Re: Need pointer to namespace discussion (rdf vs rdfs)

On Tue, 17 Sep 2002, Murray Spork wrote:

>
> Collin wrote:
> > Hi, Murray:
> >     RDF = Resource Framework Framework
> >     and
> >     RDFS = RDF Schema.
> >     But we usually use RDFS to refer both RDF and RDFS since the two concepts are very correlated.
> >    Wishes it is useful .
> >    Best Regards
> > Collin
>
> Hi Colin,
>
> Sorry - I was not clear enough in my original email - what I need to
> know is why (for e.g.) rdfs:Resource is in the rdfs namepace and not in
> the rdf namespace.

90% historical accident. The old 1997-1999 Model and Syntax RDF Working
Group tried to avoid defining more than it needed to. It gave us the
notion of rdf:type (since we had special case RDF/XML syntax for encoding
info about types in XML Markup). It gave us notion of rdf:Property, since
each arc-label in a graph is of rdf:type rdf:Property. But it left to the
RDF Schema WG the work of fleshing out the type/class system, ie. the
notion that rdf:Property is of rdf:type rdfs:Class, and that each thing
that was an rdf:type of something was itself a thing of rdf:type
rdfs:Class.  And that classes had subclass relations, etc etc. In this
context, it fell to the RDFS WG to give a name to the class rdfs:Resource.

Sometimes it's easy to slip into looking for technical / engineering
reasons for design features that are artifacts of the way these specs are
produced. It is only natural for such pragmatic considerations to affect
specs, though if things went too far alarm bells should probably ring...

Dan


ps. if rdfs:Resource was in the RDF M+S spec not RDFS, we could have lost
the rdf:Description construct from the RDF XML syntax, simplifying things a
bit. Bummer. Probably not worth worrying about now...

-- 
mailto:danbri@w3.org
http://www.w3.org/People/DanBri/

Received on Monday, 16 September 2002 20:03:25 UTC