RESOLVED DUPLICATE of bug 246913222096
Small conformance issues in web-animations/interfaces/Animation/commitStyles.html 
https://bugs.webkit.org/show_bug.cgi?id=222096
Summary Small conformance issues in web-animations/interfaces/Animation/commitStyles....
Philip Jägenstedt
Reported 2021-02-18 01:15:31 PST
Safari and WebKitGTK have some failures in WPT's web-animations/interfaces/Animation/commitStyles.html: https://wpt.fyi/results/web-animations/interfaces/Animation/commitStyles.html?run_id=5684229927075840&run_id=5715930678034432&run_id=5099062846029824&run_id=5655709666508800 Over in https://github.com/mdn/browser-compat-data/pull/7233#issuecomment-780920038, Brian Birtles looked over the failures and very helpfully broke them down. The conclusion was the issues don't look very severe, but I'm filing this issue so that his analysis can be found from this bug tracker in case someone looks into it. Inlining the comment here too for searchability: - Incorrect handling of logical properties -- fairly minor? - Incorrect handling of custom variables -- probably due to not implementing registered properties -- Firefox also fails this for that reason - Incorrect unit conversion / serialization for em units, line-height, transforms -- only an issue if you are actually reading back the value from style - Incorrect handling of animations in a stack of composited animations -- pretty minor case - Incorrect handling of redundant changes with regards to mutation observers -- edge case
Attachments
Add attachment proposed patch, testcase, etc.
Antoine Quint
Comment 1 2021-02-18 02:37:50 PST
Thanks for raising this Philip. For composited animations, we simply don't support them (yet). The other issues are known as well, but a bug tracking this test is welcome.
Radar WebKit Bug Importer
Comment 2 2021-02-25 01:16:14 PST
<rdar://problem/74735215>
Antoine Quint
Comment 3 2022-06-26 10:41:50 PDT
"Incorrect handling of custom variables" will be fixed in bug 242007.
Antoine Quint
Comment 4 2022-10-25 01:46:15 PDT
There is only one remaining failing subtest, which is tracked by bug 246913, but it will require some doing since we need do implement the mix() function from CSS Values Level 4.
Antoine Quint
Comment 5 2025-08-20 06:57:35 PDT
*** Bug 297657 has been marked as a duplicate of this bug. ***
Antoine Quint
Comment 6 2025-08-20 07:26:14 PDT
After the changes made to match the latest spec behavior, here are the remaining failures: FAIL Commits logical properties as physical properties assert_equals: expected "20px" but got "10px" FAIL Commits custom variables assert_approx_equals: expected 0.8 +/- 0.0001 but got 0.5 FAIL Commits matrix-interpolated relative transforms assert_equals: Resolved transform is correct after commit. expected "matrix(2, 0, 0, 2, 100, 0)" but got "matrix(2, 0, 0, 2, 0, 0)"
Antoine Quint
Comment 7 2025-08-22 01:20:13 PDT
As of 299046@main, the sole remaining issue is the one covered by bug 246913, so I'm going to mark this bug as dupe of that bug. *** This bug has been marked as a duplicate of bug 246913 ***
Note You need to log in before you can comment on or make changes to this bug.