[Webkit-unassigned] [Bug 63899] The content of replaced element 'img' is trimmed to the content edge curve of a rounded corner.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=63899


Benjamin Poulain <benjamin at webkit.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |benjamin at webkit.org




--- Comment #14 from Benjamin Poulain <benjamin at webkit.org>  2011-07-18 06:04:30 PST ---
That is an interesting case, but I am wondering if the inner border radius will not give us wrong pixels values on some pixels due to:
-antialiazed clipping (the background color could blend with some pixels values)
-the "rounded" algorithm differing for clipping the inner circle, and the innermost pixels of the border. (I remember issues on X11 with the algorithm for fill not matching the algorithm for border so the 2 can never match pixel to pixel).

My first idea seeing the test case was:
-the clip should be in the middle of the border (0.5 pixels for a 1 pixel border) so that the antializing of the clip does not gives us bad value. If there is padding and the padding is supposed to clip, the clip in that case would be the inner border - padding.
-the border should be on top of the rendering and not done before (I don't remember what the spec says about that in the rendering phases...)

I see the original test include padding. I think a second test with padding is necessary here since that is an interesting variation on the rendering of 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.