On 7/5/09 08:51, Martin Hepp (UniBW) wrote:
4. Until this will have happened, there is no rush to update existing applications that create output using the old vCard specification. (Due to the masses of "old" vCard data, it would not be prudent for any data-consuming application to accept only "new" vCard data anyway - see my SWOOGLE figures from the initial mail - 230,000 vs. 470 datasets - which implies that continuing to create "old"-style vCard should be tolerable.)
Given that vCard is itself again an moving target, and that the most comprehensive modern treatment of "classic addressbooks" is now Portable Contacts, the best future direction isn't entirely clear. I intend to better align the addressbook-ish portion of FOAF more closely with the Portable Contacts / hCard idiom, but also look into a handling of the "work", "home" etc distinctions that can be applied to all kinds of account (blog, bookmarks etc.), so I don't expect perfect bi-directional 1:1 mappings.
A practical question: can the older flavour of vCard in RDF plausibly be handled by SPARQL CONSTRUCT rules, or does the use of rdf:Bag etc make this tricky?
I encourage anyone interested in progressing this discussion to join the new W3C Social Web incubator group, see http://www.w3.org/2005/Incubator/socialweb/ ... we had our first meeting yesterday and could address this through a taskforce or more informal liaison with any group of collaborators who'd like to report to their peers through the XG.
cheers, Dan
