I stopped reading here. I think this is crap engineering and we should strive to do better.
That’s not engineering, that’s sketching. Literally like if I have a paper and pencil and draw something very roughly, just to see the shape. I’ve been programming for long enough to know that the first, and even second version of every code I write is a sketch. Yes, I use AI these days for the sketches. The engineering happens after that.
If I download your library and start developing against it, how am I to know that these examples are rough sketches and not the best examples of how the system should work?
I’m often left scratching my head at how zig’s std works, so I end up reading the tests. I trust that those represent the author’s best example of how to use these APIs.
I don’t see any problem with including ideas/sketches/prototypes in a project, they just need to be labelled as such.
Everything in the main branch is how it should be, the sketches I posted here are in a branch that will never be merged.
I think the Zig devs and crowd are generally against using AI, because it is soooo bad at Zig. And I’m talking the depths of hell are too high kind of bad.
Now if your sketches are language agnostic, then it’s another thing. But I do think it’s a valid approach to spare your time, and not give a second thought about AI-generated, assisted or even reviewed Zig code. Until maybe someday it get decent at it.
Let’s not turn this into AI discussion, please
I would be interested in your sketches. I’m not interested in the slop that you prompted. Naturally, it’s your perogative to outsource your brain, but it’s also my perogative to bow out of the conversation. This is me letting you know why, rather than ghosting the thread.
I’ve split off this topic as I think it’s off-topic from the original. I also am closing it as I think the AI discussion in general is off-topic for this forum, unless it is more germaine to zig programming.