Jesus Michał von Gentoo 🏔 (he)

#Gentoo has a strict #NoAI policy, so our projects are slop-free. Unfortunately, these days you cannot run an up-to-date system without tons of random slop packages, because everyone insists on using that crap. Gentoo supports a whole bunch of architectures that probably every other distribution dumped so far. Unfortunately, these days the available software list is shrinking, as everyone insists on adding some random meaningless dependencies on #RustLang (like using uv-build, which is the single most harmful meaningless #Python project ever created). Gentoo has a strong upstream cooperation policy. However, these days I'm afraid to file issues, because just too often they just trigger more slop. I miss the "good" old days where the worst you could expect is some toxic asshole. The state of #FreeSoftware is truly depressing. There is little joy in distribution work these days. You're overwhelmed with shit, burned out and end up doing a shitty job. And it's bad for your health. #FLOSS

#Python in a nutshell:

1. Let's use this new fancy #TOML thing for everything, it's so cool!
2. Everyone starts depending on third-party TOML libraries.
3. TOML library is finally added to the standard library, so people are moving to conditionally using it.
4. As old versions of Python become irrelevant, people are finally dropping fallbacks to the third-party library.
5. TOML 1.1 is introduced, and everyone has to start using a third-party library again… (at least until Python 3.15)

That's in case anyone is wondering why dev-python/tomli is suddenly getting installed on #Gentoo systems again.

Okay, let's get this straight. I have trouble taking compliments for my work. I was raised in 1990s Poland, you could say it was on the border between two eras. It was a time of newly realized freedom, when people thought that the possibilities were now endless and you had to grab them, you had to compete for them. You didn't compliment kids for their work, that would spoil them. You could compliment kids for their looks (not me, though, I apparently had "other qualities"). But not for their successes; you instead pointed out where they didn't succeed, or that they could do even better. Because that obviously motivates them to try harder, rather than causes them to stop trying at all, because what's the point. And of course you never complimented your employees, because they never deserved that, they were always the lazy little bastards, undeserving of the job you've given them, only looking how to steal your money. Well, you get my point. So when I started interacting with people from other cultures, it's been a shock. I've been raised to be an asshole. Fortunately, at least some of them knew this part of "Eastern European" culture and bore with me. And I've got compliments, and real gratification, and my mindset slowly started changing. Honestly, when someone compliments me for my work, I still am uneasy and distrustful (are they being sarcastic?), but I can believe it's honest. And here comes all the messed up #AI hype, and bootlicking machines. And most of them are mass generating false compliments and pleasantries, and what not. And people are repasting this slop everywhere, and learning from it, and talking like it. And now, whenever I see people being nice, I'm not just uneasy. I keep wondering: is this real, or is it slop? Am I talking to a person, or to a bot? Is this actually being nice, or is it awfully disrespectful? I'm so tired of this timeline. How do you people stay sane? #NoAI #NoLLM

Does crossfitness mean exercising while hanging on a cross? #shitposting

The state of #Python ecosystem right now: almost everything is slop. Some projects are complete vibe-coded slop (autobahn, chardet, cryptography). Some projects are disguised slop ("Coding agents shouldn't co-author themselves.") There is only a handful of human projects left (Flask being one bright example). And they're being killed. They're being killed by the neverending slew of slop pull requests. They're being killed by all their dependencies becoming slop. They're being killed because this whole ecosystem has became such a complete cesspit that you have zero motivation to do anything. EDIT: and honestly, this is only going to get worse. When you kill the incentive to work on projects, you only get more slop. Maintainers give in, and more projects become slop. Or they give up, and projects get forked unto slop. #NoAI #NoLLM

Let's get this straight: anything created with "AI" is #slop. It doesn't have to be vibe-coded. It doesn't have to be horrible code. It doesn't matter that you've spent hours perfecting the prompt. It doesn't matter that ten people have spent a week reviewing it in detail. As long as an "AI" took part in writing it, it's slop. It's an unethical technology, and anything created with it is tainted. It's an inhuman technology, and you're supporting it. You're being disrespectful to your users and contributors. You're putting your own convenience over humanity. You're aiding the worst murderous assholes in the history of humanity. And what's perhaps worst, you quite likely believe that you are making a net positive contribution, that you're helping people, while you're actually disrespecting and harming them. #NoAI #NoLLM

If anyone's wondering, #Xfce is a #slop junk now too, and the author is busy writing long excuses for them, with full realization of how harmful they are. https://www.spurint.org/journal/2026/07/llms-and-xfwl4 #NoAI #NoLLM

Zero days since it turned out that #GnuPG has arbitrarily changed behavior between versions in the name of "safe default", and required you to pass an additional option to make it behave properly in the rare cases it matters. Of course we still need to support old versions and alternative implementations without that option. And of course there is no clean way of checking whether it's available or not. https://bugs.gentoo.org/983021 #OpenPGP #GPG #Gentoo

I'm trying to keep an open mind. When someone I used to respect starts using LLMs, I'm thinking: okay, their moral compass is bad, but at least they're putting an effort to review that stuff and they're trying to improve other projects. When that person writes a pro-AI blog post, complaining that people aren't respecting them anymore, I'm still trying to stay neutral. But honestly, when said person starts ranting that Codeberg dared ban vibecoded shit, I can't but despise them at this point. #NoAI #NoLLM

Sigh at ensloppification of random software you use. Soon you discover that #Vim maintainers consider it a good idea to use black text on blue background now. In dark theme.

