Bug 33272 – 'abort' event still fired as a progress event

• Status: RESOLVED
• Resolution: FIXED
• Component: Media

[Eric Carlson]

https://bugs.webkit.org/show_bug.cgi?id=30513 should have changed all events fired by the HTMLMediaElement to regular events, but 'abort' was missed.

[Eric Carlson]

Created attachment 46002 [details]
Proposed patch

[WebKit Review Bot]

style-queue ran check-webkit-style on attachment 46002 [details] without any errors.

[Darin Adler]

Comment on attachment 46002 [details]
Proposed patch

Is there some test that covers this? Could you make a test that does?

I'm tempted to set review+ but not sure if I should without a test covering the change.

[Eric Seidel (no email)]

Comment on attachment 46002 [details]
Proposed patch

Can we test for lack of an abort event? Marking cq- since this couldn't be committed as-is due to the second OOPS even if you had wanted to cq it (which I doubt).

[Eric Carlson]

media/video-error-abort.html *used* to test the abort event, but it became timing dependent when we stopped firing media events synchronously [1] so it was added to the skipped lists [2]. I will update it for the current behavior, have it to check the abort event, and move it to http/tests/media so it can use the slow loading cgi and actually work reliably.

[1]  https://bugs.webkit.org/show_bug.cgi?id=24588
[2] rdar://6710625

[Eric Carlson]

Created attachment 46022 [details]
Proposed patch

With layout test.

[WebKit Review Bot]

style-queue ran check-webkit-style on attachment 46022 [details] without any errors.

[Maciej Stachowiak]

Comment on attachment 46022 [details]
Proposed patch

r=me

[Eric Carlson]

http://trac.webkit.org/changeset/52923

[Eric Seidel (no email)]

This broke the Leopard Release bot:
http://build.webkit.org/results/Leopard%20Intel%20Release%20(Tests)/r52923%20(9069)/http/tests/media/video-error-abort-pretty-diff.html

[Eric Seidel (no email)]

This also broke the Gtk and Qt builders.  I assume because they had skipped this test previously.

[Eric Seidel (no email)]

Looks like it broken windows too:
http://build.webkit.org/results/Windows%20Release%20(Tests)/r52923%20(7673)/results.html
http://build.webkit.org/results/Windows%20Debug%20(Tests)/r52923%20(8233)/results.html

[Eric Carlson]

The svn:executable flag wasn't set on the cgi, setting it fixed Windows and Leopard Release (http://trac.webkit.org/changeset/52935). 

I forgot to add the new test to the GTK Skipped list. Eric lied about me breaking the Qt build (this time).

[Eric Seidel (no email)]

Thank you for the quick fix.  I am a liar. :(