Anthony Ramine wrote:
"Backed out changeset 90036d4d378e (bug 17564) because gecko-side patch caused
bustage on Windows. r=backout on a CLOSED TREE"
I see issues in that:
- Mentioning a Mercurial changeset hash when just a PR number is enough
this is true for the servo side. we need it on the gecko side.
- Putting the reason of the backout directly in the title instead of the message
if there's only one backout in the pr the reason will be in both the pr
title and the body. this is the case for the PR you quoted.
if there's multiple backouts this isn't possible, so the pr title is
changed to a static message, and the pr body contains all the titles
concatenated.
- Saying "bug X" where X is a Servo PR number (so it should say #17564)
that's a user error; whoever created this backout wasn't paying attention.
we have an recently created bug open to try to catch this.
note the full pr body (you only quoted the first line) should contain
#17564 as that's added by automation.
- Saying "on a CLOSED TREE" which AFAIK serves no purpose
it's another gecko thing.
be aware this text is verbatim the commit description of the backout
commit; you'll see artefacts that are required to for it to be landable
on mercurial.
the only change is we look up the commit being backed out (from the
revision in the commit desc), extract the PR, and append "Backs out
#17564" to the pr's body.
--
glob — engineering productivity — moz://a
_______________________________________________
dev-servo mailing list
dev-servo@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-servo