Let's say you're building a house. Normally this involves hiring a whole bunch of experts (either directly or via some company). You need to pay a fair wage for their labor, and you have to pay for all the material they're going to use. Some of them will be more experienced than others. They will sometimes make a mistake. They will sometimes get sick. The whole construction work will probably take a few months, and may get delayed in the process. Some construction material will be wasted. There's also a risk that over the next years, you'll discover that they screwed something up, and have to fix it. But you'll get a reasonably good house. Now, a kid comes and tells you you don't have to do this. Instead of paying all these people, you can hire him. He's wearing a suit and sunshades. He never laid a brick in his whole life. But he's got this new great technology that will give you a great house in no time. Of course it will cost you, but not as much as the labor of all these people. You try to learn more, and you discover that the technology actually builds and razes a hundred of house-like things a day, until it gets close enough to something that actually resembles a house. There's a lot of wasted energy and material, but you don't have to worry because it's all subsidized (for the time being). You pay less, you get a house that looks just right on the outside and inside, and that seems to work just right. Of course, there's a real risk that in a few months you'll discover that something is wrong. You may have to pay a lot for a real expert to figure out how things were done, and how to fix them. Or they may not be able to help you. You may be stuck having to call the suit again, and he'll happily come and fix it for you. And by "fix", I mean he will just raze a whole part of the house and start randomly spitting it out again until he gets something that looks about right, and works about right. Of course, it may not really be a good fit for the rest of the house, but you don't want to pay for replacing the whole thing, do you? It may also introduce a completely new issues that will surface in a few months, but that's just how it is. Welcome to the glorious world of #AI. #NoAI #NoLLM #LLM

I don't know what's more stupid: #OpenSSH using different authentication flow depending on whether you have a .pub file in addition to the private key or not, or #GitHub suddenly starting to reject one of the two valid RFC 4252 workflows.

https://thorsell.io/2026/07/21/github-ssh-keys.html

#SSH

In the meantime, #GitPython, the package that used to be dead (because the author is busily working on their next great thing) is now slopping out 4 "security" releases a week, as their #slop machine is busily fixing unintended variable expansion in every single URL they call. Unfortunately, they probably don't have enough tokens to fix them all at once, so instead all downstreams have to deal with the churn of endless security release. Or the brain to figure out that maybe they could just disable variable expansion by default and solve them all at once. It's truly a great time to be a #Python packager. #NoAI #NoLLM

@karo@floss.social, no, I didn't think of "bullying anyone into silence", though now that makes me wonder what goal your reply has. Am I not allowed to express my feelings now? Should I just bottle up and look the other way while people I used to respect are jumping on the single most destructive bandwagon of my lifetime? Yeah, perhaps I've read that post wrong. Hard not to when I'm being swamped with slop, and spending all my free time either working on projects that are constantly DDoS-ed by slop scrapers and attacked by slop-powered spambots. When I'm spending hours of my precious human time dealing with projects that are becoming slop, and whenever I make a careful human contribution, it gets converted into slop anyway. And my only way to survive all this is to go out. Into yet another heat wave. Wondering how much more heat will my body will be able to take. What's my goal? My goal is primarily to survive. But maybe, just maybe, I will be able to get some people thinking about what they're doing. Maybe they'll one day realize that perhaps their own convenience, and the hype their employer engaged in, is not worth burning the planet and killing people. I do realize that it's unlikely that my toots actually change anyone's mind; people either already agree with all this or they don't. But at least I'm trying.

Did I ever announce gentoo-kernel-modprep? It's the newest variation on #Gentoo Distribution Kernels: the absolute minimum kernel tree necessary for building modules. Ideal for package-building containers where a full kernel is unnecessary, yet you need to be able to build kernel modules.

Also, zero days (well, minus a lot of days because I've been slacking) since a random #Python package got a new #NIH dependency on another #Python package which is actually a "fastest" #RustLang package which cannot be built because the author thinks it a great idea to require a nightly Rust compiler. https://github.com/awolverp/cachebox/issues/62

Zero days since 83 #Gentoo developers effectively lost push access over revoked PGP key signatures because of random #LDAP hiccup. Yes, we have error handling. And yes, we have a safety check in case error handling didn't work and LDAP returned empty list of developers. Yet apparently that's not enough either, so now we also check if the diff between old and new lists isn't too large; you know, in case ldapsearch randomly returned a subset of developers and considered it a success.

