelinux-clear-tmp-on-reboot/ <p>By default, Alpine Linux <a href="https://quitesimple.org/page/alpine-tmp">doesn’t (completely) clear the /tmp directory</a> on reboot (unless you mount it as tmpfs). If you would like to enable this feature, change “wipe_tmp” to “YES”:</p> <figure class="figure-highlight"> <figcaption>/etc/conf.d/bootmisc</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl"><span class="c1"># List of /tmp directories we should clean up</span> </span></span><span class="line"><span class="cl"><span class="nv">clean_tmp_dirs</span><span class="o">=</span><span class="s2">"/tmp"</span> </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="c1"># Should we wipe the tmp paths completely or just selectively remove known</span> </span></span><span class="line"><span class="cl"><span class="c1"># locks / files / etc... ?</span> </span></span><span class="line"><span class="cl"><span class="nv">wipe_tmp</span><span class="o">=</span><span class="s2">"YES"</span></span></span></code></pre></div> </figure> <p>The train-of-thought that led me there was:</p> <ol> <li>Why does it take so long to take a snapshot of my `signal-desktop` container?</li> <li><pre class="repl"><samp><span class="comment"># Check snapshot sizes</span><br /><span class="prompt">root@xps:~# </span><kbd class="input">du -chs /var/lib/incus/storage-pools/default/containers-snapshots/signal/*</kbd> 9.1G /var/lib/incus/storage-pools/default/containers-snapshots/signal/pre-patch-2026-02-01T19.54 9.1G /var/lib/incus/storage-pools/default/containers-snapshots/signal/pre-patch-2026-02-02T15.28 18.1G total</samp></pre> Wow. That doesn’t seem right.</li> <li><pre class="repl"><samp><span class="prompt">root@signal:~# </span><kbd class="input">du -chs /*</kbd> ... 7.8G /tmp ...</samp></pre> Huh…?</li> <li><pre class="repl"><samp><span class="prompt">root@signal:~# </span><kbd class="input">ls -lhaS /tmp</kbd> -rwxr-xr-x 1 root root 15.4M Jan 19 20:38 .org.chromium.Chromium.BCGOJo -rwxr-xr-x 1 root root 15.4M Jan 29 15:36 .org.chromium.Chromium.BFhFGo -rwxr-xr-x 1 root root 15.4M Jan 25 09:46 .org.chromium.Chromium.FKJadh ... ...</samp></pre> Hrm.</li> <li><pre class="repl"><samp><span class="prompt">root@signal:~# </span><kbd class="input">ls /tmp/.org.chromium.Chromium.* | wc -l</kbd> 1276</samp></pre> Sigh…</li> <li>Search the web for “.org.chromium.Chromium”</li> <li><a href="https://github.com/electron/electron/issues/16097">electron/electron/issues/16097</a> shows up: <blockquote> I have a suspicion that this happens to Electron apps, which package native modules into app.asar instead of app.asar.unpacked/node_modules/ </blockquote></li> <li>Okay, let’s check Alpine Linux’s <a href="https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/testing/signal-desktop">APKBUILD</a> for <a href="https://pkgs.alpinelinux.org/package/edge/testing/x86_64/signal-desktop">signal-desktop</a></li> <li>Oh, this package has a <a href="https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/testing/signal-desktop/README.md">README</a> with its APKBUILD: <blockquote>…<br />Signal-desktop is an electron application that is rather complex to build<br />…</blockquote></li> <li>Alright, I don’t have time for this right now. Let’s just treat the symptoms.