1152288 - Incorrect browser window toolbar transparency after switching displays on OS X Yosemite

Status: NEW

(Firefox :: Theme, defect)

Description

[j.markevicius]

Attachment: Screen Shot 2015-04-08 at 13.02.02.png

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10; rv:37.0) Gecko/20100101 Firefox/37.0
Build ID: 20150402191859

Steps to reproduce:

I dragged browser window from one display onto another


Actual results:

Browser window toolbar transparency has been partially replaced with white colour


Expected results:

Browser window toolbar transparency should have remained

Comment 1

[Loic]

Did you meet this issue with previous versions of FF?

Comment 2

[j.markevicius]

(In reply to Loic from comment #1)
> Did you meet this issue with previous versions of FF?

I haven't noticed it before, no.

Comment 3

[Loic]

Are you able to reproduce it each time?

If yes, could you download mozreg (see http://mozilla.github.io/mozregression/ for details) to find a possible regression range, please. No need to bisect, the pushlog is enough.

Comment 4

[j.markevicius]

Yes, I can reproduce it each time.

I downloaded mozregression, but I'm not sure how to get the pushlog for this. Thanks.

Comment 5

[Loic]

It's a simple tool to use, it works by dichotomy to find the last good nightly build and the 1st bad nightly build .
If you think the issue started with FF37, use "mozregression --good-release 36" (or lower).

After bisecting, th shell should provide a link like http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=dac5700acf8b&tochange=cb573b9307e5

That's the pushlog you want, no need to bisect more.

Comment 6

[j.markevicius]

Okay, so this took a little while, but got there in the end.

https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=ab137ddd3746&tochange=7f0d92595432

I looked up 'Yosemite' and found this: 6bb4109d3f5e	Wesley Johnston — Bug 1052466 - [10.10] Update titlebar/toolbar to use transparency on OS X Yosemite, r=mstange,f=gijs

I think this transparency implementation isn't as polished as it could be.

Comment 7

[j.markevicius]

In fact, this is a shorter version: https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=30276610fd29&tochange=a05b5362429f

Comment 8

[Loic]

Yes, you're right about the suspected bug. Ty.

Comment 9

[j.markevicius]

This appears to have been resolved as of FF 39.0