Re: Annotea's context property

The value of the context is a literal, not a resource. It looks like a URI, 
but it could have any other value.

I think that the previous draft of the protocol had an error and was
showing the value of the context as a resource. 

I've some net problems so I can't verify the Annotea RDF schema to see
how we declared the context there or if there's an ambiguity.

In all cases, it is a literal.  The ZAnnot server is correct.

-jose

On Mon, Jan 27, 2003 at 11:29:08AM -0000, Jim Ley wrote:
> 
> 
> "Charles McCathieNevile" <charles@w3.org> wrote in message
> news:Pine.LNX.4.30.0301262115510.8589-100000@tux.w3.org...
> >
> > in the protocol document [1] all the examples for context use a literal as
> > its value. The schema doesn't specify any domain or range.
> >
> > In my annotools I used an rdf:resource as the object, on the basis that it
> > was going to be an Xpointer or a URI. The ZAnnot server is returning this
> as
> > if it were a literal.
> 
> This changed between the old and new protocol, it's the one area of concern
> I have with the new protocol, context seems problematical, I still want to
> be able to annotate svg/raster images via my RDF img parts (like hte RDF
> generated in http://jibbering.com/svg/annotateimage.html - example:
> http://jibbering.com/rdfsvg/example.rdf ) which makes for the context to be
> a lot more complicated object than is really expressable in a literal.  I've
> been trying to think of an alternative suggestion, but not come up with one.
> 
> > And what do implementations expect?
> 
> I don't distinguish between literal/resource, if they smell the same they're
> used the same.
> 
> Jim.
> 
> 
>

Received on Thursday, 30 January 2003 11:56:10 UTC