</li> </ol> <p> <a href="mailto:chimo@chromic.org?subject=Re: Clear%20%2ftmp%20at%20reboot%20on%20Alpine%20Linux">Reply by email</a> </p> 28 Years Later: The Bone Temple https://chromic.org/watching/movies/2026-01-31-jhiofa/ Sat, 31 Jan 2026 20:03:00 +0000 chimo@chromic.org (chimo) https://chromic.org/watching/movies/2026-01-31-jhiofa/ <div class="movie h-product h-movie"> <div> <img class="u-photo" src="https://chromic.org/watching/movies/2026-01-31-jhiofa/w185.jpg" /> </div> <div> <dl> <dt>Title</dt> <dd class="p-name">28 Years Later: The Bone Temple</dd> <dt>Overview</dt> <dd class="e-description">Dr. Kelson finds himself in a shocking new relationship - with consequences that could change the world as they know it - and Spike's encounter with Jimmy Crystal becomes a nightmare he can't escape.</dd> </dl> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: 28%20Years%20Later%3a%20The%20Bone%20Temple">Reply by email</a> </p> Fallout S02E06: "The Other Player" https://chromic.org/watching/tv/2026-01-31-jckedd/ Sat, 31 Jan 2026 00:18:56 +0000 chimo@chromic.org (chimo) https://chromic.org/watching/tv/2026-01-31-jckedd/ <div class="tv h-product h-tv-episode"> <div> <img class="u-photo" src="https://chromic.org/watching/tv/2026-01-31-jckedd/w185.jpg" /> </div> <div> <dl> <dt>Series</dt> <dd class="p-series-name">Fallout</dd> <dt>Season</dt> <dd class="p-season">2</dd> <dt>Episode</dt> <dd class="p-episode">6</dd> <dt>Episode Title</dt> <dd class="p-name">The Other Player</dd> <dt>Episode Overview</dt> <dd class="e-description">I remember good people.</dd> </dl> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Fallout%20S02E06%3a%20%22The%20Other%20Player%22">Reply by email</a> </p> aerc, gpg and pinentry https://chromic.org/blog/aerc-gpg-pinentry/ Sat, 31 Jan 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/aerc-gpg-pinentry/ <p>If you’re getting an error like “Screen or window too small” while trying to use gpg (ex: <span style="white-space: nowrap">`gpg --import ./private.key`</span>), one way to fix it might be to instruct gpg to use “pinentry-tty”</p> <figure class="figure-highlight"> <figcaption>~/.gnupg/gpg-agent.conf</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">pinentry-program /usr/bin/pinentry-tty</span></span></code></pre></div> </figure> <p>Note that you might need to install <a href="https://pkgs.alpinelinux.org/package/edge/main/x86_64/pinentry-tty">pinentry-tty</a>.</p> <p>If you’re getting “failed to parse messagedetails: FAILURE sign” when <a href="https://aerc-mail.org/">aerc</a> tries to sign a message, you might need to tell aerc to use terminal pinentry:</p> <figure class="figure-highlight"> <figcaption>~/.config/aerc/aerc.conf</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl"><span class="o">[</span>general<span class="o">]</span> </span></span><span class="line"><span class="cl">use-terminal-pinentry<span class="o">=</span>true</span></span></code></pre></div> </figure> <p> <a href="mailto:chimo@chromic.org?subject=Re: aerc%2c%20gpg%20and%20pinentry">Reply by email</a> </p> Fallout S02E05: "The Wrangler" https://chromic.org/watching/tv/2026-01-25-njgidg/ Sun, 25 Jan 2026 01:43:37 +0000 chimo@chromic.org (chimo) https://chromic.org/watching/tv/2026-01-25-njgidg/ <div class="tv h-product h-tv-episode"> <div> <img class="u-photo" src="https://chromic.org/watching/tv/2026-01-25-njgidg/w185.jpg" /> </div> <div> <dl> <dt>Series</dt> <dd class="p-series-name">Fallout</dd> <dt>Season</dt> <dd class="p-season">2</dd> <dt>Episode</dt> <dd class="p-episode">5</dd> <dt>Episode Title</dt> <dd class="p-name">The Wrangler</dd> <dt>Episode Overview</dt> <dd class="e-description">Big Whoop.