Visual glitches are trigger on attempt to share the Chrome window
Categories
(Core :: WebRTC: Audio/Video, defect, P3)
Tracking
()
People
(Reporter: asoncutean, Unassigned)
Details
(Keywords: regression)
Attachments
(2 files)
Affected versions
- 79.0a1
- 78.0
Affected platforms
- Windows 10
Steps to reproduce
- Have Chrome in focus on one side of the screen
- Open Firefox on the other side
- Go to https://janus.conf.meetecho.com/screensharingtest.html
- Click on the Start button
- Insert a title and click on the Share your screen button
- Click on the Pick a window (or Share screen) button from the displayed popoup
- Select Chrome from the Window or Screen to share dropdown
Actual result
- Visual glitches are trigger on the Chrome tab bar
Expected result
- No visual glitches
Regression range
- Not sure if this regression is the correct one, since on older builds I can trigger the glitch intermittent: https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=908557c762f798605a2f96e4c943791cbada1b50&tochange=783356f1476eafd8e4d6fa5f3919cf6167e84f8d
- Note that the glitch worsen on earlier builds:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=9b47750842e6f8df0d31828f8f5bbdf2aceb9b04&tochange=daec5df56be2cdde62c2e5d3c2e02e339fc5ff9a
Additional notes
- I reproduce this issue only on one machine
- Recommended severity: s3
Reporter | ||
Comment 1•5 years ago
|
||
Reporter | ||
Updated•5 years ago
|
Comment 2•5 years ago
|
||
From the first range, bug 1284877 seems relevant, and from the second range, bug 1559011 seems relevant. Maybe bug 1315737 opened up access to some features. Either way, I expect media can better evaluate these pushlogs. There did not seem to be anything graphics that is relevant (but it is a long list and I could be wrong).
Comment 3•5 years ago
|
||
At S3 and without a clear fix, we're not going to be taking this in 78.
Comment 4•5 years ago
|
||
Hi Anca. I'm not able to repro on my Windows 10 box. Does it happen for you here?
Comment 5•5 years ago
|
||
The first regression range is too old to matter. Bug 1559011 might be relevant. Dan WDYT?
Reporter | ||
Comment 6•5 years ago
•
|
||
(In reply to Jan-Ivar Bruaroey [:jib] (needinfo? me) from comment #4)
Hi Anca. I'm not able to repro on my Windows 10 box. Does it happen for you here?
Apparently I can no longer reproduce it myself with your link or the one used when I was filling this issue on the latest Beta/Nightly builds. I will be looking more into it next week before mark it as worksforme, maybe it is some "trigger" that I am missing.
Comment 7•5 years ago
|
||
(In reply to Jan-Ivar Bruaroey [:jib] (needinfo? me) from comment #5)
The first regression range is too old to matter. Bug 1559011 might be relevant. Dan WDYT?
That is certainly possible, but I thought Chrome windows could not be shared at all before Bug 1559011 landed.
Comment 8•5 years ago
|
||
Marking P3 until we get confirmation from Anca that it is still reproducible.
Reporter | ||
Comment 9•5 years ago
|
||
I can no longer reproduce this issue, not on the latest Nightly/Beta versions, nor on the ones available at the time of submitting this problem (tried with Windows 7, Windows 8.1, and two test machines with Windows 10). I’ve tried to recreate the context in which I’ve triggered the issue initially (as far as I could remember), but with no payoff. Probably, there is indeed a particular “trigger” for this issue, but since I can’t determine the “culprit”, therefore it no longer happens on my end, I will mark the bug as Worksforme. Feel free to reopen it if the case!
Updated•5 years ago
|
Description
•