4 ms·
It is definitely not a joke. This started off as scratching our own itch in answering 'how predictable are programmers?' but it turned out to be really useful,
by videlov 4y ago
It is definitely not a joke. This started off as scratching our own itch in answering 'how predictable are programmers?' but it turned out to be really useful, so we made a site.
- mdaniel 4y agoGood to know; is there an example (other than its own GH Action repo) to see what it has approved? Given that it's a model, is there a feedback mechanism through which one could advise it (or you) of false positives? I would be thrilled to see what it would have said about: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/76318 https://gitlab.com/gitlab-org/gitlab/-/merge_requests/76318 (q.v. https://news.ycombinator.com/item?id=30872415 https://news.ycombinator.com/item?id=30872415)
- videlov 4y agoIt would have said nothing. The model's idea is to identify the bulk of easy / safe contributions that get approved without objection and let the humans focus on the tricky ones (like the example above). On the site you can give it GitHub repos where it will test the last 50 PRs and show you what it would have done (false negatives and false positives included). You can also give it a link to an individual PR as well, but GitLab is not yet supported.
- zegl 4y agoI've tried to reproduce #76318 as best as I could (using a fork of the CE version of GitLab). https://github.com/zegl/gitlabhq-cve-test/pull/1 https://github.com/zegl/gitlabhq-cve-test/pull/1 Codeball did not approve the PR! https://codeball.ai/prediction/8cc54ce2-9f50-4e5c-9a16-3bc481a3d8d3 https://codeball.ai/prediction/8cc54ce2-9f50-4e5c-9a16-3bc48...