W3C home > Mailing lists > Public > www-rdf-interest@w3.org > August 2004

RE: Reification - whats best practice?

From: <Patrick.Stickler@nokia.com>
Date: Thu, 26 Aug 2004 18:59:48 +0300
Message-ID: <1E4A0AC134884349A21955574A90A7A50A1C33@trebe051.ntc.nokia.com>
To: <chris@bizer.de>, <alberto@asemantics.com>
Cc: <www-rdf-interest@w3.org>



> -----Original Message-----
> From: www-rdf-interest-request@w3.org
> [mailto:www-rdf-interest-request@w3.org]On Behalf Of ext Chris Bizer
> Sent: 26 August, 2004 18:43
> To: Alberto Reggiori
> Cc: www-rdf-interest@w3.org
> Subject: Re: Reification - whats best practice?
> 
> 
> 
> > On Thu, 26 Aug 2004, Chris Bizer wrote:
> >
> > >
> > > From: "Dave Beckett" <dave.beckett@bristol.ac.uk>
> > > > > ... and you can even get them quite easily into 
> RDF/XML, simply by
> > > > > allowing rdf:about/rdf:ID/rdf:nodeID on the <rdf:RDF> wrapper.
> > > > > Although there are some good arguments for a 
> completetly different
> > > > > serialization syntax such as TriX...
> > > >
> > > > We thought of that.  But slipping in such a huge change to RDF
> > > > in an existing syntax wasn't where we ended up.
> > > >
> > > > I noticed that Named Graphs extends RDF in at least two ways:
> > > > 1) RDF triple subjects can be literals
> > > > 2) RDF triples are quads (sic)
> > > >
> > > > so it's really Named non-RDF Graphs.
> > > >
> > >
> > > Really?
> > >
> > > A collection of RDF/XML documents on theWeb map naturally into the
> abstract
> > > syntax of Named Graphs, by using the first xml:base 
> declaration in the
> > > document or the URL from which an RDF/XML file is 
> retrieved as a name
> for
> > > the graph given by the RDF/XML file.
> >
> > it could work if you assume your graph-name is a URI - but 
> it might not be
> > the case - and you might want eventually reference-by-description to
> > those graphs (then the graph might be a bNode)
> >
> > 
> http://lists.w3.org/Archives/Public/public-rdf-dawg/2004JulSep
> /0328.html
> >
> 
> Nice use case. I definitively think DAWG should have 
> something like ?SOURCE
> in the query language. The semantic *WEB* is about distributed data
> published by many sources and I don't see the point in having a query
> langage which can't reflect this.

+1

> 
> > or you have a stronger argument about it - i.e. why graphs 
> can not be
> bNodes?
> >
> 
> Actually we don't have an argument but just decided to 
> restrict graph names
> to be URIs in order to keep things simple. It might be useful 
> to discuss
> about loosening this restriction in the future.

Pat Hayes and Jeremy Carroll can probably provide some
arguments with more "meat" (there were some MT issues
which made things hairy), but one reason why bnodes are
disallowed as graph namess is because bnodes are graph-specific,
and the intention is that graph names are inter-graph in,
scope i.e. global. Thus statements about a particular graph
can occur in some other graph, which would preclude (in
that case at least) using a bnode.

Historically, TriX allowed bnodes as graph names.

Patrick
Received on Thursday, 26 August 2004 16:00:15 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:14:57 UTC