- From: Ben Adida <ben@adida.net>
- Date: Thu, 14 Dec 2006 00:47:06 -0500
- To: Dan Connolly <connolly@w3.org>
- CC: Fabien Gandon <Fabien.Gandon@sophia.inria.fr>, public-grddl-wg@w3.org
Dan Connolly wrote: > Yes, that would be better, I suppose, though the HTML 4 spec is reasonably > clear that when a link relationship is dominated by a profile, the profile > is what grounds the link relationship in URI space, and the link > relationship > isn't defined by HTML. Yes, in fact, I was asking about the GRDDL profile thing because, as you and I discussed a while ago, this really *should* be performed by what I've called "hGRDDL"... In other words, the PROFILE on the HEAD defines how to modify the DOM for certain reserved words, so that rel="transformation" becomes rel="grddl:transformation". But of course, hGRDDL is out of scope for GRDDL, so... :) we're going to be a bit stuck, I fear. I'll keep thinking about this one. If you've got some ideas on this, let me know. What I'd like to see is a consistent RDFa story, where this really does yield an HTML+RDFa representation of the triple: <> grddl:transformation <RDFa2RDFXML.xsl> Because that's what the document really says! -Ben
Received on Thursday, 14 December 2006 05:47:05 UTC