[quote="areusch, post:4, topic:12220"] committers have reviewed and one wants to approve the PR but not merge it until the others have looked the PR over. [/quote]
One of the main reasons of this work would be to explicitly remove this step, though it makes sense to still limit the ability to those that have already interacted with the repo in some way (i.e. they aren't a first time contributor). We could have a way for reviewers to add other required reviewers which would solve this too. Either way even if it's just committers to start with that lets us fix our commit messages which by itself is a big improvement. [quote="areusch, post:4, topic:12220"] Suppose we want to approve but hold a PR until another PR lands [/quote] This is something we just need to rely on the community for, with mostly good faith actors this workflow would work fine with comments like `accepted but please don't merge until #1234 merges` to keep the bot simple and understandable. [quote="kparzysz, post:5, topic:12220"] One question—what should happen if (1) PR is approved and looks non-controvesial, (2) author adds a merge directive, and then (3) some other reviewer requests changes? Should we expect the merge directive to be canceled? I think we should consider what is actually possible to implement in automation, maybe there is no way to handle this scenario [/quote] The time from acceptance to merge should be very short (like < 1 min), so this would be pretty rare. In any case, reverts are [quick and simple](https://github.com/apache/tvm/blob/main/docs/contribute/ci.rst#skip-ci-for-reverts) and if a committer has a good reason they should not hesitate to bounce something back to the author. A cancel mechanism is certainly possible but not worth it from a complexity perspective IMO --- [Visit Topic](https://discuss.tvm.apache.org/t/rfc-allow-merging-via-pr-comments/12220/6) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/ef5ac42a64eafa320fe0046fc681923113dd3c16693fa2adb092bbdbfdc3a919).
