Maybe instead of submitting PRs, people should submit "prompt diffs" so that the maintainer can paste the prompt into their preferred coding agent, which is no doubt aware of their preferred styles and skills, and generate the desired commit themselves.

▲

rebolek 5 hours ago | parent | next [-]

"Prompt diff" is just wish. Why not call it what it is? It’s perfectly fine to submit wishes, feature requests, RFCs or - if you want - "prompt diffs". But there’s no need for LLM to implement it, human can do it. Or not, LLM can. That’s not the point who implements it, it’s a wish.

▲

acedTrex 5 hours ago | parent | prev [-]

Why would anyone bother doing this, prompts are not code, they are not shareable artifacts that give the same results.

▲

travisjungroth 5 hours ago | parent | next [-]

Neither are bug reports or feature requests.

▲

orangesilk 5 hours ago | parent | next [-]

bug reports should be reproducable. They may even be statistically reproduceable. A bug report that cannot be reproduced is worthless.

	▲	esafak 4 hours ago \| parent [-]
		It is not worthless; that means you need to work on making it easier to detect and report bugs.

▲

mjmas 5 hours ago | parent | prev [-]

Do you accept bug reports that just say "it doesn't work" or do you require reproducibility?

▲

fc417fc802 5 hours ago | parent | prev [-]

> Why would anyone bother doing this

For the same reason a PR can be useful even if it turns out to be imperfect. Because it reduces the workload for the maintainer to implement a given feature.

Obviously that means that if it looks likely to be a net negative the maintainer isn't going to want it.