1217418 - Make wpt update script more robust in the face of patches that won't apply.

Status: RESOLVED FIXED

(Testing :: web-platform-tests, defect)

Description

[James Graham [:jgraham]]

Attachment: wpt-update.diff

Comment 1

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Comment on attachment 8677466 [details] [diff] [review]
wpt-update.diff

Review of attachment 8677466 [details] [diff] [review]:
-----------------------------------------------------------------

Thanks

Comment 2

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/f65121860ffe

Comment 3

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/f65121860ffe

Comment 4

[Pulsebot]

Backout:
https://hg.mozilla.org/integration/mozilla-inbound/rev/28db7cfb2e4a

Comment 5

[Sebastian Hengst [:aryx] (needinfo on intermittent or backout)]

Backed out in https://hg.mozilla.org/integration/mozilla-inbound/rev/28db7cfb2e4a because https://hg.mozilla.org/integration/mozilla-inbound/rev/af7f10137637#l14.8 made Linux x64 debug W-e10s(3) permafail and this patch landed between the two patches of bug 1217767 which caused the failure.

Comment 6

[James Graham [:jgraham]]

WTF Why did this get backed out when it a) had nothing to do with the permafail and b) when the permafail was clearly incorrect metadata that should just have been fixed. Given it was backed out, why didn't the bug get reopened?

Comment 7

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/762d46436477

Comment 8

[Sebastian Hengst [:aryx] (needinfo on intermittent or backout)]

(In reply to James Graham [:jgraham] from comment #6)
> WTF Why did this get backed out when it a) had nothing to do with the permafail
That wasn't obvious to two sheriffs because this checkin had been sandwiched between the two for bug 1217767 which are two pulled updates for which the second one needed an update to this script. There wasn't any information in this bug and if it affected the other one, so it also got backed out.

> b) when the permafail was clearly incorrect metadata that should just have been fixed.
Whoever would have fixed it would also have needed review (on a Saturday).

> Given it was backed out, why didn't the bug get reopened?
Sorry, that was my mistake. I hadn't noticed that it had already been merged to the other integration branches.

Comment 9

[James Graham [:jgraham]]

(In reply to Sebastian H. [:aryx][:archaeopteryx] from comment #8)
> (In reply to James Graham [:jgraham] from comment #6)
> > WTF Why did this get backed out when it a) had nothing to do with the permafail
> That wasn't obvious to two sheriffs because this checkin had been sandwiched
> between the two for bug 1217767 which are two pulled updates for which the
> second one needed an update to this script. There wasn't any information in
> this bug and if it affected the other one, so it also got backed out.
> 
> > b) when the permafail was clearly incorrect metadata that should just have been fixed.
> Whoever would have fixed it would also have needed review (on a Saturday).

Trivial bustages fixes don't need review in general.

Sorry to go on about this, but the fact that two weeks have passed since this backout has created a real mess that I need to fix when trying to do the next wpt update.

Comment 10

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/762d46436477