</dd> </dl> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Fallout%20S02E05%3a%20%22The%20Wrangler%22">Reply by email</a> </p> Fallout S02E04: "The Demon in the Snow" https://chromic.org/watching/tv/2026-01-25-iikefc/ Sun, 25 Jan 2026 00:41:33 +0000 chimo@chromic.org (chimo) https://chromic.org/watching/tv/2026-01-25-iikefc/ <div class="tv h-product h-tv-episode"> <div> <img class="u-photo" src="https://chromic.org/watching/tv/2026-01-25-iikefc/w185.jpg" /> </div> <div> <dl> <dt>Series</dt> <dd class="p-series-name">Fallout</dd> <dt>Season</dt> <dd class="p-season">2</dd> <dt>Episode</dt> <dd class="p-episode">4</dd> <dt>Episode Title</dt> <dd class="p-name">The Demon in the Snow</dd> <dt>Episode Overview</dt> <dd class="e-description">Rock to spear and so on...</dd> </dl> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Fallout%20S02E04%3a%20%22The%20Demon%20in%20the%20Snow%22">Reply by email</a> </p> Real IP when nginx, darkhttpd are behind HAProxy https://chromic.org/blog/haproxy-nginx-darkhttpd-real-ip/ Sun, 25 Jan 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/haproxy-nginx-darkhttpd-real-ip/ <p>I use <a href="https://www.haproxy.org/">HAProxy</a> as an entry point for most of the traffic hitting this server. Past that, I either have <a href="https://nginx.org/">nginx</a> fronting dynamic applications, or <a href="https://unix4lyfe.org/darkhttpd/">darkhttpd</a> fronting static sites (such as this blog). This means that, unless configured otherwise, client IP addresses are known by HAProxy but applications behind it see all traffic coming from the HAProxy instance’s IP address.</p> <h2 id="haproxy">HAProxy</h2> <p>One way you can expose the client’s IP address to the applications is using HAProxy’s “<a href="https://www.haproxy.com/documentation/haproxy-configuration-tutorials/proxying-essentials/client-ip-preservation/add-x-forward-for-header/">forwardfor</a>” directive:</p> <figure class="figure-highlight"> <figcaption>/etc/haproxy/haproxy.cfg</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-yaml" data-lang="yaml"><span class="line"><span class="cl"><span class="l">frontend main</span><span class="w"> </span></span></span><span class="line"><span class="cl"><span class="w"> </span><span class="l">option forwardfor</span></span></span></code></pre></div> </figure> <p>This tells HAProxy to add the client’s IP address to the value of “<a href="https://developer.mozilla.org/en-US/docs/Web/HTTP/Reference/Headers/X-Forwarded-For#security_and_privacy_concerns">X-Forwarded-For</a>” HTTP Header.</p> <p>Now that the information is forwarded to the rest of the stack, it can be used in the logs, for example.</p> <h2 id="nginx">nginx</h2> <p>For nginx, we can tell it to use the IP address in the “X-Forwarded-For” header as the client’s IP in its logs with the “<a href="https://nginx.org/en/docs/http/ngx_http_realip_module.html">real_ip_header</a>” directive:</p> <figure class="figure-highlight"> <figcaption>/etc/nginx/http.d/app.conf</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-javascript" data-lang="javascript"><span class="line"><span class="cl"><span class="nx">location</span> <span class="o">/</span> <span class="p">{</span> </span></span><span class="line"><span class="cl"> <span class="p">...</span> </span></span><span class="line"><span class="cl"> <span class="nx">real_ip_header</span> <span class="nx">X</span><span class="o">-</span><span class="nx">Forwarded</span><span class="o">-</span><span class="nx">For</span><span class="p">;</span> </span></span><span class="line"><span class="cl"> <span class="p">...</span> </span></span><span class="line"><span class="cl"><span class="p">}</span></span></span></code></pre></div> </figure> <h2 id="darkhttpd">darkhttpd</h2> <p>The next version of darkhttpd might include a `<a href="https://github.com/emikulic/darkhttpd/pull/88">--trusted-ip</a>` option, which tells darkhttpd to use the value of “X-Forwarded-For” as the client’s IP address if the traffic is coming from a proxy matching the “trusted-ip” value.