Remember the time when I took #security vulnerabilities seriously, rather than shrugged and whatever'd? #Linux

I've added a "contributing" file to #Gemato: https://github.com/gentoo/gemato/blob/master/CONTRIBUTING.rst I guess I'm going to use this as a template for other projects of mine, hosted on the #Gentoo infrastructure and mirrored to GitHub or Codeberg.

#TIL that another way to make #tar (GNU and libarchive) create an empty archive is to pass a file and --exclude that same file. tar -c --exclude=README.md README.md

PSA: I've just removed 344 #Gentoo #GURU committers who haven't connected to the git server in the last 6 months. This is specifically related to the change of rules. We have 205 regular committers now, plus 6 trusted contributors.

Do people actually understand #RustLang types and borrowing and whatever, or do you just write the first approximation and then follow compiler errors like me?

#TIL that #git has an --end-of-options option because they hijacked -- in git checkout. 🤦

https://nesbitt.io/2026/07/21/end-of-options.html

(thanks to @thesamesam@social.treehouse.systems for the link)

Zero days since I've discovered a random #CPython bug that triggers in very specific circumstances and I don't really have time to report it, nor can expect anyone to do it for me. This time it's a test failure and refcounts being broken while running fastjsonschema's test suite on #Python 3.15 freethreading with pytest's --junit-xml option enabled 😱. And of course removing --junit-xml makes all tests pass.

Now I really want to create a #Gentoo project called "The Butlerian Inquisition". #NoAI #NoLLM

I recommend getting an #aronia shrub. • you don't have to do anything about it most of the year • after a few years it gives enough fruit to add to ice cream the whole year round • the fruits come in large bunches and have relatively tough skin, so it's easy to collect • most of people are prejudiced against it, so you have it all to yourself #gardening

Today in things that couldn't have aged well: "Unpin flit-core dependency" 2 years ago without actually updating the metadata to use the newer pyproject.toml format. https://github.com/sethmlarson/socksio/commit/b326406915fd98a8185c1c160165c5b8963b30c1 #Python #flit

Here's a perfect #blog #AntiSpam idea: put an URL field, and say "to pass, you must leave that field empty". Challenge difficulty level: impossible. #spam #slop #NoAI #NoLLM

An average two-feeted European hearing Americans talk about all their tall feet: 🤯.

#TIL that "geek" used to mean a carnival freak. (by watching Nightmare Alley (1947)) https://en.wiktionary.org/wiki/geek #English

Okay, it's still rough around the edges, needs documentation, tests and other stuff I'll slowly be adding but here it is: gagit, or #Gentoo #AGit workflow tool. https://gitweb.gentoo.org/proj/gagit.git/ #NamingIsMyPassion

Oh great, new kernels, "all users must upgrade", except that their CDN is still completely broken and you can't fetch sources to upgrade to. #Linux

I know that some car people need big cars to compensate for their insecurity, but naming a company "cis cars"…

Did you know that you can use #PyTest 's plugin manager instance to check whether a particular plugin was loaded? You might have known that.

And did you know that pytest plugins have no "canonical" name, and every plugin's name is literally whatever was passed while it was registered? So if a plugin was autoloaded, it's its entry point name. If it was loaded via -p or PYTEST_PLUGINS, it's whatever the argument was (i.e. entry point name in the first case, object reference in the second one). And if it was loaded via an explicit call to plugin loader, then it was whatever you passed to the plugin loader.

https://github.com/syrupy-project/syrupy/pull/1142/changes

As a Senior Software Engineer 2, I can also call myself a WNI. #shitposting

Today's earworm: riffs from Strange Horizons by #ReverendBizarre. The best part is I didn't listen to it in a while, yet it randomly attacked my brain around the morning.

My mom asked me for help with #Microsoft #Excel on her work computer. I don't know which version it was, beyond being an actual program and having these damn ribbons. She wanted to make all the rows on a spreadsheet taller. I did the obvious thing: selected all the cells, dragged at some random row header edge and resized it. To my surprise, this affected only one row. I tried again, same result. WTF?! I mean, I haven't seen Excel for years, but why would they break that?! So I started digging through these damn ribbons, properties, trying to find some row height setting. Nothing. I was close to giving up, but I've tried again: selected everything, dragged… and this time it worked and affected all rows. So much for predictable software.

