Re: Graph naming?

On Wed, 25 Feb 2004 11:28:58 +0100, Eric Jain <Eric.Jain@isb-sib.ch> wrote:

> 
> > rdfstore:context properties to flag each description block
> 
> This is an excellent solution, wish I had thought about it earlier!
> 
> The only drawback I see is that this solution can end up being a bit
> verbose, but then again it's completely backard-compatible.
> 
> So, if the official serialization syntax is ever revised, defining an
> rdf:context property would in my opinion to be a better approach than
> anything that requires tools to be rewritten from scratch.

You may not be surprised to know that the RDF Core WG discussed some
of this and felt that it wasn't appropriate to add during the last round of changes.
Don't hold your breath for revisions.

Alberto's idea of adding a rdf:context property attribute is ok and
parsers should handle it fine (or they are not rdf/xml parsers).

The key issue is as long as you know what kind of context is being used;
we identified three types at a SWADE workshop[1]:
  1) source URI (where the triples came from, usually always a URI)
  2) subgraph/submodel identifier
  3) an internal/statement identifier

You have to be clear what you are defining for context.

I'd say that adding anything to rdf:RDF will break parsers; they are entitled
to assume that it can have no attributes other than those specified in the syntax.

rdf:RDF could have been used to label graphs and sub graphs but again,
the WG left this alone.  The question of "what does naming a
graph/subgraph mean?" seems the first and most important thing to answer.

Cheers

Dave

[1] http://www.w3.org/2001/sw/Europe/reports/dev_workshop_report_4/#sec-context

Received on Wednesday, 25 February 2004 06:05:31 UTC