The problem I ran into with box-shadow is that the shadow cannot show up anywhere within the contents of the box. It has to be clipped out. Maybe there's a clever way to get the same result that I'm not thinking of though. Note that CG draws shadows in response to other drawing operations (in this case a fill). As far as I can tell with CG, I can't just draw a shadow in a vacuum. It has to be implicitly cast by some other drawing operation. So I'm basically doing a fill, which casts the shadow, and then ensuring that the entire fill is clipped out so that it isn't seen.

dave

On Jan 9, 2007, at 7:04 AM, Mike Reed wrote:

Do we need to use clips for round-rects? In my implementation these
are always hard-edged (no antialiasing). Is that true for Qt as well?
CG? I would prefer being asked to draw round-rects (filled or stroked)
directly with no clipping...

On 1/8/07, David Hyatt <[EMAIL PROTECTED]> wrote:
I just landed support for the CSS3 box-shadow property.  This
entailed adding a few new methods to GraphicsContext.  I stubbed
these methods out on Qt but did not implement them.

They are:

void clipOut(const IntRect&);
        Excludes the specified rect from the clipping region.

void clipOutEllipseInRect(const IntRect&);
        Same, but for an ellipse drawn inside the specified rect.

void fillRoundedRect(const IntRect&, const IntSize& topLeft, const
IntSize& topRight, const IntSize& bottomLeft, const IntSize&
bottomRight, const Color&);
         Fills a rounded rect (with elliptical curvature in the four
corners) with the specified color.

Note that the fillRoundedRect method arguably should be cross-
platform, relying on primitives instead.   This is an area for
improvement, if someone wants to rework it to use the cross-platform
Path class etc.

dave
([EMAIL PROTECTED])

_______________________________________________
webkit-dev mailing list
[email protected]
http://lists.webkit.org/mailman/listinfo/webkit-dev


_______________________________________________
webkit-dev mailing list
[email protected]
http://lists.webkit.org/mailman/listinfo/webkit-dev

Reply via email to