[Webkit-unassigned] [Bug 51354] Don't block rendering and script execution on deferred stylesheets

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Dec 21 04:31:53 PST 2010


https://bugs.webkit.org/show_bug.cgi?id=51354





--- Comment #12 from Antti Koivisto <koivisto at iki.fi>  2010-12-21 04:31:53 PST ---
(In reply to comment #11)
> Apparently, we're breaking scripts that access "unnecessary" stylesheets via CSSOM before a load event fires. I'm ready to accept that level of brokenness, even though debugging will be a huge pain if this breaks a live Web site (clearly, the problem would be intermittent). If you can think of any remedy, that would be great.

Yeah, that is going to happen. Note that we already did this with alternate stylesheets.

> One thing that seemed worth double checking was whether the document pending sheet count was used for anything else. It can get ugly if it is, or if someone reuses it later. Speaking of which, it looks like a comment in Document.h really needs to be updated.

I searched around quite a bit and couldn't find anything. The purpose of this collection is to block rendering (and rendering dependent script execution) until sheets have been loaded. Since non-matching sheets by definition don't participate in the rendering, they shouldn't be there either. 

Load even firing is based on actual resource loads, not this collection.

> > This patch should have no impact on when load events fire.
> 
> Is this covered by existing regression tests? Seems important enough to add a new test if it isn't.

Yes, fast/css/sheet-collection-link.html covers this case

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.



More information about the webkit-unassigned mailing list