@lyse@lyse.isobeef.org

And maybe slow down a tad once you’re back.

Not sure if that’s possible in this company. 🤔 (This is the worst time to look for a new job, though. Alles im Umbruch.)

@lyse@lyse.isobeef.org Yeah, it probably would look better. I might fix that. It’s just laziness, the implementation was easier this way. 😅

Glad you find it interesting! And honestly, I agree, nobody but me would use this anyway. There are more mature and featureful toolkits out there.

@lyse@lyse.isobeef.org Aaaaahhhhhhhh! That explains it! 😅

As I said, I’m on vacation, and for about a week now, I did basically nothing but sleep. Day or night.

I can’t tell if this actually is a vacation or just therapy for burnout. 🤔

@bender@twtxt.net Yeah, no, not gonna happen. 😅

@lyse@lyse.isobeef.org Those poor unicorns. Why do they have to puke all the time? 😢🦄

(I know it’s kind of pointless to post all this stuff if the source code isn’t available, but meh, whatever. 🤪)

First draft of a file selection popup / widget:

https://movq.de/v/0955149868/vid-1781094010.mp4

Also makes use of the new Table widget.

@arne@uplegger.eu Heh, nice. 🥳

Actually, I’m stupid: I’m using the normal rsync on OpenBSD as well.

And regarding OpenRsync’s general usability:

https://marc.info/?l=openbsd-misc&m=178090751524547&w=2

Right now openrsync is limited in functionality and is primarily present
for rpki-client. The limited functionality makes it unusable for generic
use and so any diff or change like the above will not be considered since it
is simply not ready.

First problem to solve is to remove the mmap usage in openrsync. After
that modern protocol versions need to be added. Once that is in place one
can start a discussion about using openrsync as a default on OpenBSD.

Love this remix:

https://www.youtube.com/watch?v=TfviPc4ag3A

@lyse@lyse.isobeef.org Shadow on the wall!!1! 📣😅

It was an easy patch, so menus have drop shadows now:

Image

Not that big a difference in the dark theme, though.

@lyse@lyse.isobeef.org Not a fan of Mittelaltermärkte, but that sounds like an interesting idea. I wonder if they end up shooting each other on accident. 😅

@lyse@lyse.isobeef.org Oh, no, he was the one providing the feedback. 😅

mgk appears to be everywhere. His Unicode box drawing demo has been part of my unicode-test script for a long time: https://movq.de/git/bin-pub/file/unicode-test.html#l23

@itsericwoodward@itsericwoodward.com Yeah. That DORA quote is probably spot on. It’s exactly what I’m seeing here. 🫩

@prologic@twtxt.net Hm, yeah, probably. I don’t think that’s how many FLOSS projects are/were run, though, so they’ll have to find new ways to build those relationships. 😅 I mean, isn’t it usually a new person sending patches to a project, over and over, and at some point they’ve shown enough skill so they’re “promoted” to a full maintainer position? 🤔

@lyse@lyse.isobeef.org Two emails. 😅 One person asking for the source code, and the author of wcwidth (the library I’m using) contacted me to provide some input. 👌

@lyse@lyse.isobeef.org Switching to Make might be a good idea, though, because the whole thing is purely sequential at the moment … It takes close to 20 seconds (including the w3c verification which runs the Java checker). It’s not unusable, but it could be better. 😅

@arne@uplegger.eu Hat nicht so lange gehalten. 🤪

Image

(This settled at about 25k hits on the HTML page now. But only about 11k hits in total on favicon.ico and only around 7.5k hits on the image thumbnails. So I guess that, in reality, it might have gotten around 7k hits. The rest … is probably bots.)

@prologic@twtxt.net As have I. 🤔 I mean, since I left GitHub, I got basically 0 pull requests anyway.

Even during my time using GitHub, I noticed that “drive-by PRs” are rarely a good idea. People don’t really know/understand the code or the design principles/goals, so I often turned down PRs. Or I accepted them and was grumpy afterwards. 😅

What does work is having a team of maintainers/devs. The only question is: How do you build such a team if you don’t accept PRs? That’s going to be the interesting part.

