Re: RDFa and its relationship to XHTML

Benjamin,

On Jun 8, 2006, at 3:00 PM, Benjamin Nowack wrote:
> It should be compatible with existing HTML documents at least.

We wish we could make this work, but it would cause a bunch of  
important requirements to fail.

In particular, if we limit ourselves to current HTML specs, we don't  
know how to make statements about other URIs and we don't know how to  
build chunks of HTML that are fully self-contained so they can be  
copied-and-pasted. These are incredibly important to a large set of  
users (I represent one of them: Creative Commons). The reason we  
think we can make this work in XHTML1 is because we're using mostly  
attributes, which won't break current browsers (by definition). Sure,  
the XHTML1 compatibility is not done yet, but that's the direction  
we're heading in, and we've got Mark's first stab at it that's quite  
promising.

As a side note, a bit of historical perspective is in order: the  
attempt to build RDF into HTML has been ongoing since 1999. When  
Creative Commons asked W3C in 2002 how it should go about solving  
this problem, the answer was basically "we have no idea." We waited a  
year, and by end of 2003, the answer was still "we have no  
idea." (There are other examples, but I know the CC one best, of  
course.) RDFa is the first solution that actually proposes to embed  
any RDF into HTML, to maintain the context of the triples, to reuse  
the data already in the page, and to be modular enough that you can  
cut and paste chunks of HTML with complete RDFa. RDFa is the *only*  
solution to provide this combination of important features. The HTML  
spec changes are minimal and won't break any current rendering engine.

And again, I want to reemphasize: if you want to use eRDF, go for it.  
If you want to use microformats, go for it. Just make sure to use a  
profile URI in all of these cases. Then, all RDFa tools will be able  
to parse your page, too, so the work we're doing will benefit you, too.

-Ben

Received on Thursday, 8 June 2006 19:21:19 UTC