[webkit-reviews] review denied: [Bug 48297] Fix LayoutTests/canvas/philip/tests/2d.composite.uncovered.image.destination-atop.html : [Attachment 100262] Proposed patch with drawImage() changes

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Fri Jul 22 15:42:25 PDT 2011


James Robinson <jamesr at chromium.org> has denied Mustafizur Rahaman (rahaman)
<mustaf.here at gmail.com>'s request for review:
Bug 48297: Fix
LayoutTests/canvas/philip/tests/2d.composite.uncovered.image.destination-atop.h
tml
https://bugs.webkit.org/show_bug.cgi?id=48297

Attachment 100262: Proposed patch with drawImage() changes
https://bugs.webkit.org/attachment.cgi?id=100262&action=review

------- Additional Comments from James Robinson <jamesr at chromium.org>
This will produce weird results if the destination rectangle is not
pixel-aligned.	I think in general the displayTransparencyElsewhere() approach
is not going to work perfectly.

I think we should investigate using an intermediate surface for these composite
operations and see what the performance looks like.  This is what firefox does
and is closer to the rendering model in the spec.  If it's not too expensive,
it'll be a lot simpler and correct in all of these tricky cases.


More information about the webkit-reviews mailing list