Now that is an interesting move:

https://ladybird.org/posts/changing-how-we-develop-ladybird/

Maybe this is how all Free Software will look like in the future. It might not be the worst idea … ? 🤔

@lyse@lyse.isobeef.org Ah, I almost thought so (that you wrote it by hand), but then I looked at the source code and saw the TOC and I was like: “Naah, probably not. I would be way too lazy to do that manually.” 😅 And indeed … ha.

Oh god, yeah, that’s a lot of <span>. 🤔 Can’t really avoid that, I guess, especially if you want to do syntax highlighting of code blocks.

You wrote your own site generator, didn’t you?

In parts. I write everything in Markdown (it’s online, even: https://movq.de/blog/postings/2026-05-29/0/POSTING-en.md), plus a few Vim shortcuts (to generate thumbnails, for example), and then python-markdown renders it: https://pypi.org/project/Markdown/ This process is wrapped in a shell script, like “re-render every page if the .md file is newer than the .html file” and that’s mostly it. And the Atom feed generator is completely custom. 🤔

@bender@twtxt.net lol, no, please don’t send me a quackton of ducks. 😅 We use BIRD a lot at work, hence this bears some significance for me/us. 😅

You know what this is?

Image

A BIRD bird! 😅

I got it as a gift from a very friendly coworker and she, in turn, got it from Maria Matějka. 😃

@bender@twtxt.net Ugh, I don’t know. I’m having a long vacation now and I try not to think about this topic anymore. 🤣

Response by the author of rsync: https://medium.com/@tridge60/rsync-and-outrage-d9849599e5a0

Okay. I have lost the “battle” against “AI” at work and I will no longer try to “fight” any of it.

It is simply what people want. They want to use it. And that’s the end of it.

And why do they want it? Because it makes their job easier. And why is that? In very large parts, it’s because we have accumulated a metric fuckton of technical debt due to decades long mismanagement. We were (and are) operating in “emergency mode” all the time. There simply was no time to clean things up or to rethink designs. We always have to go with the cheapest and quickest solution. We are never ahead of things: Earlier this year, I started an initiative and wanted to tackle some issue that I could see coming. I was shut down because this wasn’t “urgent”. Very soon after, this exact thing became that exact problem – but now, there was no time anymore to do it properly because NOW it’s urgent, so, once again, we had to go with a quick and dirty solution.

It’s always like that and I had brought it up again and again. And now we have a huge spaghetti mess that hardly anyone understands anymore.

Nobody – except AI. It can still make some sense of this and, obviously, this is useful to people.

So, any argument I make against AI is completely pointless to begin with. I’m such a fool for not having seen this earlier.

The last argument I made today was: “Look, we already have so much technical debt and spaghetti systems, we really, really must clean this up. If we throw AI on top of this now, it’ll only get so much worse.” And once more, I was shut down. My intentions were “admirable”, but “there’s no time for that”.

Okay. Good luck with that. They’ll keep doing it this way. At some point, it’ll either explode entirely and some poor soul has to clean it up, or it’ll explode and they’ll have no other choice but to throw everything away and start from scratch – assuming they can still afford that.

In other words, none of this about AI, really, nor caused by it. Our department’s massive spike in AI usage is just a symptom of the underlying management issues. And since those aren’t being addressed, nothing will change and this whole mess will only get worse.

(I blame all this on management, because, well, that’s who’s to blame. I do not have a solution for it, though – and assigning blame without constructive criticism always sucks big time. I don’t like doing this. If you had put me into that particular management position, I wouldn’t have been able to solve any of this. The thing is, though, I’m not an expert on management and it isn’t my job – I’m just the “princess” who solves your technical issues.)

@lyse@lyse.isobeef.org By the way, which site generator are you using? I kind of miss having code blocks with syntax highlighting and that generic yellow highlighting thing is pretty cool, too.

