<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom" ><generator uri="https://jekyllrb.com/" version="4.3.3">Jekyll</generator><link href="https://muhh.lol/feed.xml" rel="self" type="application/atom+xml" /><link href="https://muhh.lol/" rel="alternate" type="text/html" /><updated>2025-06-26T17:29:26+02:00</updated><id>https://muhh.lol/feed.xml</id><title type="html">muhh internet happy place</title><subtitle>muhh internet happy place. The personal website of Markus Heurung. </subtitle><author><name>Markus Heurung</name></author><entry><title type="html">One can hope.</title><link href="https://muhh.lol/posts/one-can-hope/" rel="alternate" type="text/html" title="One can hope." /><published>2025-03-31T14:18:00+02:00</published><updated>2025-03-31T14:18:00+02:00</updated><id>https://muhh.lol/posts/one-can-hope</id><content type="html" xml:base="https://muhh.lol/posts/one-can-hope/"><![CDATA[<p><a href="https://heydonworks.com/article/poisoning-well/">poisoning well</a></p>

<blockquote>
  <p>One of the many pressing issues with Large Language Models (LLMs) is they are trained on content that isn’t theirs to consume.</p>

  <p>Since most of what they consume is on the open web, it’s difficult for authors to withhold consent without also depriving legitimate agents (AKA humans or “meat bags”) of information.</p>

  <p>Some well-meaning but naive developers have implored authors to instate robots.txt rules, intended to block LLM-associated crawlers.</p>
</blockquote>

<p>Related from my bookmarks:</p>

<ul>
  <li><a href="https://iocaine.madhouse-project.org/">iocaine</a>, which I’ve been following for quite a while since <a href="https://come-from.mad-scientist.club/@algernon">@algernon</a> does  a great job of developing in the open.
    <blockquote>
      <p>This is deliberately malicious software, intended to cause harm. Do not deploy if you aren’t fully comfortable with what you are doing. LLM scrapers are relentless and brutal, they will place additional burden on your server, even if you only serve static content.</p>
    </blockquote>
  </li>
  <li><a href="https://gts.superseriousbusiness.org/@dumpsterqueer/statuses/01JK5PXHCHWP58172Y40QKE2ZZ">tobi of gotosocial about related fediverse stats poisoning</a></li>
  <li>
    <blockquote>
      <p><a href="https://marcusb.org/hacks/quixotic.html">Quixotic</a> is a program that will feed fake content to bots and robots.txt-ignoring LLM scrapers.</p>
    </blockquote>
  </li>
  <li>
    <blockquote>
      <p><a href="https://zadzmo.org/code/nepenthes/">Nepenthes</a> […] is a tarpit intended to catch web crawlers. Specifically, it’s targetting crawlers that scrape data for LLMs - but really, like the plants it is named after, it’ll eat just about anything that finds it’s way inside.</p>
    </blockquote>
  </li>
  <li>
    <blockquote>
      <p><a href="https://anubis.techaro.lol/">Anubis</a> - Weigh the soul of incoming HTTP requests using proof-of-work to stop AI crawlers.</p>
    </blockquote>
  </li>
  <li>
    <blockquote>
      <p>The Algorithmic Sabotage Research Group (<a href="https://algorithmic-sabotage.github.io/asrg/about/">ASRG</a>) is an ongoing, conspiratorial, aesthetico-political, practice-led research framework focused on the intersection of digital culture and information technology.<br />
They maintain <a href="https://algorithmic-sabotage.github.io/asrg/posts/sabot-in-the-age-of-ai/">a good list of selected tools and frameworks</a>.</p>
    </blockquote>
  </li>
</ul>]]></content><author><name>Markus Heurung</name></author><category term="ai" /><category term="poison" /><summary type="html"><![CDATA[poisoning well]]></summary></entry><entry><title type="html">$ GOTO muhh :lol:</title><link href="https://muhh.lol/posts/goto-muhh-lol/" rel="alternate" type="text/html" title="$ GOTO muhh :lol:" /><published>2024-10-17T10:40:00+02:00</published><updated>2024-10-17T10:40:00+02:00</updated><id>https://muhh.lol/posts/goto-muhh-lol</id><content type="html" xml:base="https://muhh.lol/posts/goto-muhh-lol/"><![CDATA[<p>I’ve started testing <a href="https://gotosocial.org/">GoToSocial</a> and I think I like it!</p>

<p><a href="https://goto.muhh.lol">goto.muhh.lol</a> is a self-hosted single-user fediverse <del>instance</del> server for me.</p>

