- From: Garret Wilson <garret@globalmentor.com>
- Date: Tue, 9 Apr 2002 11:30:54 -0400 (EDT)
- To: <sean@mysterylights.com>
- Cc: <patrick.stickler@nokia.com>, <www-tag@w3.org>
> Linking to the RDF will remedy all of the problems mentioned above, so > architecturally it is the strongest choice. Augmenting the current > metadata mechanisms is also another choice. But please - don't embed. Sean, I agree with that in general. Just as we've tried to get style information out of the document and into a separate stylesheet, so we should get metadata out of the document and into a separate document---which is the point of XPackage: to use metadata externally to link various resources (HTML, stylesheets, namespaces) together. Really, then, RDDL is just a bunch of metadata telling how namespaces relate to stylesheets, DTDs, etc.---in effect, it's an XPackage document that uses XLink. What Tim and others are saying is that they want a browser to be able to see the metadata, which is why they propose XLink XHTML. This is where the response comes, "OK, then, just put your RDF, be it XPackage or whatever, in an XHTML document." The *best* solution, of course, would be to 1) use RDF to encode RDDL-like information, and 2) get browsers to automatically display RDF, just like they display HTML. (Displaying RDF is *much* easier than rendering HTML, anyway.) The information would be encoded in a semantically robust manner, and everyone would be able to see it. Perhaps putting RDF inside HTML is really just a hack that will allow us to wait until major browsers allow RDF viewing. I agree---in a perfect world, the metadata should go outside the document. Garret P.S. Maybe IE and Netscape could just go ahead and support RDF browsing, and we would all be happy.
Received on Tuesday, 9 April 2002 14:52:07 UTC