@lyse@lyse.isobeef.org lol, “garbo” 😅 Took me a moment. 🤣

@bender@twtxt.net That certainly sounds much better in English, yeah. 😅

@bender@twtxt.net It started out as me calling myself “Princess Valium” because I’m so tired and braindead today, but then someone misheard that because a garbage truck drove by, and, so … one thing lead to another. 🤪 Sadly, it kind of fits, because I’m often the one who cleans up shit. 😬

@kiwu@twtxt.net In-cred-ib-ly tired. 😂

Aha, my nickname at work now appears to be “Princess Garbage Disposal” (“Prinzessin Müllabfuhr”). 🤦‍♀️ 🥴

It’s not that much traffic, of course. One hit per second on average. (Plus the images.) The nasty bots are much worse. 😅

We’re at close to 20k hits now, but it has slowed down considerably. Nobody cares about page 2. 😅

@bender@twtxt.net The good thing is that it’s already pretty battle-tested. 😅 There was this dumpster fire a few years back: https://news.ycombinator.com/item?id=31114554 This was on their front page for quite a while, just look at the number of comments … 😂

@arne@uplegger.eu Indeed. I’m glad that it’s all just static HTML. The most expensive part about this is probably TLS. 🤷‍♀️

@bender@twtxt.net

oh, I knew it wasn’t you. It is just nice to see your hobby was noticed. :-)

Ah, I see! 🤗

@bender@twtxt.net Doing tail -f access.log looks like a Matrix screensaver at the moment. Whoooooosh …

@bender@twtxt.net … boom, 5500+ hits on that blog post. 🤣 Should I start monetizing this shit?! 🤪 (Don’t worry, I won’t. German law gets super annoying if you do that kind of thing.)

@bender@twtxt.net Oh, well, thanks, I guess? 😅 (This “zdw” person isn’t me. I don’t even have an account at HackerNews. 😅)

@lyse@lyse.isobeef.org Oh, nice. That was quite the ride. :-) And all that because of locales. 😳

But, did I understand that correctly? All Atom feeds were broken, right? Because they all use that same code path with that strftime/strptime dance in it?

Ambient noise. Crows, (wild) parrots, pigeons, the occasional blackbird, some traffic, individual raindrops, thunder, heavy rain on lots of trees.

A welcome change from the daily noise of the construction site nearby.

(I wish I had better equipment. As usual.)

https://movq.de/v/9b7df4ea0a/2026-05-31–regen.flac

@lyse@lyse.isobeef.org Nature is cruel.

And the humidity sucks. It’s been a horrible day. 🥴

@prologic@twtxt.net Jesus, that’s expensive. 🥴

@prologic@twtxt.net lol, well, better than nothing, eh? What did the tickets cost? 😅

@lyse@lyse.isobeef.org @tftp@tilde.town Someone has pointed out that there’s OpenRsync:

https://www.openrsync.org/

Since I run OpenBSD on my servers, I actually do use that and have never noticed any incompatibilities with the “normal” rsync.

Haha, someone had a similar idea … https://lpcvoid.com/blog/0018_why_i_am_against_genai/index.html

@prologic@twtxt.net You actually did? 😅 Good luck. 😅 I never dared to, I’d probably get addicted. 🤣

@prologic@twtxt.net Oh yeah, same here. 😞 Let’s all just win the lottery and stop with this damn work thing. 🤣

@prologic@twtxt.net (I hope I’m not too incoherent. I didn’t sleep very well recently and have a lot of unrelated stuff on my mind. 🤣)

@prologic@twtxt.net Ah, so that’s what “Bob” is. I saw that popping up in email notifications. 😅

@prologic@twtxt.net

it’s “probabilistic” not “deterministic”

Yep, I know. And when I tell that to people and tell them “if we use AI here, we lose the ability to debug this stuff”, then all I get is: “But it’s good enough. We don’t need to debug this. Non-deterministic computing has its use cases.”

But that is just not how I’d like to model/implement our business processes. 🤔 I want something reliable, not “it mostly works”.

@prologic@twtxt.net Ahh, I see. Okay, I’m with you there. On this high level, I can understand how the thing works.

Maybe my wording isn’t good. 🤔 Let’s take a real life example from what we do at work.

There’s this AI chatbot. It gets support requests from users, so the user says something like “I need access to a particular system”. This triggers the bot to “run” the instructions stored in a large Markdown file, like “check if the user is authorized to do this, then issue the following API requests”, and so on. This is essentially like running a little script, except it’s written in natural language (German) and there’s no “script interpreter” but just the AI.

Now, suppose that the AI doesn’t quite do what was intended. There’s some subtle bug. How do you debug this? How do you find out how the AI came to the “conclusion” to run step A instead of step B? And how do you find out how exactly you have to change your prompt so this doesn’t happen again next time?

