On Tue, 05 May 2009 00:16:50 +0200, ddailey <[email protected]>
wrote:
Maybe I misunderstood (or, more precisely, I am about to state my
probable misunderstanding):
Does this mean for example, that the author could take a chunk of <svg>
code or its equivalent and pass it to canvas.drawImage() (and friends)
and then have a raster version in JavaScript of the <svg> ? If so, then
would that not allow, client-side, things like interrogating pixel
values at point (x,y) within an SVG? That's something I'd like to see. I
understand that if done improperly this can yield security problems
across domains, but it seems in this case we're only talking one domain.
The SVG fragment could still contain a cross-origin reference of some
sorts so it does not help with that. It does allow you to create some SVG
and draw it on a <canvas> without having to first create a data URL or
some such out of it.
--
Anne van Kesteren
http://annevankesteren.nl/
