[Webkit-unassigned] [Bug 93755] ShadowRoot insertion point change aborts css transition

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=93755





--- Comment #18 from Takashi Sakamoto <tasak at google.com>  2012-08-24 00:09:37 PST ---
(From update of attachment 160103)
View in context: https://bugs.webkit.org/attachment.cgi?id=160103&action=review

Thank you for reviewing.

>> Source/WebCore/ChangeLog:10
>> +        shadow host's style clear. So when distribution is changed, no css
> 
> The meaning of "clear" here is unclear :)
> 
> Do you mean it clears the style of host's style?

Yes. I would like to say so. I fixed the description.

>> Source/WebCore/dom/ElementShadow.cpp:92
>> +    // a renderer in Element::recalcStyle.
> 
> This comment probably belongs in a bug. Reference bug here and enjoy using a much shorter explanation :)

Sure. I created a new bug 94905.

>>> Source/WebCore/dom/ElementShadow.cpp:95
>>> +        shadowHost->lazyAttach();
>> 
>> Didn't we have something called re-attach on Node?
> 
> That isn't lazy. But I agree that it might be nice to have lazyReattach().

I have already tried shadowHost->reattach() here and I found that the reattach breaks the following layout tests:
fast/dom/shadow/input-with-validation.html - crash
fast/dom/shadow/shadowdom-for-progress-dynamic.html - image diff
fast/dom/shadow/shadowdom-for-progress-multiple.html - image diff    
fast/dom/shadow/shadowdom-for-progress-with-style.html - image diff
fast/dom/shadow/shadowdom-for-progress.html - image diff

At least, the oldest shadow root will be reattached too often.

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