[uf-discuss] Potential for Microformats.org to work with W3C and RDFa Task Force

[Ben Ward]

On 28 Aug 2008, at 12:24, Manu Sporny wrote:

> It will take a couple of weeks to give examples of how this will all
> work, but I wanted to get feedback from this community before
> proceeding. We have a fantastic opportunity in front of us now - who  
> in
> this community thinks that we should work with the W3C on this  
> endeavor?

I'm not sure I completely see the benefit in this, and seeing your  
examples would be very helpful in getting a better idea of what you're  
proposing. From your bullet points, it seems to suggest taking  
microformat vocabularies and expressing them in RDFa, rather than  
HTML? It seems redundant for publishers.

However, I do have a somewhat related issue that you might consider  
part of this effort. Some discussions I've had lately revealed  
usefulness in being able to _map_ microformats into RDF, for the  
purpose of combining microformats with other RDF vocabularies in a  
back-end somewhere (so, conversion for processing, rather than  
publishing. Publishing remains in HTML where it is most effective).

I'm told that RDF ‘versions’ of vcard and icalendar are out of date  
compared to the microformats. As such, it strikes me that rather than  
maintaining duplicate specifications, it would instead make sense to  
develop a set of standard transformations so that any microformat can  
be transformed from HTML to RDF, without requiring duplicate effort to  
maintain another spec. This I'm sure would relate closely to GRDDL,  
since that already deals with transformation.

This latter issue seems valuable, and preferable to a situation where  
every processor of microformats and RDF comes up with their own  
incompatible conversions.

Note, I'm talking about mapping rules, not separate specs. For  
example, we have the ‘jCard’ page on the wiki, which I still feel  
should be more generic ‘JSON Mapping Rules’ page that can cover  
parsing from any format, not just hcard. If this RDF mapping effort is  
pursued by anyone, I would again favour ‘RDF Mapping Rules’, rather  
than ‘rCard’, ‘rCal’ and ‘rListing’ — duplicate specs not based in  
HTML are not something that this community was founded to produce.

Cheers,

Ben