| 17:15 | <Chris de Almeida> | small typo? rebase? boom, all reviews dismissed |
| 17:38 | <ljharb> | also, even a rebase, with the diff remaining identical, will clear reviews - and shouldn't imo. iow i think that the only thing that should maybe require rereview is changes in the diff, and even then it's probably too annoying |
| 17:45 | <Mathieu Hofman> | Some bots exist to carry forward the approval when they judge the content of the PR has not "significantly changed", which is intrinsically a language and repository specific concept |
| 17:54 | <ljharb> | true. with github tho and CODEOWNERS, they could certainly determine when a diff has changed and only invalidate reviews on those files |