[Webkit-unassigned] [Bug 158019] Fix ResourceTiming XHR and initiator flakiness

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=158019

--- Comment #15 from Yoav Weiss <yoav at yoav.ws> ---
(In reply to comment #14)
> Comment on attachment 279645 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=279645&action=review
> 
> Would it work with a timeout of 0?

I believe it would.

> I think a better approach would be to
> create the xhr after you're done with the first runTest call to make sure it
> will never start before the first one.
> 
> > LayoutTests/http/tests/performance/performance-resource-timing-xhr-single-entry.html:35
> > +            setTimeout(runTest, 10);
> 
> Why do you call runTest twice but only expect it to be run once?

I copied this mechanism from another test where the XHR call is racing with window's onload event (as XHR is not blocking onload). So, I wanted the test to run once at the later one of those events, but not inside XHR's onload event, as the entries are not yet added at this point. It's true that for the xhr-single-entry I could just eliminate the window.onload part altogether.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20160524/e604daac/attachment.html>