|Summary:||LayoutTests/transitions/zero-duration-with-non-zero-delay-end.html is flaky|
|Product:||WebKit||Reporter:||Julie Parent <jparent>|
|Component:||Tools / Tests||Assignee:||Julie Parent <jparent>|
|Severity:||Normal||CC:||cmarrin, dimich, eric|
|Version:||528+ (Nightly build)|
Description Julie Parent 2009-11-18 14:04:01 PST
This test uses setTimeouts to measure the number of active transitions at given times. Since this is a 0 duration transition, it is not possible to use transitionEnd events instead (they are not dispatched), but it should be possible to improve reliability of this test by increasing the final timeout. This test fails ~50% of the time on the Chromium bots.
Comment 1 Julie Parent 2009-11-18 14:14:35 PST
Created attachment 43459 [details] Increase timeout from 500ms to 750ms (transition delay is set to 500ms).
Comment 2 Eric Seidel (no email) 2009-11-18 15:31:07 PST
Comment on attachment 43459 [details] Increase timeout from 500ms to 750ms (transition delay is set to 500ms). There is no event we could listen for instead? Ideally we would add a comment next to the setTimeout about why it's required. Especially since we're trying to kill all the other setTimeouts, seems we should document why this one is OK.
Comment 3 Eric Seidel (no email) 2009-11-18 15:31:30 PST
Also, you probably want to cc Chris on transition bugs. :)
Comment 4 Julie Parent 2009-11-18 15:33:01 PST
Since the transitionEnd event isn't dispatched, I'm not sure what other event we could use. I'll add a comment with the setTimeout. What is Chris's email address?
Comment 5 Eric Seidel (no email) 2009-11-18 15:42:38 PST
cmarrin. I CC'd him already. Thanks.