Repository navigation
Validate empty teams - #2913
Validate empty teams#2913clarfonthey wants to merge 2 commits into
Conversation
55df602 to
f13de76
Compare
|
Okay, this one is confusing me:
|
f13de76 to
ca3b12c
Compare
|
I'm just going to mark this as ready for review because I genuinely don't understand this issue, and presumably one of the five people mentioned by CODEOWNERS can help. It should be otherwise ready. |
|
The tests run on fake teams and repos in the |
|
cc @nikomatsakis as T-spec lead to approve this Essentially, the approval would be for moving the (empty) spec-contributors team into the archive. Note it can be un-archived if new members are added, although the permission changes would also need to be reverted. (Given how the team is three people atm, I have a feeling that this extra team is probably too much extra maintenance anyway.) |
ca3b12c to
0b8d43a
Compare
|
This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed. Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers. |
Dry-run check results |
Would require T-spec signoff: #t-spec > spec-contributors subteam
Mostly just solving this bug we found a while back: #t-website > Fun edge case: alumni-only team that isn't archived
Essentially, there are two cases in the code where it seems valid to leave a team without active members:
launching padis a parent team who has no direct members, which is finerustconf-emailsandfoundation-email-redirectsare marker teams with no members that only exist to track email redirectsHowever,
spec-contributorssatisfies none of the following:So, the proposal here is to say we should archive any team where this is the case, and specifically, to archive spec-contributors as the current sole example. But, I'll mark this as a draft until we get approval that this is okay from the spec team.