1377873 - Issues with videos when changing device orientation

Status: RESOLVED INCOMPLETE

(Firefox for Android Graveyard :: Audio/Video, defect, P5)

Description

[Sorina Florean [:sflorean]]

Environment: 
Device: Asus ZenPad 8 (Android 6.0.1);
Build: Nightly 56.0a1 (2017-07-03);

Steps to reproduce:
1. Launch Fennec and open a video from www.nytimes.com; 
2. Put the video in full mode;
3. Change device orientation from portrait to landscape;

Expected result:
The video keeps playing and no interruptions or glitches are triggered.

Actual result:
When device orientation is changed the video has a white area. 

Notes:
See the video: https://drive.google.com/open?id=0B0Rtl9OVJNMCUmFMN0l6aFQyTDQ

Comment 1

[Blake Wu [:bwu][:blakewu]]

Sorina,
I think this issue should happen on non-HLS content as well. Can you check if you can see this problem on Youtube or not?

Comment 2

[Sorina Florean [:sflorean]]

I can see but it's more obvious on HLS video, it remains stuck for 2-3 seconds while on Youtube it's a glitch. 
 I will change the name of bug. Will be fixed for HLS as well?

Comment 3

[James Cheng[:JamesCheng]]

Hi Sorina,

Thanks for reporting this issue and I can also reproduce it easily by playing Youtube video with white area appeared when changing orientation.

Would you please share the tool you record the device? It seems handy and the quality is great.


Thanks.

Comment 4

[Blake Wu [:bwu][:blakewu]]

Xindorn,
jhlin mentioned offline that it could be a full screen layout problem. 
Can you help us check it?

Comment 5

[Blake Wu [:bwu][:blakewu]]

(In reply to Blake Wu [:bwu][:blakewu] from comment #4)
> Xindorn,
Sorry. I typed too fast.. Xidorn! :-)

Comment 6

[Xidorn Quan [:xidorn] UTC+10]

Attachment: expensive page to reflow

I can reproduce this.

The white area's appearing temporarily after device orientation change doesn't seem to be something specific to fullscreen. It happens whenever the page takes long to render. It is more obvious in fullscreen because the area is white while background in fullscreen is black.

You can see the attached testcase, which is an example of page which is expensive to reflow after device orientation change.

So then the question is, does it render particularly slow when in fullscreen, or it always renders that slow, just more obvious with fullscreen? If it is only slow in fullscreen, there might be some bug with fullscreen, otherwise it should just be a general performance issue in layout. Either way, it would be helpful if there could be some profiling.

Another question is, can we improve user experience in this case even if we cannot make it faster? That says, can we block the orientation change on the user interface until rendering finishes, so that user would not see any glitch after orientation changes? That would probably be a question for Fennec frontend developer.

Comment 7

[Sorina Florean [:sflorean]]

(In reply to James Cheng[:JamesCheng] from comment #3)
> Hi Sorina,
> 
> Thanks for reporting this issue and I can also reproduce it easily by
> playing Youtube video with white area appeared when changing orientation.
> 
> Would you please share the tool you record the device? It seems handy and
> the quality is great.
> 
> 
> Thanks.

Hi, James
After I connect my device to PC I use this command [1] and the video is saved on device. And CTRL+C to stop recording.
[1] adb shell screenrecord /sdcard/record7.mp4

Comment 8

[Blake Wu [:bwu][:blakewu]]

We should fix bug 842782 and see if we still see this problem.

Comment 9

[Wesly Huang (EPM)]

[triage@0712] +, Media team is tracking it.

Comment 10

[Blake Wu [:bwu][:blakewu]]

I can see this problem when playing CNN News video which doesn't use HLS, so this is a common problem, not specific on HLS.

Comment 11

[Anthony Jones (:ajones, :kentuckyfriedtakahe, :k17e)]

Note that Chrome locks to landscape when in fullscreen.

Comment 12

[Blake Wu [:bwu][:blakewu]]

(In reply to Anthony Jones (:kentuckyfriedtakahe, :k17e) from comment #11)
> Note that Chrome locks to landscape when in fullscreen.
Yeah. That's what we would like to fix in bug 842782.

Comment 13

[Emma Humphries ☕️🎸🧞‍♀️✨ (she/they) [:emceeaich] (Pacific Time) use needinfo]

This is a P1 bug without an assignee. 

P1 are bugs which are being worked on for the current release cycle/iteration/sprint. 

If the bug is not assigned by Monday, 28 August, the bug's priority will be reset to '--'.

Comment 14

[Blake Wu [:bwu][:blakewu]]

Mass change P1->P2 to align with new Mozilla triage process.

Comment 16

[Firefox Bug Husbandry Bot]

Re-triaging per https://bugzilla.mozilla.org/show_bug.cgi?id=1473195

Needinfo :susheel if you think this bug should be re-triaged.

Comment 17

[csheany]

Botond, do you have any thoughts on this in relation to Bug 1509552 and 1510029?

Comment 18

[Xidorn Quan [:xidorn] UTC+10]

See comment 6, this is because of slow reflow which causes intermediate state to be rendered.

Comment 19

[Botond Ballo [:botond]]

I agree with Xidorn. There is no relation to bug 1509552 or bug 1511029.

This is a temporary artifact of orientation changes which is fairly normal. I would almost consider it "wontfix", except for solutions like comment 12 (locking the orientation to landscape in fullscreen mode) which sidesteps the issue.

Comment 20

[BMO Automation]

We have completed our launch of our new Firefox on Android. The development of the new versions use GitHub for issue tracking. If the bug report still reproduces in a current version of [Firefox on Android nightly](https://play.google.com/store/apps/details?id=org.mozilla.fenix) an issue can be reported at the [Fenix GitHub project](https://github.com/mozilla-mobile/fenix/). If you want to discuss your report please use [Mozilla's chat](https://wiki.mozilla.org/Matrix#Connect_to_Matrix) server https://chat.mozilla.org and join the [#fenix](https://chat.mozilla.org/#/room/#fenix:mozilla.org) channel.