[uf-discuss] eRDF <=> microformats?

[Ian Davis]

Hi Ernie,

On 31/05/2006 18:09, Dr. Ernie Prabhakar wrote:
> 
>  From my admittedly naive perspective, eRDF looks like it *could* be 
> used in a way that is compatible with microformats.  That is, not *all* 
> eRDF schemas and documents *are* necessarily microformats, but many of 
> them _could_ be.  Conversely, it seems like _most_ microformats could be 
> described using eRDF schemas.
> 
> Is that true?  If so, what is the dividing line?

I'm the author of eRDF so I'll take a stab at answering this. I agree 
with your analysis but there are some differences between the mf 
approach and the eRDF one. The main difference is that eRDF has and 
requires a namespacing capability. This is to support the RDF notion 
that important relationships are named with URIs. This is opposed to a 
key mf principle, but is necessary to bridge the two worlds. In eRDF 
only the class names and rel/rev attributes that match a namespace 
prefix are considered significant, all others are ignored. This would 
make it difficult to distinguish mf classes without special case parsing.

However, eRDF was designed to work comfortably with mf in an HTML 
document and Dan Connolly has done some interesting work mixing the two 
together (see source of http://www.w3.org/people/connolly/)

I've been an RDF advocate for a long time but I'm not a bigot :) I like 
to use appropriate technology and although the RDF and mf worlds orbit 
very closely they address different audiences and different needs. We 
use RDF extensively at Talis but we want to expose metadata in lots of 
different ways. So, our directory of libraries and web services is built 
with an RDF backend but we expose hCards for library addresses.

> -- Ernie P.

Ian
-- 
http://purl.org/NET/iand
Blogging at... http://iandavis.com/blog
Working on... http://directory.talis.com/


> 
> P.S.  For what its worth, this seems _way_ more promising to me than RDFa.

P.S. Thanks, but I don't position it as a competitor to RDFa. For a 
start eRDF doesn't attempt to support all possible RDF statements 
whereas RDFa does which is why it has new elements and attributes. eRDF 
works without modifying the HTML DTD but you can't always embed all the 
metadata you might want to.

P.P.S. BTW we met very briefly at a mf dinner at the Web 2.0 conference 
last year. It was that meeting and a couple of conversations with Tantek 
et al. that prompted me to develop eRDF.