RE: rdf:text review



> -----Original Message-----
> From: Axel Polleres [mailto:axel.polleres@deri.org]
> Sent: 29 April 2009 15:51
> To: Seaborne, Andy
> Cc: 'RDF Data Access Working Group'
> Subject: Re: rdf:text review
> 
> Andy,
> 
> we seem to be almost there from my side. I just was tempted to
> relativize 1/ as well, since I believe it is likewise remedied by
> resorting to D-entailment effects only.
> Do you agree?

No - because I don't understand "relativize".

If you mean an rdf:text aware system can choose to expose rdf:text, then I do not agree.  The client system may well be an RDF one.  If they choose to negotiate a different format (i.e. SPARQL results with rdf:text), then they are welcome to do that (it's a pairewise agreement) but the base line should be to maximise compatibility.  rdf:text processors are already required to accept the DRF serialization.  RDF-Graph exchange requires RDF serialization; so the same should apply here.

If later standards come along and put rdf:text in RDF, then change all the related bits and piece then not plan for an unclear future.  Leaving "if", "buts" and "maybes" now just hurts inoperability.

(Also, if the rdf:text-entailment is mixed with non-rdf:text-entailment in one query, what happens?)

 Andy

> 
> As for 2/ and 3/, I am fine. We have to revisse the current response
> text accordingly, only started with that so far.
> 
> Axel

Received on Wednesday, 29 April 2009 15:04:02 UTC