[Webkit-unassigned] [Bug 61636] TestFailures page should have links to file bugs for failing tests

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=61636





--- Comment #5 from Adam Roben (:aroben) <aroben at apple.com>  2011-05-27 09:30:49 PST ---
(From update of attachment 95178)
View in context: https://bugs.webkit.org/attachment.cgi?id=95178&action=review

Thanks for reviewing (unofficially)!

>> Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:192
>> +            if (parsedFailingBuildName.revision - parsedPassingBuildName.revision > 1)
> 
> Should this be >= ? Or are you explicitly making sure that there are two failing revisions so as to try to discard flakey tests? If so, you should probably add a comment to that effect.

This code is building up the string that goes inside the parentheses in "REGRESSION ():". If the test passed in r12344 and failed in r12345, we want to show "REGRESSION (r12345):". If the test passed in r12344 and failed in r12346, and we don't know about r12345, we want to show "REGRESSION (r12344-r12346):". So we only include the passing revision if the failing revision is two or more revisions later.

>> Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:210
>> +            short_desc: 'REGRESSION (' + regressionRangeString + '): ' + failingTests.join(', ') + ' failing on ' + tester.name,
> 
> I don't we can assume that the failing tests are regressions, as they might be new tests that are failing since their introduction. It would be nice if we could have a way of figuring out whether the failing tests are new (but not necessarily something that needs to be done in this patch).

I'll add a FIXME and file a bug.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.