</p> <p>I tweaked the <a href="https://gitlab.alpinelinux.org/alpine/aports/-/tree/3.23-stable/main/darkhttpd">Alpine Linux darkhttpd APKBUILD</a> and created my own <a href="https://git.srht.chromic.org/~chimo/apkbuilds/tree/master/item/darkhttpd-git">darkhttpd-git package</a> to test this out. It’s currently serving this blog and seems to work as intended:</p> <figure class="figure-highlight"> <figcaption>/etc/conf.d/darkhttpd</figcaption> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl"><span class="nv">darkhttpd_args</span><span class="o">=</span><span class="s2">"--trusted-ip x.x.x.x"</span></span></span></code></pre></div> </figure> <p> <a href="mailto:chimo@chromic.org?subject=Re: Real%20IP%20when%20nginx%2c%20darkhttpd%20are%20behind%20HAProxy">Reply by email</a> </p> OpenRC zap https://chromic.org/blog/service-rc-zap/ Sun, 18 Jan 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/service-rc-zap/ <p>I don’t remember how I stumbled on this, but today I learned about the `zap` OpenRC command.</p> <p>From the <a href="https://github.com/OpenRC/openrc/blob/master/user-guide.md#modifying-service-scripts">OpenRC User Guide</a>:</p> <blockquote> <p>There is a special command zap that makes OpenRC ‘forget’ that a service is started; this is mostly useful to reset a crashed service to stopped state without invoking the (possibly broken) stop function of the service script.</p> </blockquote> <p>The Gentoo wiki has a <a href="https://wiki.gentoo.org/wiki/OpenRC#Manually_recovering_crashed_services">usage example</a>:</p> <pre class="repl"><samp><span class="prompt">root # </span><kbd class="input">rc-service docker status</kbd> * status: crashed <span class="prompt">root # </span><kbd class="input">rc-service docker start</kbd> * WARNING: docker has already been started <span class="prompt">root # </span><kbd class="input">rc-service docker stop</kbd> * Caching service dependencies ... [ ok ] * Stopping docker ... * Failed to stop docker [ !! ] * ERROR: docker failed to stop <span class="comment"># To remedy this situation, zap the service</span><br /><span class="prompt">root # </span><kbd class="input">rc-service docker zap</kbd></samp></pre> <p> <a href="mailto:chimo@chromic.org?subject=Re: OpenRC%20zap">Reply by email</a> </p> Fallout S02E03: "The Profligate" https://chromic.org/watching/tv/2026-01-11-bkbbdg/ Sun, 11 Jan 2026 00:32:33 +0000 chimo@chromic.org (chimo) https://chromic.org/watching/tv/2026-01-11-bkbbdg/ <div class="tv h-product h-tv-episode"> <div> <img class="u-photo" src="https://chromic.org/watching/tv/2026-01-11-bkbbdg/w185.jpg" /> </div> <div> <dl> <dt>Series</dt> <dd class="p-series-name">Fallout</dd> <dt>Season</dt> <dd class="p-season">2</dd> <dt>Episode</dt> <dd class="p-episode">3</dd> <dt>Episode Title</dt> <dd class="p-name">The Profligate</dd> <dt>Episode Overview</dt> <dd class="e-description">Most kids are dead by this age.</dd> </dl> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Fallout%20S02E03%3a%20%22The%20Profligate%22">Reply by email</a> </p> Automatically Pushing pet Snippets to an Existing git Repository https://chromic.org/blog/pet-git/ Sun, 11 Jan 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/pet-git/ <p>I recently started using `<a href="https://github.com/knqyf263/pet">pet</a>` as a snippet manager (thanks to <a href="https://perrotta.dev/">Thiago Perrotta</a>’s excellent <a href="https://perrotta.dev/2025/12/snippets-manager/">blog post</a>).