- From: Timur Mehrvarz <timur.mehrvarz@web.de>
- Date: Tue, 7 Feb 2006 15:57:22 +0100
- To: Bert Bos <bert@w3.org>
- Cc: public-cdf@w3.org
On 7. Feb 2006, at 12:16, Bert Bos wrote: > So do I understand correctly that > > <p>...<img src="image.svg" alt="foo">... > > is meant to have the same effect as > > <p>...<object data="image.svg">foo</object>... > > (assuming the same CSS rules for both)? What makes this a bit difficult, is that both cases are not specified. WICD only talks about the use of html:object for referencing SVG. WICD never mentions html:img and it never requires for both to behave the same. > If so, can that be explicitly > added (and tested)? Because I think it should indeed work like this, > but many browsers currently seem to treat them the same for *some* > types of content but not for others. Let's say, one particular user agent (Safoperfox whatever) would render both samples identically, but only when simple SVG content is being used (static, no animations, no interactivity, no focusable links, etc.). But when the referenced SVG content is interactive or makes use embedded links (which would require the agents focus to travel through the SVG), it could be that both cases are somehow handled differently. What would be gained if WICD would require img and object element to behave exactly the same (wrt SVG)? What advantage would such a duality bring to authors and vendors? Wouldn't it be simpler (=better), if agents would only be required to support one method for referencing SVG content and authors would only use this one method? Timur
Received on Tuesday, 7 February 2006 14:57:27 UTC