Right now in head.js's afterAllTabItemsUpdated we call update on each of them, essentially bypassing the TabPriorityQueue. Adding the _updateTestFunc hook to check that they're all updated and then calling update() on them is much more graceful. Rewrite afterAllTabItemsUpdated in terms of this kind of tabPriorityQueue-obeying logic, and making sure all the tests still pass.

bugspam

Comment on attachment 524910 [details] [diff] [review] patch v1 Passed try: http://tbpl.mozilla.org/?tree=MozillaTry&pusher=tim.taubert@gmx.de&rev=f908b24a2c25

Comment on attachment 524910 [details] [diff] [review] patch v1 Looks good to me.

Comment on attachment 524910 [details] [diff] [review] patch v1 Looks fine, but can you explain why those two tests had to be changed?

(In reply to comment #5) > Looks fine, but can you explain why those two tests had to be changed? browser_tabview_expander.js: The problem is that newly created tabs with the url "about:blank#something" have a readyState == "uninitialized". So TabItem.isComplete() will never return true and the update will never occur. That is why I changed the url to "about:home". Alternatively we could also remove the hash. browser_tabview_bug594958.js: In bug 625561 (http://hg.mozilla.org/mozilla-central/rev/20ffcaad9700) these values got changed and I can't really remember why. So what I did was to re-think the whole test and to re-check the given values so that they're now correct and the thumbnails are rendered as expected. I think this will also help us the continue with bug 641188.

(In reply to comment #6) > I think this will also help us the continue with bug 641188. It does. I'll post in bug 641188.

Comment on attachment 528001 [details] [diff] [review] patch v2 Review of attachment 528001 [details] [diff] [review]: Cool, thanks

bugspam

bugspam

bugspam (Fx7 was branched, removing open bugs from Fx7 meta bug, we won't create new meta bugs for upcoming Fx versions)

Is there any concerns why this patch is not checked-in?

Yeah, alas, this patch does not pass try. I tried to find the error source but wasn't very successful so far and didn't have much time to try it further (because that's such a minor issue).

Fixed by bug 659594.