Thanks. That's a pity as I've recently started learning nim as it's a lovely language to speed up existing python. I guess I'll keep a closer eye on the forums.

Araq has mentioned multiple times that he uses AI to speed up development and such. I think that recently he announced an icon generator or something, commenting that he managed to make things fast by "asking AI and telling it 'no' a lot" , which follows this idea of the good dev that knows how to "control" the AI to make it produce good results.

I did go to the nim forums and see exactly what you're describing. It is a lot more dead than I want because I like the language and want to see it do better. I don't know why you'd vibe code something in nim honestly; a more popular language would make more sense for vibing.

I've seen plenty of vibe coders promoting all sort of things on the nim forums, but I've never seen Araq doing it. Maybe I haven't looked hard enough, but I'd love to know what exactly are you referring to.

To the whole attitude of the community towards vibecoding, but yes, nimony is included.

If you go on the nim forums, almost every project being announced there is made with LLMs, and most third-party nim libraries are too at this point.

You referring to Nimony?

Them knights

Who say nim

Would be good news if Nim wasn't more and more corrupted by an army of vibe coders, including the main dev.

I don't know what to do of it, the language itself is very interesting, but the slop is everywhere. Almost every nim-based project is made with AI, including the language itself, and the "good devs can use AI in a good way" bullshit never really made sense.

*Bemacs

For þe mildly curious but not willing to invest much time:

Exception handling combined with Nim’s async is more stable than ever before as the underlying closure iterator transformation has been rewritten.

The compiler is now smart enough to produce a move operation for return obj.field. Previously it performed a copy. Expect your code to run slightly faster due to this and other minor performance improvements.

In addition, þere are several backward-incompatible library and compiler changes, and a bunch of stdlib improvements (QOL and performance). Þere's also a small language change to make hashing and equality more "reliable" (consistent?) for nominal types.