Closed
Bug 1258313
Opened 8 years ago
Closed 8 years ago
Images in slider on http://www.industrialthemes.com/explicit/ tremble during animation
Categories
(Core :: Graphics: Layers, defect)
Tracking
()
VERIFIED
FIXED
mozilla48
Tracking | Status | |
---|---|---|
firefox45 | --- | unaffected |
firefox46 | --- | wontfix |
firefox47 | + | fixed |
firefox48 | --- | verified |
People
(Reporter: Ovidiu, Assigned: mattwoodrow)
References
()
Details
(Keywords: regression, Whiteboard: gfx-noted)
Attachments
(3 files)
2.49 MB,
video/mp4
|
Details | |
1.08 KB,
patch
|
jrmuizel
:
review+
ritu
:
approval-mozilla-aurora+
|
Details | Diff | Splinter Review |
2.09 KB,
patch
|
jrmuizel
:
review+
|
Details | Diff | Splinter Review |
[Note]: This is a regression. [Affected versions]: Firefox 45 is not affected Tested on Nightly 48.0a1 [Affected platforms]: Tested on MAC 10.10 and Linux 15.04 [Steps to reproduce]: 1. Go to http://www.industrialthemes.com/explicit/ 2. Check the master slider area as shown in the attached video 3. Change the slides [Expected result]: When the slides are changed you should not see trembling. [Actual result]: When the slides are changed you see the trembling. [Regression range]: Last good revision: ddd566eb39b7a39d0ff630343f179fedc8fcafd5 First bad revision: 7e18014be68d9e13760b2814862745911caf49a5 Pushlog: https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=ddd566eb39b7a39d0ff630343f179fedc8fcafd5&tochange=7e18014be68d9e13760b2814862745911caf49a5
Comment 1•8 years ago
|
||
This absolutely seems like a regression from bug 1168263.
Blocks: 1168263
Whiteboard: gfx-noted
Reporter | ||
Updated•8 years ago
|
Keywords: regression
Updated•8 years ago
|
Assignee: nobody → matt.woodrow
Updated•8 years ago
|
Component: Graphics → Panning and Zooming
Comment 2•8 years ago
|
||
I'm assuming this affects 46 and 47 as well since bug 1168263 is in those branches. The trembling is very minor. More obvious than that is that the left side of the slider over-scrolls as you are scrolling (by "over-scrolls" I mean the reverse of "lags" - it scrolls too much and gets detached from where it's supposed to be). I can repro with and without APZ on Nightly on OS X, so we should probably try to fix this in 46 even though APZ won't be going to release on that branch.
status-firefox45:
--- → unaffected
status-firefox46:
--- → affected
status-firefox47:
--- → affected
Updated•8 years ago
|
Version: Trunk → 48 Branch
Comment 3•8 years ago
|
||
(In reply to Kartikaya Gupta (email:kats@mozilla.com) from comment #2) > More obvious than that is that the > left side of the slider over-scrolls as you are scrolling (by "over-scrolls" > I mean the reverse of "lags" - it scrolls too much and gets detached from > where it's supposed to be). Markus pointed out that this is likely because they have special handling of wheel events. Behaviour is wonky in Safari and Chrome as well. We probably use this bug only for the "trembling" issue which is really very minor, at least for me.
Summary: Go to http://www.industrialthemes.com/explicit/ the images have a display problem → Images in slider on http://www.industrialthemes.com/explicit/ tremble during animation
Comment 4•8 years ago
|
||
Since this occurs without APZ as well I'm moving it to Layers. Although maybe Layout is more appropriate? It looks like some of the animations are in the compositor and some are not, so it could be either.
Component: Panning and Zooming → Graphics: Layers
Updated•8 years ago
|
Flags: needinfo?(matt.woodrow)
Don't care if we tag it Layers or Layout as long as we get it taken care of :) Matt, are you the right person for this (I assigned to you because it appears to be a regression from bug 1168263, but let me know if we should find somebody else.)
