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 19:04:33 +0300
Message-ID: <1E4A0AC134884349A21955574A90A7A50A1C34@trebe051.ntc.nokia.com>
To: <dave.beckett@bristol.ac.uk>
Cc: <chris@bizer.de>, <otto@math.fu-berlin.de>, <Eric.Jain@isb-sib.ch>, <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 Dave Beckett
> Sent: 26 August, 2004 18:52
> To: Stickler Patrick (Nokia-TP-MSW/Tampere)
> Cc: chris@bizer.de; otto@math.fu-berlin.de; Eric.Jain@isb-sib.ch;
> www-rdf-interest@w3.org
> Subject: Re: Reification - whats best practice?
> 
> 
> 
> On Thu, 26 Aug 2004 18:40:15 +0300, 
> <Patrick.Stickler@nokia.com> wrote:
> 
> > > From: "Dave Beckett" <dave.beckett@bristol.ac.uk>
> ...
> 
> > > > I noticed that Named Graphs extends RDF in at least two ways:
> > > > 1) RDF triple subjects can be literals
> > 
> > This extension is (while IMO worthy of consideration) not
> > part of named graphs, but of TriX, TriG, etc.
> > 
> > Consideration of literal subjects is disjunct from consideration
> > of named graphs.
> 
> Yeah, but I did notice it the footnote describing it as a "legacy
> constraint".  It is the RDF triple definition we shipped.  Not legacy.

"Legacy" as in, when the papers on TriX, Named Graphs, etc. were
written, the new RDF specs were full Recs, and therefore, any constraint
defined in such specs was, and is, "legacy". The footnote is not in
reference to M&S.

> 
> > 
> > > > 2) RDF triples are quads (sic)
> > 
> > No. RDF triples are RDF triples. Graphs are sets of triples.
> > Graphs, just as any resource, can be named/denoted by URIs. 
> > Statements made about graphs are triples, where the URI denoting
> > the graph occurs as the subject.
> > 
> > In any storage implementation or serialization, however, membership 
> > of statements in a given graph can be captured as quads (among other
> > possible methods) consisting of graph, subject, predicate, 
> and object.
> 
> Yes.  Maybe triples associated with a Named Graph name (URIref).

Yup. I used 'quads' because the thread originally was discussing
storage/serialization/representation issues that related to the
use of quads.

The actual model is, as you say, simply triples associated with a
named graph.

 
> > Such quads are not "extended triples" insofar as triples are
> > defined by the RDF specs. They are simply a data structure
> > for expressing triples within the (possibly named) graphs to
> > which they belong.
> > 
> > > > so it's really Named non-RDF Graphs.
> > 
> > No. As I hope the above clarifies, it is certainly
> > named *RDF* graphs.
> 
> I apologise.  You are of course, correct.

No apology necessary. I simply wanted to keep things straight 
(and as the terminology used in the thread has been a bit free,
it's no wonder some degree of confusion might arise).

Cheers,

Patrick
Received on Thursday, 26 August 2004 16:04:49 UTC

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