503681 - browser_bug134911.js times out intermittently

Status: VERIFIED FIXED

(Core :: Layout: Form Controls, defect)

Description

[(no longer active)]

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1247329235.1247335617.25038.gz
Linux mozilla-central unit test on 2009/07/11 09:20:35

The previous run was from a different changeset <http://hg.mozilla.org/mozilla-central/rev/20be4e9999cf> but the only changesets in the range is the backout of  a font change, which can't be relevant here; so this should be intermittent.

Comment 1

[Boris Zbarsky [:bzbarsky]]

The test uses a setTimeout (twice, acually).  Why?  It looks to me like if either timeout is too short (which it could easily be on a loaded machine) the test will throw and never reach finish()...  Why are we not observing events instead of relying on timeouts here?

Comment 2

[Simon Montagu :smontagu]

(In reply to comment #1)
> The test uses a setTimeout (twice, acually).  Why?

Because I knew no better at the time?

Fixed to use event listeners both in browser_bug134911.js and the very similar browser_bug92473.js, pushed in http://hg.mozilla.org/mozilla-central/rev/b9012a18d2c6

A possible question mark is that I use the load event both for the original page load and for changing charset. I'm not sure if there is any other event that would work instead, and I'm not sure if there is a guarantee that changing charset will fire a load event, e.g. if we fix bug 61363.

Comment 3

[Simon Montagu :smontagu]

http://hg.mozilla.org/releases/mozilla-1.9.1/rev/d08559185681

Comment 4

[Simon Montagu :smontagu]

Checking in docshell/test/browser/browser_bug134911.js;
/cvsroot/mozilla/docshell/test/browser/browser_bug134911.js,v  <--  browser_bug134911.js
new revision: 1.2; previous revision: 1.1
done
Checking in docshell/test/browser/browser_bug92473.js;
/cvsroot/mozilla/docshell/test/browser/browser_bug92473.js,v  <--  browser_bug92473.js
new revision: 1.2; previous revision: 1.1

Comment 5

[Boris Zbarsky [:bzbarsky]]

> and I'm not sure if there is a guarantee that changing charset will fire a load
> event, e.g. if we fix bug 61363.

If we change that, this test will just fail reliably, right?  We can worry about it then.

Comment 6

[Mike Beltzner [:beltzner, not reading bugmail]]

Simon: can you verify that this is resolved for mozilla-1.9.1?

Comment 7

[Simon Montagu :smontagu]

I verify that there are no SetTimeouts in http://mxr.mozilla.org/mozilla1.9.1/source/docshell/test/browser/browser_bug134911.js. Not sure what other verification is possible in this case.