Canadians: Bill C-22 threatens the privacy and security of all Canadians’ communications. This dangerous surveillance bill can still be fixed. @OpenMediaOrg has a petition demanding the Canadian Senate defend our digital rights. (Petition hosted by OpenMedia; its privacy policy applies.) https://action.openmedia.org/s/10016908/znS7mbZnv

Le Ministre de l’Europe et des Affaires étrangères de France dénonce des « campagnes de harcèlement massif en ligne par Elon #Musk » pour influencer les élections en #Europe.

Si vous aviez suivi la campagne #HelloQuitteX, vous seriez au courant depuis 2 ans : https://helloquittex.com/-Je-cherche-dans-la-FAQ-.html

😮‍💨En France, il faut donc 2 ans pour que l'information remonte de la périphérie jusqu'au cerveau !

Malgré les dangers de #X pour la démocratie, #X reste le principal réseaux pour la communication gouvernementale. Le gouvernement doit-il privilégier d'autres réseaux ? Partagez si vous êtes d'accord🙏

Friends don't let friends work at google or uber ....

Or meta, palantir, openai, anthropic, mistral, spacex, microsoft, aws, coreweave, nvidia, apple, adobe ...

#tired

Ma boîte est, comme malheureusement trop de boîtes en ce moment, à fond sur l'IA. Cette année, on a des objectifs liés à l'IA. 😠
Mon chef, connaissant mes réticences, m'a simplement demandé "d'expérimenter l'IA". L'autre jour, pour réparer un outil logiciel, on a expérimenté ensemble, ma conclusion était sans appel : ça ne fait rien de plus qu'un moteur de recherche en plus lent, plus chiant et moins écolo.
Il m'a demandé de recommencer en améliorant un autre outil.
Va falloir que je lui rappelle que je dois expérimenter, mais pas que ce soit concluant.

Some #NetBSD questions that keep coming up.

We don't use modules for HW drivers in the default install. All baked into the kernel. If you're curious about desktop hardware, for x86 there are lists https://wiki.netbsd.org/laptops/ https://wiki.netbsd.org/gpus/ (and the pci/usb manual pages)

x86 virtual machines. https://www.netbsd.org/docs/guide/en/chap-virt.html#chap-virt-guest

For disk encryption there's different options, such as creating a small root with the installer, then adding encrypted partitions later (see man cgdconfig). Also: cgdroot.

Running commercial games? It is possible, but you'll probably have better luck with DRM-free indie games. https://washbear.neocities.org/valhalla

No audio output? It is possible your default device or output DAC needs to be changed, see the instructions on https://www.netbsd.org/docs/guide/en/chap-audio.html

WireGuard® VPNs? See the EXAMPLES section of https://man.netbsd.org/wg.4

ZFS? Yes, we are using it in production. No, it isn't the same branch as Linux or FreeBSD.

Je garde aussi l'image des livres conservés par Internet Archive après numérisation non destructive, à mettre en regard de leur destruction par Anthropic.

Le fichier de la Willis Library au Texas

Dans mon disque : quelques images de de gens, de fichiers et de bibliothèques… en fil… L'ancienne bibliothèque publique de Cincinnati…

P.P.S. NetBSD is a very small project, it is incredible that it supports as much hardware as it does, even if it might be incomplete (e.g. no hardware acceleration on many recent video cards, or no WiFi 6 support on many laptops that requires a USB dongle), and it is almost definitely worth improving on and appreciating.

https://tech.lgbt/@nina_kali_nina/117455437804769408

@prefec2 @nixCraft AI is best for tasks where quality doesn't matter, where the result doesn't matter.

Which ideally should be none of them.

Just so you know. Linus who is adopting AI tools. Here is how it is going right now. Linus Torvalds let Claude Code fix the Linux page cache. Then reviewers found a race https://thenewstack.io/torvalds-claude-code-kernel/

Linus Torvalds shares his thoughts on Anti-AI coding rules on Linux mailing list.

https://lore.kernel.org/lkml/CAHk-=whTs09_ozN+e09GU68N1wPP56MUzDjrg9rnKxQUa0A_4Q@mail.gmail.com/

He is doubling down on AI. This will come back to bite him. He thinks people using LLMs for coding are actually checking their work 😂 Code reviewers are just rubber stamping PRs. Nobody in their right mind can look at 100s of PRs per day &d manually go through them especially since they don't even know if they'll have a IT job left at the end of the day. It's a race to the bottom now

torvalds is really speedrunning all the major “dickhead manager discovers the slopbot” points

and of course, observers know what comes next: trying to shout over every breakage, every maintainer that leaves, every instance where using LLMs has proven to be a horrible fucking idea

nobody operating in good faith thinks an LLM is like a compiler. this is an AI booster meme from early 2026. but torvalds has a captive audience, so his bad takes don’t need to be new. expect to see this crap parroted once again by the worst guys you know. https://mastodon.social/@nixCraft/117451805824031590

@f4grx @freecad

I was going to suggest looking into SolveSpace, but then I checked their GitHub and FUUUUUUUUUUU

So yeah don't.

https://github.com/solvespace/solvespace

@f4grx @freecad This is disappointing. I was just learning #freecad for 3D print models coming from #onshape.

LLMs are pushed heavily at my work. They seem productive but all it does is push the bottleneck to reviewers, who now have the burden of making sense of slop. You're in a situation where you're tempted and pushed to ship faster: read less, skim more, and just approve. The LLMs are great at adding lines of code, but every line is liability, they miss obvious bigger picture abstractions.

Things _will_ slip by, you're facing cognitive bias no one can completely overcome. Please reconsider.

@rzbrk FWIW OpenSCAD have a similar AI policy to FreeCAD.

@fluepke @f4grx @freecad Since you mentioned Forking: there is codeberg.org/xwzCAD I am not sure what they think about AI but I am sure they could use some Help.

@paco @fluepke @gigabecquerel @f4grx @freecad FreeCAD is already great and improving noticeably with every release. I see no need to accelerate progress at the expense of our planet and people. Keep it ethically sound and artfully crafted by humans.

@fluepke @f4grx @freecad Nobody owes anyone anything here.
But I consider freecad a genuinely good thing and I really like what they're currently doing, especially for makers.
However, using, and thereby supporting AI is a genuinely bad thing in my world view, so this feels like finding out a very kind neighbour who's always helped you votes for fashists.

I don't like (or own) cars either, but I do like good CAD software and would like to own that.