2 ms·
I develop/maintain several a few small FOSS libraries. I've not received any AI-generated patches/PRs - that I know of, anyway - but I have always had a bias ag
by einpoklum 5mo ago
I develop/maintain several a few small FOSS libraries. I've not received any AI-generated patches/PRs - that I know of, anyway - but I have always had a bias against people coming out of the blue with a PR, as opposed to:
1. Opening an issue.
2. Talking about what they want/need that's not catered to right now.
3. Asking for my thoughts or suggestions - even if they already have a potential PR to submit.
and that is for a small codebase where changes are rarely that big of a deal in terms of amount of effort.
I've gotten a few decent 'cold-submit' PRs as well, but my bias has usually borne out, in that these are usually PRs to reject, and only some of the time get adapted into something useful, following some back-and-forth of course.
So, on the one hand, the measure the LB people are taking seems extreme to me; but the previous state of affairs they allude to seems equally weird. (I mean, unless it's a "here is a two-liner fix for a bug" kind of patches).