Mark Birbeck wrote:
Hi Ivan,- HTML4 is not in XML, right? I just wonder whether the very DOM oriented processing steps would be appropriate. What seems to be relatively straightforward is an XHTML1.0+RDFa, the question is whether doing the corresponding DTD would be that easy.I think XHTML+RDFa is easier to define...that's true. And in terms of processing RDFa on a server, it's also slightly easier to implement than HTML+RDFa, because you can use XML tools. But a consequence of XHTML not being 'standard' across browsers, is that there is no difference between implementing an XHTML+RDFa and an HTML+RDFa parser. This is because, although an enormous number of documents are created as XHTML on the server, they are delivered to the client as "text/html", which will switch the browser into HTML mode. So anyone writing a client-side parser for XHTML+RDFa is almost certainly going to have to write it so that it works in 'HTML mode'.
Well... my implementation starts with an XML parser, and does everything on the resulting DOM tree. Ie, it will not work on an HTML+RDFa version unless it is proper XML...
:-( Ivan
(Which is incidentally why I used the DOM idea to define the processing, because it works with both HTML DOMs and XHTML DOMs.) I'm not saying anything here about the original question -- I think we'll get to the HTML+RDFa side when we're ready. I'm merely pointing out that technically it's a no-brainer, because we took care to make sure that this was so. As Shane says, the only piece missing to create an HTML+RDFa Syntax is a way to set the prefix mappings. (And even then, that's only because it seems odd to use the @xmlns mechanism in HTML.)- The real issue is, however, HTML5. And that is only where a crystal ball would help:-)Indeed. :) Regards, Mark
-- Ivan Herman, W3C Semantic Web Activity Lead Home: http://www.w3.org/People/Ivan/ PGP Key: http://www.ivan-herman.net/pgpkey.html FOAF: http://www.ivan-herman.net/foaf.rdf
smime.p7s
Description: S/MIME Cryptographic Signature