<p>The only profile over there is called <a href="https://goto.muhh.lol/@muhh">@muhh@goto.muhh.lol</a>. I don’t have a final opinion on the many muhhs in there yet. It kinda annoys me, I guess.</p>

<p>Since it isn’t Mastodon, but can talk with it, it might look unfamiliar when you view this on its website. Your replies and some interactions will show up here. It’s supposed to just work.</p>

<p>I don’t have any blocks or bans in place right now - I just started, but plan to apply them freely, even switch to exclusive allow lists. Reading about the concept of <a href="https://writer.oliphant.social/oliphant/islands-an-opt-in-federated-network">federated islands</a> got me curious about how to apply that.</p>]]></content><author><name>Markus Heurung</name></author><summary type="html"><![CDATA[I’ve started testing GoToSocial and I think I like it!]]></summary></entry><entry><title type="html">Why?</title><link href="https://muhh.lol/posts/why/" rel="alternate" type="text/html" title="Why?" /><published>2024-05-10T10:23:00+02:00</published><updated>2024-05-10T10:23:00+02:00</updated><id>https://muhh.lol/posts/why</id><content type="html" xml:base="https://muhh.lol/posts/why/"><![CDATA[<p>It’s happened again. I finished the early morning tasks at home and didn’t immediately leave for a walk. I dropped on the couch.<br />
Now it’s more than two hours later, I handled one important customer request, but the rest was lost to distractions.</p>

<p>Sigh.</p>

<p>I need to go now. Make at least a short round. Otherwise the whole day might be lost.</p>

<p>I’m furious how my brain works against me.</p>]]></content><author><name>Markus Heurung</name></author><category term="weblogpomo2024" /><summary type="html"><![CDATA[It’s happened again. I finished the early morning tasks at home and didn’t immediately leave for a walk. I dropped on the couch. Now it’s more than two hours later, I handled one important customer request, but the rest was lost to distractions.]]></summary></entry><entry><title type="html">I failed</title><link href="https://muhh.lol/posts/i-failed/" rel="alternate" type="text/html" title="I failed" /><published>2024-05-07T17:34:00+02:00</published><updated>2024-05-07T17:34:00+02:00</updated><id>https://muhh.lol/posts/i-failed</id><content type="html" xml:base="https://muhh.lol/posts/i-failed/"><![CDATA[<p>I failed hard. I noticed it on day three of the <a href="https://muhh.lol/posts/it-s-weblog-posting-month/">challenge</a>, but was not able to write. With good reasons, but they are not important.<br />
So I continued to fail. The day after and the next day as well. Weekend went by.<br />
On Monday I went into full work mode and enjoyed the things I’m not failing at. Like doing an early morning walk with the dog to clear my brain to work around the constant interruptions that make my life harder than necessary. Which is one of the best results of therapy so far. There are more. Maybe I’ll write about them one day.</p>

<p>And I’m still failing some more. It’s ok to fail.</p>

<p>I learned to stand up and walk on. Finally.</p>]]></content><author><name>Markus Heurung</name></author><category term="weblogpomo2024" /><summary type="html"><![CDATA[I failed hard. I noticed it on day three of the challenge, but was not able to write. With good reasons, but they are not important. So I continued to fail. The day after and the next day as well. Weekend went by. On Monday I went into full work mode and enjoyed the things I’m not failing at. Like doing an early morning walk with the dog to clear my brain to work around the constant interruptions that make my life harder than necessary. Which is one of the best results of therapy so far. There are more. Maybe I’ll write about them one day.]]></summary></entry><entry><title type="html">how I currently run this site</title><link href="https://muhh.lol/posts/how-i-currently-run-this-site/" rel="alternate" type="text/html" title="how I currently run this site" /><published>2024-05-02T22:55:00+02:00</published><updated>2024-05-02T22:55:00+02:00</updated><id>https://muhh.lol/posts/how-i-currently-run-this-site</id><content type="html" xml:base="https://muhh.lol/posts/how-i-currently-run-this-site/"><![CDATA[<p>The easiest way for me to better understand things, is to explain them. The way to explain things here is to write them down.</p>

<p>So I want to write down how I run this blog and how I post, to make next identify what I’m want to improve and what is missing.</p>

<p>This HTML (with some CSS and one image right now) get’s served from an OpenBSD mini VM located at Hetzner. It uses the included relayd and httpd to deliver the statically generated files.<br />
The files get pushed to a folder on this server using rsync over ssh.<br />
Before that they get generated by Jekyll, my static site generator of choice (&lt;3 Ruby!). This is happening on my main workstation that also serves as a home server and always running. This means I can add and create new posts or notes by SSH’ing into that machine and typing things into my terminal window.</p>

