Per discussion with Dave, we found that css2.1/t090204-display-change-01-b-ao.html is trying to attach load event listener on document instead of window. We should fix that.
This should be reported on public-css-testsuite@w3.org
Created attachment 106618 [details] fixes the test
(In reply to comment #1) > This should be reported on public-css-testsuite@w3.org That's a good point but I'm not subscribed to that mailing list. Could you or someone else familiar with the matter do that on behalf of me?
Comment on attachment 106618 [details] fixes the test r=me
http://www.w3.org/Bugs/Public/ might help :)
Committed r94696: <http://trac.webkit.org/changeset/94696>
(In reply to comment #5) > http://www.w3.org/Bugs/Public/ might help :) Ah, thanks for the pointer. Filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=14064.
Is it no longer the policy to not modify third-party test suites?
I don't see this test in the the current 2.1 test suite: <http://test.csswg.org/suites/css2.1/20110111/html4/toc.html>
(In reply to comment #8) > Is it no longer the policy to not modify third-party test suites? (In reply to comment #9) > I don't see this test in the the current 2.1 test suite: <http://test.csswg.org/suites/css2.1/20110111/html4/toc.html> I am confused now. What is the correct solution then?
We have an old snapshot of the CSS 2.1 suite. We should import the (huge) current one.
(In reply to comment #10) > (In reply to comment #8) > > Is it no longer the policy to not modify third-party test suites? > > (In reply to comment #9) > > I don't see this test in the the current 2.1 test suite: <http://test.csswg.org/suites/css2.1/20110111/html4/toc.html> > > I am confused now. What is the correct solution then? Unless the bug in the test is causing it to (a) crash or hang the test tool or (b) produce inconsistent results, I think the change should be reverted. If the bug in the test is causing it to be invalid or not cover what it is supposed to test, then the solution is to add a valid test somewhere outside the css2.1 directory.
(In reply to comment #12) > (In reply to comment #10) > > (In reply to comment #8) > > > Is it no longer the policy to not modify third-party test suites? > > > > (In reply to comment #9) > > > I don't see this test in the the current 2.1 test suite: <http://test.csswg.org/suites/css2.1/20110111/html4/toc.html> > > > > I am confused now. What is the correct solution then? > > Unless the bug in the test is causing it to (a) crash or hang the test tool or (b) produce inconsistent results, I think the change should be reverted. If the bug in the test is causing it to be invalid or not cover what it is supposed to test, then the solution is to add a valid test somewhere outside the css2.1 directory. That’s right.
(In reply to comment #11) > We have an old snapshot of the CSS 2.1 suite. We should import the (huge) current one. Also a great idea.
(In reply to comment #12) > Unless the bug in the test is causing it to (a) crash or hang the test tool or (b) produce inconsistent results, I think the change should be reverted. If the bug in the test is causing it to be invalid or not cover what it is supposed to test, then the solution is to add a valid test somewhere outside the css2.1 directory. Okay. Then let's revert this change and add a fixed test in fast/css.
i.e. I'm going to swap css2.1/t090204-display-change-01-b-ao.html and fast/css/line-after-floating-div.html.
Created attachment 106667 [details] swaps line-after-floating-div.html and t090204-display-change-01-b-ao.html
Reopen the bug.
Created attachment 106668 [details] Fixed the change log
Thanks for the timely review. Will land now.
Committed r94724: <http://trac.webkit.org/changeset/94724>