This happens on Firefox 3.5: Linux and Windows, but not MacOSX. { http://tinderbox.mozilla.org/showlog.cgi?log=Firefox3.5-Unittest/1245465905.1245467120.17203.gz Linux mozilla-1.9.1 test everythingelse on 2009/06/19 19:45:05 http://tinderbox.mozilla.org/showlog.cgi?log=Firefox3.5-Unittest/1245463043.1245464514.12573.gz WINNT 5.2 mozilla-1.9.1 test everythingelse on 2009/06/19 18:57:23 Running chrome://mochikit/content/browser/browser/base/content/test/browser_tabs_owner.js... TEST-UNEXPECTED-FAIL | ... | 4 tabs are open - Got 5, expected 4 TEST-UNEXPECTED-FAIL | ... | newly opened tab is selected - Got [object XULElement], expected [object XULElement] TEST-PASS | ... | owner is selected TEST-PASS | ... | selecting a different tab clears the owner relation TEST-PASS | ... | owner relatitionship persists when tab is moved } (Does not happen on Firefox 3.6.)

The failure is really in one of the earlier running tests... one of them doesn't cleanup the tab strip when it's done, either due to sloppiness or because that test has real issues.

And that other test is probably the one bug 496534 is about...

(In reply to comment #1) > one of them doesn't cleanup the tab strip when it's done Arguely, browser_tabs_owner.js could detect/warn/handle such an initial situation. (In reply to comment #2) > And that other test is probably the one bug 496534 is about... Agreed, very probably.

verified1.9.1.2, per tinderboxes. You're welcome to file a follow-up bug if you care to make this test more bulletproof.

As per comment 4, verified1.9.1

Nothing broken on trunk or 1.9.2 => verified.