<p>A new note gets created by typing <code class="language-plaintext highlighter-rouge">rake note</code> which opens neovim and I can type away and thing a bit. If I’m extra easy going I can directly add the note text by using <code class="language-plaintext highlighter-rouge">rake note[$NOTETEXT]</code>.<br />
Blog posts are set up using <code class="language-plaintext highlighter-rouge">jekyll compose -l default "$TITLE"</code> which generates a file I can open and edit like before.</p>

<p>A <code class="language-plaintext highlighter-rouge">rake deploy</code> builds the site and does the rsync magic to deliver the build to the web server.</p>

<p>You can see the <a href="https://code.muhh.lol/muhh/muhh.lol">jekyll “code” in my forge</a>.</p>

<p>It’s a mess. I love it :)</p>]]></content><author><name>Markus Heurung</name></author><category term="weblogpomo2024" /><summary type="html"><![CDATA[The easiest way for me to better understand things, is to explain them. The way to explain things here is to write them down.]]></summary></entry><entry><title type="html">It’s Weblog Posting Month!</title><link href="https://muhh.lol/posts/it-s-weblog-posting-month/" rel="alternate" type="text/html" title="It’s Weblog Posting Month!" /><published>2024-05-01T07:54:00+02:00</published><updated>2024-05-01T07:54:00+02:00</updated><id>https://muhh.lol/posts/it-s-weblog-posting-month</id><content type="html" xml:base="https://muhh.lol/posts/it-s-weblog-posting-month/"><![CDATA[<p>It’s time to put some effort into my little home on the internets. 
A couple days I noticed <a href="https://anniegreens.lol/">Anne</a> starting a sort of challenge to post 
to your own blog every day for the whole of May, the <a href="https://weblog.anniegreens.lol/weblog-posting-month-2024">“WeblogPoMo2024”</a>.<br />
I was intruiged from the start, but also afraid since I already failed a
couple of times with these kind of miniprojects.</p>

<p>I’m going to try again though. Where’s the fun otherwise?</p>

<p>So, here’s my plan: write a short post (not a <a href="https://muhh.lol/notes/">note</a>, it at least needs a title…) every day. I have a few ideas queued, but I’ll try to focus on moving this site forward. There are so many ideas and hopes I want to bring into manifestation and if I can write about it and improve my still clunky english, there’s nothing to loose.</p>

<p>Have a great International Workers Day!</p>]]></content><author><name>Markus Heurung</name></author><category term="weblogpomo2024" /><summary type="html"><![CDATA[It’s time to put some effort into my little home on the internets. A couple days I noticed Anne starting a sort of challenge to post to your own blog every day for the whole of May, the “WeblogPoMo2024”. I was intruiged from the start, but also afraid since I already failed a couple of times with these kind of miniprojects.]]></summary></entry><entry><title type="html">Blogging is a conversation</title><link href="https://muhh.lol/posts/blogging-is-a-conversation/" rel="alternate" type="text/html" title="Blogging is a conversation" /><published>2024-04-09T16:13:00+02:00</published><updated>2024-04-09T16:13:00+02:00</updated><id>https://muhh.lol/posts/blogging-is-a-conversation</id><content type="html" xml:base="https://muhh.lol/posts/blogging-is-a-conversation/"><![CDATA[<blockquote>
  <p>Just write. Then share.</p>

  <p>Blogging is a conversation. It’s a conversation with yourself and it’s a conversation with others.</p>
</blockquote>

<p>– <a href="https://bjhess.com/posts/you-re-a-blogger-not-an-essayist">Barry Hess - You’re a Blogger, Not an Essayist</a></p>]]></content><author><name>Markus Heurung</name></author><category term="quote" /><summary type="html"><![CDATA[Just write. Then share. Blogging is a conversation. It’s a conversation with yourself and it’s a conversation with others.]]></summary></entry><entry><title type="html">Ha!</title><link href="https://muhh.lol/posts/ha/" rel="alternate" type="text/html" title="Ha!" /><published>2024-02-19T01:14:00+01:00</published><updated>2024-02-19T01:14:00+01:00</updated><id>https://muhh.lol/posts/ha</id><content type="html" xml:base="https://muhh.lol/posts/ha/"><![CDATA[<p>This is it, this is the new site. Still heavily under construction, but it’s here.</p>

<p>And it’s mine. My precious.</p>

<p>Danke, <a href="https://muan.co">Mu-An</a>! So much of this site is based on hers.</p>]]></content><author><name>Markus Heurung</name></author><summary type="html"><![CDATA[This is it, this is the new site. Still heavily under construction, but it’s here.]]></summary></entry></feed>