There is a lot of consternation lately about the growing onslaught of low-quality (spam) and completely AI submitted bug reports, pull requests, etc. This is unrelated to AI scrapers which can be another problem, but I'm more interested in the concerns about the bottleneck of "code review". Perhaps you can't take my advice and shutdown your public repos. I've experienced firsthand the burden of dealing with 100% human-made spam getting published on our GitLab which no amount of captcha or email verification can stop. Unfortunately, "open registration" is a pretty big vulnerability. If you opt instead to require admins to approve accounts, you're just asking someone to sit there all day and click APPROVE or DENY hundreds to thousands of times. (It really gets that bad!) We don't yet have a functional "federated" git forge and I'm skeptical it will ever work well. To me this feels like reinventing the wheel, and the options appear limited. We're Pretty Good At Dealing With Spam I mean…
No comments yet. Log in to reply on the Fediverse. Comments will appear here.