Closed
Bug 735178
Opened 12 years ago
Closed 12 years ago
CSS3 animation at http://www.cssplay.co.uk/menu/cssplay-3d-box-slide.html blinks
Categories
(Core :: Web Painting, defect)
Tracking
()
RESOLVED
FIXED
mozilla15
People
(Reporter: desiradaniel2007, Assigned: mattwoodrow)
References
Details
(Keywords: regression)
Attachments
(3 files, 1 obsolete file)
3.29 KB,
application/x-zip
|
Details | |
835 bytes,
patch
|
roc
:
review+
|
Details | Diff | Splinter Review |
6.47 KB,
patch
|
roc
:
review+
|
Details | Diff | Splinter Review |
User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:13.0) Gecko/20120312 Firefox/13.0a1 Build ID: 20120312031136 Steps to reproduce: I ran http://www.cssplay.co.uk/menu/cssplay-3d-box-slide.html on Nightly. Actual results: I noticed this example blinks. Expected results: It seems like it should not blink as I also tested in Chrome Canary. I haven't looked at the code yet though.
Comment 1•12 years ago
|
||
I can reproduce in http://hg.mozilla.org/releases/mozilla-beta/rev/0df2e93ff663 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0) Gecko/20100101 Firefox/11.0 ID:20120310173008 http://hg.mozilla.org/releases/mozilla-aurora/rev/8f1eb954ff37 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:12.0a2) Gecko/20120312 Firefox/12.0a2 ID:20120312042012 http://hg.mozilla.org/mozilla-central/rev/5ec9524de1af Mozilla/5.0 (Windows NT 6.1; WOW64; rv:13.0) Gecko/20120312 Firefox/13.0a1 ID:20120312031136 http://hg.mozilla.org/releases/mozilla-beta/rev/0df2e93ff663 Mozilla/5.0 (X11; Linux i686; rv:11.0) Gecko/20100101 Firefox/11.0 ID:20120310173008 http://hg.mozilla.org/releases/mozilla-aurora/rev/8f1eb954ff37 Mozilla/5.0 (X11; Linux i686; rv:12.0a2) Gecko/20120312 Firefox/12.0a2 ID:20120312042012 http://hg.mozilla.org/mozilla-central/rev/1ca7a94573f2 Mozilla/5.0 (X11; Linux i686; rv:13.0) Gecko/20120313 Firefox/13.0a1 ID:20120313031112 Regression window(m-c), Not flickers: http://hg.mozilla.org/mozilla-central/rev/30161b298513 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0a1) Gecko/20111117 Firefox/11.0a1 ID:20111117030939 Flickers: http://hg.mozilla.org/mozilla-central/rev/e7d5dd9efeca Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0a1) Gecko/20111117 Firefox/11.0a1 ID:20111117015450 Pushlog: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=30161b298513&tochange=e7d5dd9efeca Regression window(m-i), Not flickers: http://hg.mozilla.org/integration/mozilla-inbound/rev/aa05d607c95e Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0a1) Gecko/20111116 Firefox/11.0a1 ID:20111116173248 Flickers: http://hg.mozilla.org/integration/mozilla-inbound/rev/b201f2434265 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0a1) Gecko/20111116 Firefox/11.0a1 ID:20111116194849 Pushlog: http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=aa05d607c95e&tochange=b201f2434265 First bad changest: 17c9363d801a Matt Woodrow — Bug 701656 - Include the preserve-3d parent content in preserve-3d sorting. r=roc
Blocks: 701656
Status: UNCONFIRMED → NEW
Component: Untriaged → Layout: View Rendering
Ever confirmed: true
Keywords: regression
OS: Windows 7 → All
Product: Firefox → Core
QA Contact: untriaged → layout.view-rendering
Version: 13 Branch → 11 Branch
Fixed by bug 730166?
Comment 3•12 years ago
|
||
(In reply to Robert O'Callahan (:roc) (Mozilla Corporation) from comment #2) > Fixed by bug 730166? No, bug 730166 did not fix this issue. I can reproduce in latest Nightly. http://hg.mozilla.org/mozilla-central/rev/e94edfdb1f5b Mozilla/5.0 (Windows NT 6.1; WOW64; rv:14.0) Gecko/20120318 Firefox/14.0a1 ID:20120318031036
Assignee: nobody → matt.woodrow
Assignee | ||
Comment 4•12 years ago
|
||
Interestingly, this regression makes our behaviour the same as chrome. This looks like a bug (or more likely a limitation) of our sorting code, which is only showing up now since previously we (incorrectly) didn't include the grid background in the sorting list. If anyone can make a reduced test case (without animations), I can take a closer look at this, I haven't been able to reproduce the problem without the animations. This could likely be a dupe of bug 689498.
