496087 - setTimeout stops working in small two frame test case.

Status: NEW

(Core :: DOM: Core & HTML, defect, P5)

Description

[Ronan Klyne]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.9.0.10) Gecko/2009042316 Firefox/3.0.10 (.NET CLR 3.5.30729)
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.9.0.10) Gecko/2009042316 Firefox/3.0.10 (.NET CLR 3.5.30729)

I have imtermittently observed setTimeout ceasing to work in a particular tab after browsing seemingly random pages. Only closing the tab and using a different one would make it work again.

I've managed to reduce the problem to four small HTML files, which are attached.

This is not a problem in Firefox 2, but has been seen in 3.0.1 and 3.0.10. I've only tried this on Windows XP.

Reproducible: Always

Steps to Reproduce:
1. Get these attached files: ft1.html, ft2.html, ft3.html, ft4.html. Put them in the same directory.
2. Open ft1.html
3. Click the link "Test me"
4. The page will try to navigate one of the two frames, which will in turn navigate away from the frameset.
5. The final page has a message "Test failed", which should be changed to "Test passed" by a setTimeout call.
Actual Results:  
Message reads "Test failed", and the test cannot be re-run in the same tab.
After going to any other web site using this tab, setTimeout still does not work.

Expected Results:  
Message reads "Test passed".

Comment 1

[Ronan Klyne]

Attachment: Test case part 1/4 - frameset

Comment 2

[Ronan Klyne]

Attachment: Test case part 2/4 - frame 1

Comment 3

[Ronan Klyne]

Attachment: Test case part 3/4 - frame 2a

Comment 4

[Ronan Klyne]

Attachment: Test case part 4/4 - frame 2b

Comment 5

[Ria Klaassen (not reading all bugmail)]

Confirmed with Latest trunk and Firefox 3.0 on Windows XP, but in older builds the results are inconsistent. Sometimes I get the combination of a Boo! alert and a Test Passed result.

Comment 6

[Ronan Klyne]

(In reply to comment #5)
> Confirmed with Latest trunk and Firefox 3.0 on Windows XP, but in older builds
> the results are inconsistent. Sometimes I get the combination of a Boo! alert
> and a Test Passed result.

I was perhaps slightly unclear - when I said that this had been seen in 3.0.1, I did not mean that I'd run this test case on 3.0.1, but rather that the problem had been observed whilst using 3.0.1.
My testing (and test case writing) has been on 3.0.10 only - I don't have a 3.0.1 setup to hand...

Comment 7

[Townsend]

Confirmed in Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2) Gecko/20100115 Firefox/3.6.

However, the test case passes if Firebug (1.5.0) is installed and enabled (though not necessarily activated).

Comment 8

[a.hendriks]

Observing similar problems in 3.6.12 - timeouts in the tabs stop working. It's easily verified by entering 

window.setTimeout('alert("hoi")', 100)

in the firebug java console for the tab - it won't fire in the specific tab where all timeouts fail, but it works fine in the other tabs. (so firebug isn't working around the problem anymore)

Haven't yet found a testcase to reproduce what we're seeing.

Comment 9

[indeepestdarkest]

Does anybody know if the original attachments with this bugzilla can still be made available? I am trying to reproduce the problem to see if it related to the problem I am experiencing with my own code (which is more difficult to reproduce reliably but is definitely related to setTimeout() ceasing to work. Firefox 3.6.13).

Comment 10

[Ronan Klyne]

I dug out the original files and made them available here: https://github.com/rklyne/firefox-bug-496087

I don't know why they went missing.

Tested on OSX 37.0.2 and found to be working.

Comment 11

[Firefox Bug Husbandry Bot]

https://bugzilla.mozilla.org/show_bug.cgi?id=1472046

Move all DOM bugs that haven’t been updated in more than 3 years and has no one currently assigned to P5.

If you have questions, please contact :mdaly.