Assignee | ||
Comment 7•8 years ago
|
||
This is really hard to see on a retina MacBook pro. Reducing the screen resolution to 1440x900 makes it look like the video. The problem is that perspective layers are handled the same way as preserve 3d, so we're not flattening the combined effective transform to 2d on the transform layer [1]. This in turn disables pixel snapping for the transform, since we don't do it for 3d transforms [2] with perspective [3]. The page is kind of weird, where it's using a pair of nested that are meant to cancel each other out. The outer one has perspective and isn't being snapped, and the inner one is always snapping. [1] http://mxr.mozilla.org/mozilla-central/source/gfx/layers/Layers.cpp#1383 [2] http://mxr.mozilla.org/mozilla-central/source/gfx/layers/Layers.cpp#629 [3] http://mxr.mozilla.org/mozilla-central/source/gfx/layers/Layers.cpp#648
Assignee | ||
Comment 8•8 years ago
|
||
Attachment #8737711 -
Flags: review?(jmuizelaar)
Comment 9•8 years ago
|
||
Comment on attachment 8737711 [details] [diff] [review] Allow snapping of 'pseudo' 3d transforms Review of attachment 8737711 [details] [diff] [review]: ----------------------------------------------------------------- A test sure would be nice.
Attachment #8737711 -
Flags: review?(jmuizelaar) → review+
Assignee | ||
Comment 10•8 years ago
|
||
Attachment #8737990 -
Flags: review?(jmuizelaar)
Updated•8 years ago
|
Attachment #8737990 -
Flags: review?(jmuizelaar) → review+
Comment 11•8 years ago
|
||
https://hg.mozilla.org/integration/mozilla-inbound/rev/5d456db9e320 https://hg.mozilla.org/integration/mozilla-inbound/rev/482a89ef31e8
Comment 12•8 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/5d456db9e320 https://hg.mozilla.org/mozilla-central/rev/482a89ef31e8
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla48
Comment 13•8 years ago
|
||
Too late for 46 but if you think this would be safe for aurora please request uplift.
Assignee | ||
Comment 14•8 years ago
|
||
Comment on attachment 8737711 [details] [diff] [review] Allow snapping of 'pseudo' 3d transforms Approval Request Comment [Feature/regressing bug #]: Bug 1168263 [User impact if declined]: Incorrect pixel snapping for transforms with perspective (but no actual depth). [Describe test coverage new/current, TreeHerder]: Manually tested, added reftest. [Risks and why]: Super low risk, just lets us take the snapping path when compositing layers. [String/UUID change made/needed]: None
Flags: needinfo?(matt.woodrow)
Attachment #8737711 -
Flags: approval-mozilla-aurora?
Hi Ovidiu, could you please verify this issue is fixed as expected on a latest Nightly build? Thanks!
Flags: needinfo?(ovidiu.boca)
Comment on attachment 8737711 [details] [diff] [review] Allow snapping of 'pseudo' 3d transforms Has baked on Nightly for ~10 days, recent regression, Aurora47+
Attachment #8737711 -
Flags: approval-mozilla-aurora? → approval-mozilla-aurora+
Reporter | ||
Comment 17•8 years ago
|
||
Hi Ritu, I verified this issue on Mac OS X 10.10 and Ubuntu 15.04 with latest Nightly 48.0a1 (2016-04-17) and I can't reproduce it any more.
Comment 18•8 years ago
|
||
bugherder uplift |
https://hg.mozilla.org/releases/mozilla-aurora/rev/ae136454b6b7
(In reply to ovidiu boca[:Ovidiu] from comment #17) > Hi Ritu, > > I verified this issue on Mac OS X 10.10 and Ubuntu 15.04 with latest Nightly > 48.0a1 (2016-04-17) and I can't reproduce it any more. Thank you for a prompt verification.
You need to log in
before you can comment on or make changes to this bug.
Description
•