</p> <p>It stores the snippets to a text file (`~/.config/pet/snippet.toml` by default) which makes it easy to copy/sync/version/etc. It natively supports <a href="https://github.com/knqyf263/pet?tab=readme-ov-file#sync-snippets">synchronizing</a> the snippet file as a <a href="https://github.com/knqyf263/pet?tab=readme-ov-file#gist">GitHub Gist</a> or <a href="https://github.com/knqyf263/pet?tab=readme-ov-file#gitlab-snippets">GitLab Snippet</a> (manually or <a href="https://github.com/knqyf263/pet?tab=readme-ov-file#auto-sync">automatically</a>).</p> <p>I’m already keeping my dotfiles in a <a href="https://git.srht.chromic.org/~chimo/dotfiles">git repo</a>, however, and would like to keep my snippet.toml file in the same spot as everything else. I also don’t want to have to remember to manually push the changes every time I add/delete/edit something.</p> <p>I run a lot of my applications in <a href="https://linuxcontainers.org/incus/">Incus containers</a> (including `pet`) and have <a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/main/item/.local/bin">wrapper scripts</a> in `~/.local/bin/` to <a href="https://chromic.org/blog/containers-on-my-machine/#interacting-with-the-containers">run them</a>.</p> <p>So I <a href="https://git.srht.chromic.org/~chimo/dotfiles/commit/136ac3fdd5c077e6831edfb019f26c388a837ba0">modified</a> my `pet.sh` wrapper script so that on every invocation, it:</p> <ol> <li><a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/main/item/.local/bin/libs/pet.sh#L15-21">Runs `pet`</a> like it’s meant to be used.</li> <li><a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/136ac3fdd5c077e6831edfb019f26c388a837ba0/item/.local/bin/libs/pet.sh#L24-30">Checks</a> whether `snippet.toml` has local modifications.</li> <li>If so, prompt if we want to: <ol> <li>See <a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/136ac3fdd5c077e6831edfb019f26c388a837ba0/item/.local/bin/libs/pet.sh#L8-12">the diff</a>.</li> <li><a href="https://git.srht.chromic.org/~chimo/dotfiles/tree/136ac3fdd5c077e6831edfb019f26c388a837ba0/item/.local/bin/libs/pet.sh#L15-21">Commit</a> the changes.</li> <li>Skip (do nothing).</li> </ol> </li> </ol> <p> <a href="mailto:chimo@chromic.org?subject=Re: Automatically%20Pushing%20pet%20Snippets%20to%20an%20Existing%20git%20Repository">Reply by email</a> </p> Fallout S02E02: "The Golden Rule" https://chromic.org/watching/tv/2026-01-10-mfddnb/ Sat, 10 Jan 2026 23:35:48 +0000 chimo@chromic.org (chimo) https://chromic.org/watching/tv/2026-01-10-mfddnb/ <div class="tv h-product h-tv-episode"> <div> <img class="u-photo" src="https://chromic.org/watching/tv/2026-01-10-mfddnb/w185.jpg" /> </div> <div> <dl> <dt>Series</dt> <dd class="p-series-name">Fallout</dd> <dt>Season</dt> <dd class="p-season">2</dd> <dt>Episode</dt> <dd class="p-episode">2</dd> <dt>Episode Title</dt> <dd class="p-name">The Golden Rule</dd> <dt>Episode Overview</dt> <dd class="e-description">You can't put a price on family…</dd> </dl> </div> </div> <p> <a href="mailto:chimo@chromic.org?subject=Re: Fallout%20S02E02%3a%20%22The%20Golden%20Rule%22">Reply by email</a> </p> Chic 'N' Stu https://chromic.org/blog/advertising-causes/ Tue, 06 Jan 2026 00:00:00 +0000 chimo@chromic.org (chimo) https://chromic.org/blog/advertising-causes/ <p>My “go-to” gas station (i.e.: the one that’s most conveniently located for me) started showing advertisement videos on the small monitors built-in the gas pumps.</p> <p>They are too loud and the audio is of terrible quality.</p> <p>It is no longer my go-to gas station.</p> <p><small>"<cite>what a splendid pie.</cite>"</small></p> <p> <a href="mailto:chimo@chromic.org?subject=Re: Chic%20%27N%27%20Stu">Reply by email</a> </p>