Closed Bug 468261 Opened 12 years ago Closed 12 years ago

[SeaMonkey] test_bug457672.html fails; (which can make test_draggableprop.html fail too)

Categories

(Core :: DOM: Events, defect)

defect
Not set
normal

Tracking

()

VERIFIED FIXED
mozilla1.9.1b3

People

(Reporter: sgautherie, Assigned: smaug)

References

Details

(Keywords: intermittent-failure, verified1.9.1)

Attachments

(1 file)

See
{
http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1228585326.1228590231.12701.gz
Linux comm-central dep unit test on 2008/12/06 09:42:06

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1228586420.1228592740.18027.gz
MacOSX 10.4 comm-central dep unit test on 2008/12/06 10:00:20

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1228584850.1228590048.12408.gz
Win2k3 comm-central dep unit test on 2008/12/06 09:34:10

*** 19350 ERROR TEST-UNEXPECTED-FAIL | /tests/content/events/test/test_bug457672.html | Window should have got blur event when opening a new tab! - got 0, expected 1
}

***

http://hg.mozilla.org/releases/mozilla-1.9.1/rev/ff614c159864
disable the test for bug 457672, because SM can't handle it.
Flags: blocking1.9.1?
I'm not sure we should block 1.9.1 on this; the content/ side of the feature is sound, SM just doesn't utilize it [by default?].
Iirc, smaug said
{
the failing testcase needs FF "open-new-window" behavior

hows a ifndef MOZ_SUITE for now? [or ifdef for firefox itself]
}

*****

Interestingly,
http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1228589618.1228594125.22473.gz
Win2k3 comm-central dep unit test on 2008/12/06 10:53:38
which doesn't yet have the backout,
is green.

"Intermittent" ?? (But Linux is still orange...)
(In reply to comment #1)

The blocking is to get the test reenabled (on the FF 3.1 tree at least).
I'm just compiling SM so that I can check where/why the test is failing.
Oh, and ftr, the regression timeframe:
{
good: rev:8f9c89baa8d1
bad :     4211efe014b4
This is only bug 463176, which is TB specific.

good: moz:9a3768626bd7
bad :     c2d8e86e461d
}
SM doesn't open new window in a tab by default, and the test is
sort-of testing that, so not a big surprise there is a failure.
I'll fix the test...
(In reply to comment #2)
> hows a ifndef MOZ_SUITE for now? [or ifdef for firefox itself]

{
<wg9s>	Callek: couldn't the testcase chack for the user agent and just pass if it is not firefox rather than putting more conditionals in the build?
}
Yes, it could report "todo" until fully "fixed".

> *****
> 
> Interestingly,
> http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1228589618.1228594125.22473.gz
> Win2k3 comm-central dep unit test on 2008/12/06 10:53:38
> which doesn't yet have the backout,
> is green.
> 
> "Intermittent" ?? (But Linux is still orange...)

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1228593694.1228598457.678.gz
Win2k3 comm-central dep unit test on 2008/12/06 12:01:34
fails again, and still doesn't have the backout.

Then, yes, "intermittent" :-/
Blocks: 438871
Summary: [SeaMonkey] test_bug457672.html fails → [SeaMonkey] test_bug457672.html fails; (which makes test_draggableprop.html fail too)
Note that the passing machine had less tests run for some reason (I decided not to load the entire log just yet)  So I suspect for _some_ reason we just never ran the test.
Comment on attachment 351717 [details] [diff] [review]
this should fix it
[Checkin: Comment 11 & 12]

looks good from a SM side -- I'm not a content peer though
Attachment #351717 - Flags: review+
I landed the patch to 1.9.1. If SM tboxes stay green, I'll land it also
to trunk.
Fixed also on trunk.
Assignee: nobody → Olli.Pettay
Status: NEW → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED
Keywords: fixed1.9.1
(In reply to comment #2)
> http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1228589618.1228594125.22473.gz
> is green.

{
*** 19349 INFO Running /tests/content/events/test/test_bug457672.html...
search "file" not found - skipping.
Document http://localhost:8888/tests/content/events/test/bug457672.html loaded successfully
*** 19350 INFO TEST-PASS | /tests/content/events/test/test_bug457672.html | Window should have got blur event when opening a new tab!
}

(In reply to comment #8)
> Note that the passing machine had less tests run for some reason

Not sure what you are referring to:
the one green (Windows) build did have usual test numbers and did (successfully) run this test ... surprisingly :-|

*****

verified1.9.1 as the three boxes are now green (wrt this bug) :-)
Flags: blocking1.9.1? → in-testsuite+
Summary: [SeaMonkey] test_bug457672.html fails; (which makes test_draggableprop.html fail too) → [SeaMonkey] test_bug457672.html fails; (which can make test_draggableprop.html fail too)
Target Milestone: --- → mozilla1.9.1b3
Comment on attachment 351717 [details] [diff] [review]
this should fix it
[Checkin: Comment 11 & 12]

>+    try {
>+      prefs.clearUserPref("open_newwindow");
>+    } catch (ex) {}

Shouldn't you error somehow in the catch case ?
(to warn following tests which could potentially suffer from this)

Moreover, shouldn't you rather getIntPref() and restore that ?
(in case the testing profile would set it to some non default value ?)
Attachment #351717 - Attachment description: this should fix it → this should fix it [Checkin: Comment 11 & 12]
Keywords: verified1.9.1
sorry for the inconvenience.  i meant to detect any bugs before the branch merge that were still RESO fixed, but had a verified1.9.1 keyword next to them.
Keywords: verified1.9.1
(In reply to comment #13)
> (In reply to comment #2)
> > http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1228589618.1228594125.22473.gz
> > is green.
> 
> {
> *** 19349 INFO Running /tests/content/events/test/test_bug457672.html...
> search "file" not found - skipping.
> Document http://localhost:8888/tests/content/events/test/bug457672.html loaded
> successfully
> *** 19350 INFO TEST-PASS | /tests/content/events/test/test_bug457672.html |
> Window should have got blur event when opening a new tab!
> }
> 
> (In reply to comment #8)
> > Note that the passing machine had less tests run for some reason
> 
> Not sure what you are referring to:
> the one green (Windows) build did have usual test numbers and did
> (successfully) run this test ... surprisingly :-|
> 
> *****
> 
> verified1.9.1 as the three boxes are now green (wrt this bug) :-)

Hi serge, can you also check to see if this is fixed on trunk?  im unsure how to test this.  Thanks.
(In reply to comment #14)
> (to warn following tests which could potentially suffer from this)

I understand it is rather up to the other tests to (re)set whatever settings they rely on.

(In reply to comment #16)
> Hi serge, can you also check to see if this is fixed on trunk?  im unsure how
> to test this.  Thanks.

SeaMonkey does not have m-c boxes atm.

[Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.9.2a1pre) Gecko/20090110 SeaMonkey/2.0a3pre] (experimental/_m-c_, home, optim default) (W2Ksp4)
(http://hg.mozilla.org/mozilla-central/rev/6acaaa957e0a
 +http://hg.mozilla.org/comm-central/rev/865f907bb16b)

{
19402 INFO Running /tests/content/events/test/test_bug457672.html...
19403 INFO TEST-PASS | /tests/content/events/test/test_bug457672.html | Window should have got blur event when opening a new tab!
}

V.Fixed
Status: RESOLVED → VERIFIED
Whiteboard: [orange]
Whiteboard: [orange]
You need to log in before you can comment on or make changes to this bug.