Re: reification vs. named graphs

On 14 Aug 2007, at 10:24, Story Henry wrote:

>
> Does the current standard not in fact allow graphs through the use  
> of XML/RDF literals? If an RDF document contains a relation  
> pointing to an RDF/XML literal, then that RDF/XML is playing  
> somewhat the role of a graph, no? I mean it is opaque in the same  
> way a graph is...
[snip]

I've pushed this from here to Sunday with no takers. I think we  
actually had it in a version of OWL-S for a while to handle  
preconditions and effects. We ended up taking it out, IIRC, because  
1) tool support was poor, 2) you don't end up with and obvious  
"object" in the object (other than the literal itself) so some common  
parsing patterns (i.e., over graphs,  not over the serialization) got  
much trickier, and 3) there are issues, though I don't recall what  
they are, about, e.g., namespaces or something.

A solution, nevertheless, worth investigating.

Cheers,
Bijan.

Received on Tuesday, 14 August 2007 10:16:01 UTC