2000s: big game companies publish their old games for free download, because they aren't selling them anymore, and who would buy them anyway? Also, that's good marketing to sell new games. 2020s: big game companies are impotent to make new games that aren't crap. Instead, they keep selling old games, remasters of old games (both "enhancing" and "retro" remasters) and reboots of old games. #games

And the next big blocker for #Python 3.15 in #Gentoo is time-machine. Which is obviously broken. There's a patch but it's complex and it's sitting for 3 months already with no reply. Because obviously patching CPython internals is so much a better idea than freezegun ever were. All these time-based tests need all the ricing you can get; portability doesn't matter. https://github.com/adamchainz/time-machine/pull/618

What's worse than books written with temperatures expressed in degrees Fahrenheit? Books written with temperatures expressed in "degrees", and you have to guess the scale from context.

Zero days since #CPython broke the extension ABI at the very last minute after asking everyone repeatedly to test prereleases. And of course they don't bother tracking ABI changes in any way, so they could at least give people heads up that important pieces of their system will start segfaulting after the upgrade. All they can offer is a knee-jerk response of "we never promised ABI stability". And yes, this is a huge PITA for #Gentoo users who opted in to helping out. https://github.com/python/cpython/pull/152448#issuecomment-5016129511

#TIL that the #FAT #filesystem doesn't really store anywhere whether it's #FAT12, #FAT16 and #FAT32. What you're supposed to do is to look at the number of clusters, and determine which range it falls into. More specifically, you're supposed to look at the partition size and divide it by the number of sectors per cluster (which is stored in consistent position irrespective of FAT type). The obvious implication of that is that the volume sizes are limited by possible cluster size ranges. There's some overlap here, though: you get more clusters (and therefore use a "longer" FAT) by using a smaller cluster size, or fewer clusters by using a larger cluster size. https://elm-chan.org/docs/fat_e.html#fat_determination However, there's a curious trick: there's a 16-bit "FAT size" field that indicates how many sectors are used for FAT on FAT12/FAT16 filesystems. On FAT32 this field is zero (there is a separate field for 32-bit FAT size there), and apparently some tools recognize the filesystem as FAT32 in that case, irrespective of number of clusters. This is what mkfs.vfat does when you force FAT32 on a small block device. https://github.com/torvalds/linux/blob/1590cf0329716306e948a8fc29f1d3ee87d3989f/fs/fat/inode.c#L1675-L1680 Why did I learn all this? Because a friend tried to make a very small EFI partition and was surprised it didn't work. I was compelled to get to a bottom of this!

Back when I was diagnosed with #diabetes, I've found two apps to help me. Both were proprietary. The first one came from the glucometer's manufacturer, and it featured the ability to copy its readings over Bluetooth. It also had a pretty useful bolus calculator. However, it was also annoying in a number of ways: it required an account with all the implied data sharing, was premiumware (though using their glucometer implied a free "pro" version), was quite childish in design, didn't respect disabled animations or dark theme. The second one was an independent Polish app to compute carbs from meals. It wasn't perfect, but it had a rich database of products and quite a few convenient features (like copying meals or calculating carbs based on recipes). Unfortunately, the authors went full way into AI hype, and while I didn't use any of the "AI" features (they were premiumware anyway), the app itself was becoming increasingly crappy. Eventually, this motivated me to look for a new app. I've settled for #Diaguard. It's just got the basics: meal composition and computing carbs, plus letting me log blood sugar and insulin doses manually. It used to have a bolus calculator before I used it, but it was removed over legal concerns. Still, it's open source, it's nice, it's got no crap and it respects the dark theme. And honestly, given that I've already reached the point of overriding bolus calculator, I've figured out it's enough for me. In fact, it's so "enough" that I'm not even using it regularly. I mean, if my blood sugar is predictable and my meals are predictable, there's no reason to waste time logging them. The app is there to assist me when I need it, not force me to use it. I still keep the two other apps installed. The first one in case I had doubts over insulin doses and wanted to use the bolus calculator. The second one over my database of meals; though I had already exported it, so I just need to figure out if the export is complete enough. https://github.com/Faltenreich/Diaguard #Android

When you accidentally push the right mouse button while moving it down to a browser window, and you accidentally perform a "scroll down" gesture which, as it turns out, was precisely what you were going to need.

Neoliberals: "We agree with all what you said, but this needs to be handled via regulation. We won't be inconveniencing ourselves with boycotting it." Neoliberals when regulation occurs: "They're taking our freedom!" #NoAI #NoLLM

Sing a pore? I mean, how do you do that, actually?

I love how people keep shitting up #slop comments that praise my #LLM criticisms. I mean, I probably wouldn't have figured it out if not there were two suspiciously similar comments, and both ofc linked some random slop SaaS. #AI #NoAI #NoLLM

#Python Hypothesis package now requires #RustLang. This is a scale of reverse dependencies I can't handle. I guess this means it's the end of WD40 profiles on #Gentoo, and therefore the end of support for Alpha, ARM

#GitHub Status: there was no downtime recently. My mailbox: two hours worth of cronjob failures that disagree.