[Webkit-unassigned] [Bug 64412] didFirstVisuallyNonEmptyLayout dispatched too early

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=64412





--- Comment #5 from Antti Koivisto <koivisto at iki.fi>  2011-07-13 02:57:08 PST ---
> There is nothing in the change log or code that explains why you added a new check of didLayoutWithPendingStylesheets here.

- Wait until stylesheets are loaded.  <--- this one, but i'll expand a bit

> Would like a why comment explaining where this threshold came from. There isn’t even a comment explaining why we made visually non-empty a heuristic rather than a more brittle but well defined mechanism. There should be.

Ok.

> > Source/WebCore/page/FrameView.h:470
> > +    m_visuallyNonEmptyPixelCount += size.width() * size.height();
> 
> Any risk of overflow here?

Yes, but it is unlikely to happen and even more unlikely to cause problems.

> Does this do the right thing for incrementally loading images? Seems like it will only do the pixel count thing once, perhaps with a smaller size than the actual final size?

It uses the flnal size of the image even when loading incrementally. I don't think anything fancier is worth the effort.

> > +    view()->frameView()->incrementVisuallyNonEmptyCharacterCount(m_text.length());
> 
> Not worried about cases where there is nothing but a lot of whitespace?

I don't want to take the (small) performance hit unless there is a proof that this is a real world issue.

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