1483457 - moz-phab changes r? to r= before review is granted

Status: RESOLVED FIXED

(Conduit :: moz-phab, defect)

Description

[Karl Tomlinson (:karlt)]

rebasing 431458:863b579d4cc7 "bug 1483174 use float constants with float delay time variables r?padenot"
rebasing 431464:40e83d47a3d3 "Bug 1483174 use float constants with float delay time variables r=padenot" (tip)

Comment 1

[Mark Côté [:mcote]]

Yeah this is an interesting case.  r? (in the commit message) is a convention we invented for MozReview.  People who didn't use MozReview probably wouldn't use it.  That said, it's probably fairly obvious, so perhaps we should always rewrite to (or add) r?<reviewer>.

Comment 2

[Karl Tomlinson (:karlt)]

FWIW, the convention would have existed prior to 2011 at least.
e.g. https://bugzilla.mozilla.org/show_bug.cgi?id=497498#c5

https://wiki.mozilla.org/EngineeringProductivity/Projects/MozReview
implies that MozReview didn't exist until 2014.

Comment 3

[Mark Côté [:mcote]]

I just meant putting r?<user> into a commit message.  It wouldn't have made much sense pre-MozReview since we were only uploading patch files at that time.  Putting that into the commit message with MozReview triggered the system to assign the reviewer(s) automatically, which was new functionality.

Anyway, as I say, no reason not to use it; it should be pretty clear given its long history in BMO's flag system, as you indicated.

Comment 4

[Karl Tomlinson (:karlt)]

FWIW, I also meant putting r?<user> into a commit message.
https://bug497498.bmoattachments.org/attachment.cgi?id=512352

Perhaps it was bzexport that started it.  I don't know the timing of bzexport.

Comment 5

[Piotr Zalewa [:zalun]]

Attachment: commitmsg: Change reviewers flag from `r=` to `r?`

Comment 6

[:glob ✱]

Comment on attachment 9004770 [details]
commitmsg: Change reviewers flag from `r=` to `r?`

Byron Jones ‹:glob› 🎈 has approved the revision.