If this was an actual script/program instead of AI, you could repeat the request and attach a debugger or throw in some printf() or whatever. How do you do that kind of thing with AI? How do you pinpoint exactly what the problem was?

(Or is this just a stupid idea? Do we have to give up that way of thinking when using AI? Is the era of debuggability over?)

@prologic@twtxt.net Yeah, it’s hard to get my point across here. I tried to address that a few paragraphs down.

Yes, I can tinker with AI techniques on a general level. That’s cool but not really my area of interest.

What I certainly can’t do is learn how specific AI products work. I can’t possibly find out why Claude Code produced that particular line of code. Claude is just a magic box that does something and I have to trust it.

@lyse@lyse.isobeef.org Alright. 😅

Yeah, don’t waste time on this. I have a vacation coming up and I won’t touch this subject, either. Fuck this shit.

I really like your style of writing, btw. It’s much calmer and less aggressive then mine. :-) When I turned my bullet points into paragraphs, I got a bit mad in the process.

This is like the 32nd iteration of that list and it was much worse in the beginning. 😂

RIP, rsync. https://social.hails.org/@hailey/116657391001259044

I might eventually post this on the Fediverse to get more feedback. (Or maybe not. Depends on my mood. 😅)

@lyse@lyse.isobeef.org (Do you want to be linked on that page? Do you want your name to be there at all? 🤔)

@lyse@lyse.isobeef.org Thanks! There are a few points in there that I’ll add to my list.

Your very first point is obviously crucial. “Writing code” is just the means to an end for many people and they don’t really care about it or like it, so they love AI. I had this in another draft (it refers to the other list I posted):

https://movq.de/v/614f14c3ef/ramble.txt

And this right here is so important:

simplicity is the real art and much harder to achieve.

Finding an elegant, simple solution is waaaaaaaaaaaaaaaaaay harder than anything else. And here’s the thing: I don’t get why nerds/techies don’t get “nerd-sniped” by this. A lot of people love building big stuff and then brag about being clever/competent because they were able to build that big thing – but once you realize that this approach is the lazy one, shouldn’t you make finding the elegant solution your goal? Doesn’t that give you more bragging rights?

(Am I being clear? Do you understand what I mean? 😅)

@lyse@lyse.isobeef.org Uhhh, yes, I have one single script to build the website and I ran that while writing that noai.html page. Apart from the global updated field in my feeds (that one got changed), everything else should be stable, though.

Maybe this helps narrow things down?

Image

We’re currently at about 28-30°C, but the relative humidity is at a crazy low level of 20%. 😳 This actually feels pretty nice. If it only were always like that …

@arne@uplegger.eu Lol, indeed, now that you mention it … “You can’t do that!” “You really should $foo.”

@thecanine@twtxt.net I love these. Pixel art is amazing. It looks so simple, but it’s really, really hard. 😳

@bender@twtxt.net Oh, no. That’s my baby. 😅 It was built in 1989 and I probably had it since the early 1990ies (it came with an IBM PS/2 Model 30 that they handed out to employees, IIRC).

@prologic@twtxt.net I don’t know how to phrase the answer without sounding too bitter. 🤣 Let’s just say, nope, it won’t work.

@lyse@lyse.isobeef.org Yeah, I’ve heard of that option (“Aufhebungsvertrag”). I guess the real challenge will be finding something else that isn’t just as silly.

And on the bright side, you don’t even have to hand over anything.

They actually say that with a straight face. (Did I mention that already?) “The age of maintainership is over. Anyone can now contribute to any project at any time.”

@lyse@lyse.isobeef.org Oh, please post that list.

By the way, did you know that I have a five month notice period? Starting next year, it’ll be six months. Germany is the opposite of “hire and fire”, but it applies to both parties.

I’ve started collecting reasons against AI usage here, so I don’t have to repeat myself all the time:

https://movq.de/noai.html

@prologic@twtxt.net There is nothing I can do – except quit. The company explicitly wants it that way.

@lyse@lyse.isobeef.org Found him: https://movq.de/v/66f49383d8/vid-1779857134.mp4

It’s official now: People are vomiting AI code into a repo that I’m supposed to maintain. At the same time, I don’t have the authority to decline those PRs.

RIP.