Comment 5•12 years ago
|
||
Comment 6•12 years ago
|
||
When page zoom-in/out, cubes appears and disappears
Assignee | ||
Comment 7•12 years ago
|
||
(In reply to Alice0775 White from comment #5) > Created attachment 609362 [details] > reduced html (non animation) Amazing. Thank you! When sorting 3d layers, we determine a collection of 2d points (points contained within the other line, and line intersections - when both layers are projected to a 2d plane), and then find which point has the greatest z difference. We determine an ordering between 2 layers based on this point. In the case where the biggest difference is tiny (less than 0.1 pixels), then we should assume that it was caused by rounding differences and treat the planes as being parallel (using the original ordering). This fixes this site by getting the bottom edges of the cubes onto the correct side of the grid, and then stops cycles being formed which fixes the rest of the test case. This still isn't a brilliant solution, having smarter resolution of cycles (current we just break the cycles arbitrarily) would help, and having layer splitting would be the best (and most involved) fix.
Attachment #609932 -
Flags: review?(roc)
Assignee | ||
Comment 8•12 years ago
|
||
Attachment #609933 -
Flags: review?(roc)
Assignee | ||
Comment 9•12 years ago
|
||
Comment on attachment 609933 [details] [diff] [review] Add coloring of sorted 3d layers to help identification Review of attachment 609933 [details] [diff] [review]: ----------------------------------------------------------------- ::: gfx/layers/LayerSorter.cpp @@ +174,5 @@ > +#define MAGENTA 5 > +#define CYAN 6 > +#define WHITE 7 > + > +#define USE_XTERM_COLORING Will remove this line, since it probably won't work on a lot of machines.
Comment on attachment 609933 [details] [diff] [review] Add coloring of sorted 3d layers to help identification Review of attachment 609933 [details] [diff] [review]: ----------------------------------------------------------------- ::: gfx/layers/LayerSorter.cpp @@ +213,5 @@ > + > +static void print_layer(FILE* aFile, Layer* aLayer) > +{ > + print_layer_internal(aFile, aLayer, aLayer->GetDebugColorIndex()); > +} This is kind of cool but it will mess with people who use terminal with black backgrounds. I don't know how to fix that. ::: gfx/layers/Layers.h @@ +946,5 @@ > PRUint32 mContentFlags; > bool mUseClipRect; > bool mUseTileSourceRect; > bool mIsFixedPosition; > + PRUint32 mDebugColorIndex; DebugOnly<> ::: gfx/layers/basic/BasicLayers.cpp @@ +2018,5 @@ > + case 4: color = gfxRGBA(0.0, 0.0, 1.0, 1.0); break; > + case 5: color = gfxRGBA(1.0, 0.0, 1.0, 1.0); break; > + case 6: color = gfxRGBA(0.0, 1.0, 1.0, 1.0); break; > + case 7: color = gfxRGBA(1.0, 1.0, 1.0, 1.0); break; > + } how about color = gfxRGBA((aLayer & 1) ? 1.0 : 0.0, (aLayer & 2) ? 1.0 : 0.0, (aLayer & 4) ? 1.0 : 0.0, 1.0);
Attachment #609932 -
Flags: review?(roc) → review+
Assignee | ||
Comment 11•12 years ago
|
||
(In reply to Robert O'Callahan (:roc) (Mozilla Corporation) from comment #10) > This is kind of cool but it will mess with people who use terminal with > black backgrounds. I don't know how to fix that. > I'll remove #define USE_XTERM_COLORING, so that it's not enabled by default. I doubt it'll work well for people other than me really.
Comment on attachment 609933 [details] [diff] [review] Add coloring of sorted 3d layers to help identification Review of attachment 609933 [details] [diff] [review]: ----------------------------------------------------------------- OK, r+ with that taken out and the other stuff fixed
Attachment #609933 -
Flags: review?(roc) → review+
Assignee | ||
Comment 13•12 years ago
|
||
Attachment #609933 -
Attachment is obsolete: true
Attachment #609949 -
Flags: review?(roc)
Comment on attachment 609949 [details] [diff] [review] Add coloring of sorted 3d layers to help identification v2 Review of attachment 609949 [details] [diff] [review]: ----------------------------------------------------------------- ::: gfx/layers/LayerSorter.cpp @@ +166,5 @@ > #ifdef DEBUG > static bool gDumpLayerSortList = getenv("MOZ_DUMP_LAYER_SORT_LIST") != 0; > > +#define BLACK 0 > +#define RED 1 Fix indent @@ +179,5 @@ > +#ifdef USE_XTERM_COLORING > + > +#define RESET 0 > +#define BRIGHT 1 > +#define DIM 2 Fix indent
Attachment #609949 -
Flags: review?(roc) → review+
Comment 15•12 years ago
|
||
2weeks Ping
Assignee | ||
Comment 16•12 years ago
|
||
https://hg.mozilla.org/integration/mozilla-inbound/rev/7a077df0c773 https://hg.mozilla.org/integration/mozilla-inbound/rev/302c5a565965
Comment 17•12 years ago
|
||
https://hg.mozilla.org/mozilla-central/rev/7a077df0c773 https://hg.mozilla.org/mozilla-central/rev/302c5a565965
Status: NEW → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla15
Comment 18•12 years ago
|
||
I see this regression back i Mac OS X mountain lion with Gecko 19.0a1 (2012-11-09) Can someone test and check perhaps reopen — I do not have right to reopen this bug.
Can you file a new bug for that?
Oh, Alice filed bug 810685
Updated•6 years ago
|
Component: Layout: View Rendering → Layout: Web Painting
You need to log in
before you can comment on or make changes to this bug.
Description
•