- From: Ian Hickson <ian@hixie.ch>
- Date: Thu, 3 Jul 2003 11:07:32 -0700 (PDT)
- To: James Craig <work@cookiecrook.com>
- Cc: "www-style@w3.org" <www-style@w3.org>
On Thu, 3 Jul 2003, James Craig wrote: > > XML example: > > <myImage mySource="img/foo.gif" myAlt="foo" /> > > myImage { > display: image; > text-alternative: attr(myAlt); > source: attr(mySource); > } With the CSS3 Generated Content draft, that would be: myImage { content: attr(mySource, url), attr(myAlt); } > I thought of an issues that might arise. For example, should the source > path be relative to the document or stylesheet? Since it's from the > document, I would say document, but URL values from a stylesheet are > usually evaluated to be relative to the stylesheet. The form "attr(mySource, url)" will be resolved relative to the document. > Should there be a mime-type spec in the CSS for these images? > > myImage[mySource$=".gif"] { > mime-type: "image/gif"; > } > myImage[mySource$=".png"] { > mime-type: "image/png"; > } I don't understand what this means. > If this idea pans out, what kind of proposals might there be for > displaying other kind of things that would typicially be held in HTML > <object> elements? Why do they need any special casing? video { content: url(movie.mpeg); } flash { content: url(game.swf); } test { content: url(test.svg), url(test.html), url(test.mathml), url(test.smil); } Of course there's nothing stopping UAs right now from implementing: body { background: url(test.swf); } Cheers, -- Ian Hickson )\._.,--....,'``. fL "meow" /, _.. \ _\ ;`._ ,. http://index.hixie.ch/ `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 3 July 2003 14:07:15 UTC