[Webkit-unassigned] [Bug 19652] [CAIRO] wrong drawing of border-image

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=19652





------- Comment #10 from vbs85 at gmx.de  2008-09-26 12:29 PDT -------
(In reply to comment #9)
Thanks for your reply. It is a great idea. I tested it with many modifications
and spend much time on it, but didn't get it to work.

The problem is, the geometry of the context isn't reset. That means, the
dimension of the context is different on every time.
The second problem is: You don't only move the the image by -tileRect.x(),
-tileRect.y(), you need a part of it, e.g. if you have a 90x90 surface, you
need 5 parts. One would be (30, 0, 30, 30)/(x,y,width,height). I wasn't able to
realise it with push_to_group. One thing why i didn't manage it was that a rect
with the coordinate 0,0 wasn't placed at the top,left of the context, it was
somewhere out of the view area.
I have to admit, that this behavior makes it difficult to understand the sense
of push_to_group.

In summary:
I need a context with a height and width of the surface. Otherwise I'll get
problems with the pattern. It will give more than just this part of a surface.
If it is possible to get the width and height of the current context and the
coordinates of the first pixel in the top/left of the context, you can scale
and transform the created image (pop_to_group).
But I don't believe that it will be more affective than just creating a new
surface with the height and width given by tileRect and draw the part of the
image into it's context.

If you have more suggestions, please post it. It would be great to do it
without an new surface or ImageBuffer and an acceptable complexity.


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