Add MR review policy
This adds an explicit policy describing expectations and responsibilities for MR review, based on a slight restructuring of the earlier draft. While this is phrased as a normative policy, obviously it is only my suggestion until it is merged, and I'm happy to make changes in response to feedback.
@simonpj @bgamari @mpickering I'd appreciate GHC HQ input into this. Perhaps we should circulate it to ghc-devs or more widely as well?
Merge request reports
Activity
- review-policy.mkd 0 → 100644
11 README](https://gitlab.haskell.org/ghc/ghc-hq/-/blob/main/README.mkd) for 12 background). Membership of the GHC Team is open to everyone actively 13 contributing to GHC. Of course, many GHC contributors are not members of the GHC 14 Team, but they are still strongly encouraged to review MRs where possible. 15 16 17 ## 1. Responsibilities of MR Authors 18 19 The authors of merge requests are the part of the life-blood of GHC. Our goal 20 is that their efforts should be focused on the MRs themselves; we do not want 21 them to get discouraged by a lack of reviews, or inability to land a MR. 22 23 At the same time, reviewing is painstaking work, also done by volunteers, and 24 authors cannot expect a blank cheque on reviewers time. 25 26 MRs often go through an extended period of drafts, sharing with specific This blocked-on-review thing feels a bit out of place in the flow.
Move to the end of this sub-section?
Also (important) you need a link to https://gitlab.haskell.org/ghc/ghc/-/wikis/Contributing-a-Patch, and a pointer back in the other direction. Do we want these two to be separate? Maybe one is "policy" and one is "mechanism"?
Possibly the whole blocked-on-review thing belongs in the mechanism wiki page?
..later: aha I see that it is linked at the top; it should definitely be here as well
Can we merge
- https://gitlab.haskell.org/ghc/ghc/-/wikis/Contributing-a-Patch
- https://gitlab.haskell.org/ghc/ghc/-/wikis/gitlab/merge-requests
(Maybe worth doing now, to avoid polluting this doc with pointers that will become dead.)
- review-policy.mkd 0 → 100644
50 responsible for moving the MR forward in some way. 51 52 Individuals assigned to perform review (whether in the GHC Team or not) should 53 either do so promptly, or if they are not in a position to review (e.g. due to 54 lack of time or necessary expertise) reassign the MR appropriately. Being 55 assigned as a reviewer does not compel the individual to submit a review, but 56 they should feel an obligation not to be a bottleneck. 57 58 “Moving the MR forward in some way” includes requesting changes or giving the 59 author feedback that the MR is unlikely to be accepted, then assigning it back 60 to the author. Reviewers are not necessarily expected to finish off MRs that are 61 not yet in a state to merge. 62 63 64 ## 3. Responsibilities of the GHC Team 65
