- From: Ian Hickson <ian@hixie.ch>
- Date: Sat, 2 Dec 2006 02:40:07 +0000 (UTC)
On Fri, 1 Dec 2006, Robert Sayre wrote: > > I do value the ability to add user-defined tags in a decentralized way, > but the only method I care about is using a container element with an > attribute containing a URI to delineate the user-defined content. Aaaaah! Sorry, I hadn't understood this. So, to put it another way, you want a way to introduce author-specific semantics into your HTML documents? > <http://listserver.dreamhost.com/pipermail/whatwg-whatwg.org/2006-December/008171.html> > > But it is a question, not a request. I don't want to request something > that would be harmful. So, what is the downside of the example in that > earlier email? Well, SVG itself would arguably be bad because it is poor from a semantic standpoint. However, as far as generic author-defined semantics go, that's what the "class" attribute is for. Microformats.org, for example, use the "class" attribute to introduce calendar semantics and the like into HTML. You take the closest fitting HTML element, semantically, and then augment it with your classes. Introducing author-defined markup into HTML would be bad from a semantic and accessibility point of view because UAs would not be able to derive any meaningful information or default presentation from the content, and thus users would not be able to access the data. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Friday, 1 December 2006 18:40:07 UTC