Open Bug 1356627 Opened 7 years ago Updated 2 years ago

Intermittent /css/css-transitions/properties-value-implicit-001.html | border-top-width length-em(em) / events - assert_equals: Expected TransitionEnd events triggered on .transition expected "border-top-width:0.5s" but got ""

Categories

(Core :: CSS Parsing and Computation, defect)

Product:
Core
Component:
CSS Parsing and Computation
Type:
defect

Tracking

()

People

(Reporter: intermittent-bug-filer, Unassigned)

References

Details

(Keywords: intermittent-failure, leave-open) 

Pushed by archaeopteryx@coole-files.de:
https://hg.mozilla.org/mozilla-central/rev/ff6abe200419
Disable css-transitions-1 web-platform-tests for failing frequently. r=test-disabling a=frequent-orange-fix
We disabled the directory in https://hg.mozilla.org/mozilla-central/rev/ff6abe20041933a67f636b748ed698e9616a6833 and synced it back to the integration trees because more >50% of wpt11 runs were failing (mostly debug, but also saw opt) and these were at least 4 different failures.
Flags: needinfo?(james)
birtles - I think you are interested in using these tests; are you able to investigate why they're so unstable? A simple possibility is that they just need a longer timeout although it's not clear why they would *fail* in that case.
Flags: needinfo?(james) → needinfo?(bbirtles)
(I wish people would resist the urge to go and write their own complicated test harness. It always leads to spending more time debugging the test harness than the code under test. This one even does its own vendor prefixing and timeout handling :/ )

Looking at assertExpectedEventsFunc in generalParallelTest.js, it doesn't actually test if the transitionend event has been received or not. If it hasn't been received then the string it will compare will be an empty string, hence the failure.

So I'd say we're simply failing to wait long enough for the transitionend event here.

It's not worth spot-fixing this, however. This test, and I suspect many others in this folder, should be rewritten to use testharness.js and generally be a *lot* simpler. This test shouldn't even be testing events in the first place.

I'm quite ok with all these tests being disabled until we rewrite them.
Flags: needinfo?(bbirtles)
Huh, I thought we disabled all the transitions tests. I guess not. I'll get around to rewriting this test in the next month or two I expect (when I rewrite *all* the remaining transitions tests).
Summary: Intermittent /css/css-transitions-1/properties-value-implicit-001.html | border-top-width length-em(em) / events - assert_equals: Expected TransitionEnd events triggered on .transition expected "border-top-width:0.5s" but got "" → Intermittent /css/css-transitions/properties-value-implicit-001.html | border-top-width length-em(em) / events - assert_equals: Expected TransitionEnd events triggered on .transition expected "border-top-width:0.5s" but got ""
Component: web-platform-tests → CSS Parsing and Computation
Product: Testing → Core
Version: Version 3 → Trunk
(In reply to Brian Birtles (:birtles, traveling 29 June - 6 July) from comment #10)
> Huh, I thought we disabled all the transitions tests. I guess not. I'll get
> around to rewriting this test in the next month or two I expect (when I
> rewrite *all* the remaining transitions tests).

Brian, do you have an update for us? 6 months passed by meanwhile. :) Thanks.
Flags: needinfo?(bbirtles)
Depends on: 1472172
No. I still intend to rewrite all the transition tests, but it is fairly low on my list of things to do. It would be fine to disable this until then.
Flags: needinfo?(bbirtles)
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.