Re: SWIG F2F during W3C TPAC week, Oct 20/21 (Cannes, France)

----- "Danny Ayers" <danny.ayers@gmail.com> wrote:

> 2008/9/3 Phil Archer <parcher@fosi.org>:
> 
> > Actually the responses to this are interesting IMO. It all began
> when I put
> > my head above the parapet and suggested that a possible discussion
> point for
> > TPAC /might/ be testing the water to see if there was a critical
> mass of
> > folk that wanted to think about RDF 2, or at least, looking at some
> of the
> > issues that crop up on this list and elsewhere from time to time
> that folk
> > seem to agree need fixing.
> >
> > I've got the message that the answer is a firm 'no' - OK - I'll move
> on...
> 
> Sorry Phil, I for one didn't want to give that impression. Discussion
> of RDF 2 sounds a great discussion point for a f2f, whatever the
> opinions.
> 
> Without any thought I can say drop RDF's version of reification,
> introduce named graphs. Are Lists working ok? I dunno.

One thing here to contribute. With named graphs in SPARQL, much of the power currently comes from the ability to define this at the database level to fit the granularity desired by a particular user. If Named Graphs were decided by publishers, it would be nice not to lose the granularity options in RDF Databases, and SPARQL queries on these databases, which are currently configurable completely by users to fit their needs.

On the other hand, it might be nice to formalise some idea about what a Named Graph means in RDF terms, although inevitably I guess this will mean moving from RDF triples to RDF quads as the basic language element! ;-) From an OWL point of view, the move from Triples to Quads would also have to be taken into account I guess in relation to deeper theory. Is it worth disturbing the ecosystem to gain a formal specification of what a Named Graph is if it comes to that?

Cheers,

Peter

Received on Thursday, 4 September 2008 02:37:00 UTC