Happy 5.25 day to anyone who celebrates.

(“Colour diskettes”. Ha.)

Image

Image

@itsericwoodward@itsericwoodward.com Just a phone? These things have become shockingly good …

And we have a search function now. This should cover most of the basic features, I think.

https://movq.de/v/05c2fd04eb/vid-1779689792.mp4

I just missed the 20 year anniversary of my blog. 😬 What a stupid long time to do this.

This started out as a PHP page with user comments, MySQL as a database, a PHP webadmin … can you believe that? Totally unnecessary. But everything was “LAMP” back then, so that’s what I was using as well. I kicked out MySQL in 2011 (it just stored files since then) and eventually switched to static HTML pages in 2015.

RSS feeds have only been there since 2009, because I was late to the party. For a long time, I didn’t understand what they were good for. 🤦

@tftp@tilde.town Ah, I see. I have a feeling that a lot of stuff is going on under the hood all the time and it’s mostly the userland-visible things that stay the same? 🤔 But yeah, some stuff is really, really old, like the TCP code I’ve recently (tried to) read.

@itsericwoodward@itsericwoodward.com Those are some really nice shots. I mean, the cat and also the shots themselves. Which cam is that? (Don’t tell me it’s an iPhone, that’d be boring. 😂)

@lyse@lyse.isobeef.org … this reminds me that I should finally write that tutorial on how to make your own X11/Wayland keymaps. That helps a lot in mitigating stupid design issues like these.

@tftp@tilde.town Please elaborate. 😅

I did the live-preview thing now:

https://movq.de/v/120a8db205/vid-1779458258.mp4

Also shows the old value for comparison.

@kiwu@twtxt.net Hooray, congratz 🥳

Ah, there’s even a term for it:

https://en.wikipedia.org/wiki/Generation_effect

The generation effect is a psychological phenomenon whereby information is better remembered if it is generated from one’s own mind rather than simply read.

hfgl with your coding agents

@prologic@twtxt.net Bloated websites? Naaaaah, that’s unheard of. 😅

It’s time soon to buy a new PC. Why? Because browsers are getting slow.

@arne@uplegger.eu All the flavors, all at once! 😅

I just realized that this book, which I’m still using as a reference every now and then, is from 2005.

In other words, it’s over 20 years old now. 😬

Image

@arne@uplegger.eu (Mixing lovely fruit with bitter things like olives or cheese is something that only Germans do, right? 😅)

@lyse@lyse.isobeef.org Yeah, that was an interesting debugging session. :-) (Those have become rare, lately …)

@tftp@tilde.town Apparently you can. 😅 Welcome. 👋

@prologic@twtxt.net Yep, that’s what I’m going for. 😃 I mean, I don’t want to make an exact clone, just catch some of the DOS vibe.

And editing bytes as a binary number: https://movq.de/v/4c3617bbd1/vid-1778939973.mp4

I don’t need this feature that often. Maybe it would be nice to have a live-preview of the new value in hex/dec. We’ll see.

This is probably a better demo: https://movq.de/v/71218c59b2/vid-1778924229.mp4

Progress: My hex editor how has undo and redo. https://movq.de/v/3af465b29a/vid-1778918267.mp4

@lyse@lyse.isobeef.org LOL! That’s brilliant. (We’re going back to that. Some Amazon employees probably already are.)

@lyse@lyse.isobeef.org

So, it’s plenty good enough for them.

Yeah, but on the other hand, you can’t even log in normally to a Matrix/Element account. I mean using username + password. It’s not expected that you ever log out or lose your browser session. If you do, you must use a one-time backup code (that you must create and save beforehand) to log in again.

To be fair, I can’t say that I fully understand what Matrix is doing in the first place. The text that I quoted reads like they have your keys. But they also claim that they only store this stuff encryped: https://element.io/en/help#encryption5 So … encrypted with what? Only option here is my password, isn’t it? (But if my password was good enough to reclaim an account … why do all the other stuff …)

Matrix takes end-to-end encryption seriously. When I ran a Matrix server for the family, the family members would regularly lose their keys, because they didn’t pay attention to something. That’s on purpose! Or rather, that was on purpose. Maybe it’s different these days?

No clue.

Forgot the source: https://hachyderm.io/@robpike/116557975987213548