Re: namedNode? in predicate position?

There seems to be some divergence of view on this.  Lets make it an issue.

Does the model theory take a position on prince arcs?


Brian

Dave Beckett wrote:
> 
> >>>jos.deroo.jd@belgium.agfa.com said:
> > while at next telecon agenda items,
> > what about the N-triples/MT related questions
> > 1. predicate ::= uriref  versus predicate ::= uriref | namedNode?
> 
> Does RDF allow, let's call it, non-URI-ref for predicates?
> I don't think so, at present.  In the graph model in the original
> M&S, predicates are arrows with URIs, they are never empty circles.
> 
> > 2. why do we use the term namedNode for a node which is in fact not named?
> 
> It was anonNode - which was probably worse, so I changed it.
> How about princeNode?
> 
> It is just a token in the N-Triples grammar and if it will reduce any
> implied meaning by changing the characters of the token, let's do it.
> 
> > ps question 1 is in fact related with the problem to write
> >    e.g. the following N-triples
> >      <#a> _:x <#d>.
> >      _:x <#b> <#c>.
> >    in RDF/XML syntax
> >    Jeremy/Dave, do you have trouble with that?
> 
> Well at present I say it is illegal N-Triples.
> 
> Dave