RE: The X Datatype Proposal

> -----Original Message-----
> From: ext Graham Klyne [mailto:Graham.Klyne@MIMEsweeper.com]
> Sent: 15 November, 2001 14:47
> To: Stickler Patrick (NRC/Tampere)
> Cc: w3c-rdfcore-wg@w3.org
> Subject: RE: The X Datatype Proposal
> 
> 
> At 11:32 PM 11/14/01 +0200, Patrick.Stickler@nokia.com wrote:
> > > Well, yes, I have in the past found it useful to think 
> about RDF in terms
> > > of reified statements.  (e.g.
> > > 
> http://public.research.mimesweeper.com/RDF/RDFContexts.html.) 
>  But I have
> > > always seen any such view as being underpinned by the 
> basic RDF triple
> > > model, rather than vice versa.
> >
> >But the problem arises that, in any context where you must qualify
> >statements (and that's darn near everywhere, really) you have to
> >reify the statements, yet that is an extra step with the current
> >model, and should be provided IMO for free.
> 
> Well, I think you need to be able to use statements in some 
> way without 
> asserting them.  

I fully agree, which is the default for the X model. A 
statement is just a statement. It is not asserted until
and unless someone makes a statement about that statement
that indicates that it was asserted, e.g. via the
proposed property assertedBy (some authority).

> It is possible that the mechanism that RDF calls 
> "reification" is one way to achieve this.  I have found that 
> it is not 
> difficult to invent other mechanisms using RDF that can have 
> the same effect.
> 
> I think a lot of this debate is simply about the viewpoint 
> one chooses to 
> adopt.  I can see why you might find it attractive to use a reified 
> statement as your "primitive";  others view the  unreified triple as 
> primitive. I think many people here feel that RDF is not 
> necessarily the 
> best possible starting point for where we're trying to get 
> to, but it is 
> what we have and, most importantly IMO, it has a body of 
> active consensus 
> across a broad range of interested parties (theorists, developers, 
> information designers) that it is a not-wholly-unreasonable 
> starting point.

I'm not arguing that. I find the present triple based model
to be much easier to think in terms of from a resource
perspective, and certainly easier for notations.

As I hope I made clear in several posts, but which I 
state again here, I see the X model as a layer below
the triple based model. It does not itself use triples,
and the constructs are not themselves statements, but
they model statements. 

That's an important distinction (I think).

Patrick

Received on Thursday, 15 November 2001 12:32:50 UTC