Time to move on?
https://lists.debian.org/debian-devel/2026/10/msg00345.html

@sunweaver Thank you for your recent message to d-devel! I don't have more to say myself and I don't want to throw more oil into the fire, so I will remain quiet myself but I think it is useful if more well written mails like yours are sent out. Unfortunately, as you know, your arguments have all been brought up during the GR discussion on d-vote already and the result is still the one it is. So yes, I guess it is time to move on. Luckily, there are a lot of FOSS things out there that are fun :)

@sunweaver You phrase this as a question, but you debian-devel post suggests you've come to an answer already.

I am not going to try to change your mind regarding LLMs; I'm not sure that my position on them would necessarily boil down to disagreeing with you anyway. But I would like to say that I hope you don't leave the project.

@sunweaver Maybe I'm just being optimistic (if so, that's out of character) but I strongly believe that principled people in FLOSS will become more important than ever, and I reject the notion that Debian is doomed to be overrun with slop just because we sometimes have to fight it. Debian is stronger with people like you in it.

@sunweaver Danke. Gut und nüchtern geschrieben.

Wenn jetzt aber schon Linux upstream AI-Slop akzeptiert, dann fällt damit IMO schon automatisch jede Linux-Distribution raus. …

"Schön" geht da irgendwie anders.

@chbmeyer We used to have Debian/kFreeBSD. The patches were in glibc until relatively recently https://sources.debian.org/src/glibc/2.36-8/debian/patches/kfreebsd and I'm told that these patches would be useful for NetBSD support... I hear some dedicated people started working on that... Though I wonder how motivating it would be to have slop-packaging on top of a non-slop kernel... @sunweaver

@josch
Thanks a lot for your hint to the BSD direction. This is valuable.
I think the ongoing discussion is important for the whole FOSS ecosystem and to find a position where to go.

Anyway, I think when you say 'free as in free speech' and not as in 'free beer', then your position and your way is clear.

Let's hope that enough people see it similar. That could be a good motivation as well. But it's becoming much harder. …

@sunweaver

@bandali @josch @chbmeyer @werdahias My current impulse is to start to think more outside the box. Look around, read other distro's AI policies, read public communications, learn more then one (packaging) trick, etc. At FS GmbH, we have a real business case to become better in NixOS as a company team, but also openSUSE is important for OSS in Schleswig-Holstein. Mentally, I guess, I have kicked Debian of its throne it had in my personal life. Disillusioning, but clearing sight, too.

• Gentoo: https://wiki.gentoo.org/wiki/Project:Council/AI_policy
• Guix: https://codeberg.org/guix/guix/commit/dd3fa6231f12f0f0b911b19484ce2cc5de86a100
• AlpineBSD: https://wiki.alpinebsd.org/ai-policy

@bandali @sunweaver @josch @chbmeyer @werdahias

Nura is imo a gold standard when it comes to the policy: https://docs.nura.eco/policies-and-processes/development/ai-policy.html

Void is my personal favorite usage-wise (it's also systemd-free and rolling but stable): https://github.com/void-linux/.github/blob/master/CONTRIBUTING.md#ai-usage

@sunweaver Right, I think I'm unfortunately a bit more "stuck" with Debian than you are then... Since my Master's Thesis my software (botch, mmdebstrap) is tightly integrated with how the Debian dependency system works and cannot easily or at all be ported to other package managers. So for me, this is also questioning whether I leave projects of mine behind which I spend decades of my life on. And I also don't want to do a disservice to the users of said software.. @bandali @chbmeyer @werdahias

@bandali @sunweaver @josch @chbmeyer @werdahias

Chimera Linux: https://chimera-linux.org/community/

Chimera has a strong no-AI policy for contributions in all of its repositories. Contributors found using LLMs will be barred from further contributing to the project.
We take objection against the use of LLMs primarily on ethical grounds. This is based on many factors, including societal impact of LLM use, ecological impact, and others. We reject arguments about “useful tools” and similar, because technical aspects can never be considered in isolation.
LLM-using contributions additionally have a disproportionate burden on reviewers, their mental health, as well as project resources, which we have limited amounts of.
Chimera is a small project and has limited control over what upstreams do, particularly due to dependencies and widespread use of LLM usage in projects nowadays. We try to do the right thing when possible, preferring appropriate alternatives and so on.

@bandali @sunweaver @josch @chbmeyer @werdahias it's also built atop the slopless musl libc among other things

(Planning on trying it on a laptop of mine this weekend.)

EDIT: Gentoo also has a profile for musl libc based installs, but it apparently doesn't have any binary packages available for it so you will be compiling everything.

@bandali @josch @chbmeyer @sunweaver @werdahias I don't know if the surviving source code would be especially useful given code rot over time but there was at one point (ended 9 years ago) a Gentoo/NetBSD project for a minimal slop Linux upstream to match a NetBSD kernel: https://wiki.gentoo.org/wiki/Gentoo_NetBSD

@werdahias @chbmeyer @josch @sunweaver @bandali going by Chimera Linux (which uses musl)'s repositories, you can get a lot of stuff to run another libc: https://pkgs.chimera-linux.org/packages

@gourd @werdahias The rebootstrap project maintained by Helmut Grohne has a musl libc crossbuild bootstrap environment set-up in jenkins. It needs some love though. The crossbuild of cracklib for musl got recently fixed by Guillem Jover and it now gets further: https://jenkins.debian.net/job/rebootstrap_musl-linux-arm64_gcc16/10 @chbmeyer @sunweaver @bandali