- From: David Dailey <ddailey@zoominternet.net>
- Date: Thu, 10 Nov 2011 09:06:59 -0500
- To: "'SVG public list'" <www-svg@w3.org>
- Message-ID: <001001cc9fb2$036cf890$0a46e9b0$@net>
I suspect that what I'm seeing here is just a bevy of browser bugs, rather than underspecification in the spec, but bear with me. http://granite.sru.edu/~ddailey/svg/feTile.svg If the IRI is local (as it is here) then Firefox (8) and Opera (12) seem to give up. Both seem happy with an external image. Chrome does something thoroughly bizarre, but interesting (note how the second star but not the image is incorporated into the pattern). Safari (windows) doesn't do anything. IE/ASV does rather what I'd hoped. The trouble doesn't seem to go away if the local image in feImage is not transformed though the browsers present yet another set of curiosities. The overall question I was trying to ask was "does feTile really add anything that <pattern> doesn't? And does <feImage> really do anything that <image> doesn't?" I'm relatively certain that the answer is yes, but it's hard to convince myself in the absence of such oddities. Is ASV's behavior the proper behavior here? I don't see anything in the testsuite at http://www.w3.org/Graphics/SVG/Test/20110816/harness/htmlObjectApproved/inde x.html that would address this issue. Incidentally, playing around with the filter's bounding box and primitive units and the like didn't seem to make a difference. Cheers David
Received on Thursday, 10 November 2011 14:07:40 UTC