[Webkit-unassigned] [Bug 16853] http/tests/loading has different results when run with run-webkit-tests -1

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Sun Jan 13 12:36:37 PST 2008


http://bugs.webkit.org/show_bug.cgi?id=16853





------- Comment #9 from freyther at handhelds.org  2008-01-13 12:36 PDT -------
(In reply to comment #8)
> This problem is not limited to http/tests/loading and webarchive/loading - I've
> just fixed something similar in fast/loader
> <http://trac.webkit.org/projects/webkit/changeset/29340>. I think fast/dom may
> have some tricky loading cases, too.

You are wrong, at least my problem is limited to the dumping of
FrameLoaderClient callbacks and this applies only to paths containing
"loading/", which applies only to the above two directories and tests calling
layoutTestController.dumpFrameLoadCallbacks (which is one test in a loading/
directory).

What I didn't do was to compare the results of "run-webkit-tests -1" to
"run-webkit-tests", I only compared http/tests/loading because I want to add a
test case there and refuse to pick a file name that will hide this issue (which
will break on filesystems giving a listing in random order anyway, e.g.
reiser4).


> Indeed. Perhaps it would be possible to disable delegate call logging when we
> no longer need it?

I sincerly believe that my proposed fix is the proper one. It doesn't
artificially change the tests, it keeps them simple and clear, and is removing
dead code from DRT. The downside is that we spend some more time on these 7
tests.

On the other hand we can try to change the tests and DRT. This will make the
tests itself more complex, e.g. adding a tear down phase, will add more
complexity to DRT. Which includes one command to make the DRT blind. DRT has to
wait until the tear down was completed, has to verify that all subframes are
gone and this has to work for Mac,Win,Gtk+ and Qt. I don't think the added
complexity to add a test-case is helping us in anyway.


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



More information about the webkit-unassigned mailing list