W3C home > Mailing lists > Public > www-rdf-interest@w3.org > August 2001

RE: Summary of the QName to URI Mapping Problem

From: <Patrick.Stickler@nokia.com>
Date: Fri, 17 Aug 2001 21:25:09 +0300
Message-ID: <2BF0AD29BC31FE46B788773211440431245BFB@trebe003.NOE.Nokia.com>
To: aswartz@upclink.com
Cc: SCranefield@infoscience.otago.ac.nz, www-rdf-interest@w3.org


> -----Original Message-----
> From: ext Aaron Swartz [mailto:aswartz@upclink.com]
> Sent: 17 August, 2001 16:09
> To: Stickler Patrick (NRC/Tampere)
> Cc: SCranefield@infoscience.otago.ac.nz; www-rdf-interest@w3.org
> Subject: Re: Summary of the QName to URI Mapping Problem
> 
> 
> On Friday, August 17, 2001, at 08:03  AM, 
> Patrick.Stickler@nokia.com wrote:
> 
> >> Yes, there is a bug in RDF/XML with regards to this. Use
> >> N-Triples or N3 or something instead.
> >
> > Well, I'd rather use the standard (and see it fixed). I could
> > use those, or just as well roll my own serialization model, but
> > that doesn't mean that I can expect that any arbitrary SW agent
> > is going to be able to eat my data...
> 
> Well, it seems unlikely that the current WG is going to fix 
> RDF/XML to address this issue. Meanwhile, N-Triples is shaping 
> up to be very useful for this kind of interchange format. Does 
> the spec address your issues?
> 
> http://www.w3.org/2001/sw/RDFCore/ntriples/

Ummm, sorry if I keep sounding like a broken record here, but
if the solution is not *the* standard that is used by *every*
SW agent, then there will be no reliable and successful SW
(at least in the global, unconstrained sense, such as the WWW).

There is a difference between core standards that provide
the common infrastructure and applications and auxiliary
standards that extend the core functionality, but true
global interoperability will be only as successful as the
core standards, so even if N-Triples is shaping up nicely
and does everthing including make the coffee, if I need
functionality that N-Triples only offers and the standard
RDF serialization doesn't, and not every SW agent understands
(or is required to understand) N-Triples then its utility
is zero to me insofar as the global value of my data.

The SW is about global information interchange, right? And the
mechanism for that interchange is serialized XML, right? Even
if the conceptual model for processing that information is
independent of the serialization model, the SW as a whole is
*not* independent of the need for a single consistent and
reliable serialization model. 

If the W3C formally announces that N-Triples or N3 or whatever
will be a serialization that all RDF parsers must eat, then fine,
I'm happy (enough) and we can get on with business. If not, and
if the only official, standards mandated serialization model
is the present one, then we're still going to keep falling into
that hole (even if some have been fortunate enough to avoid it
thus far ;-)

> >> There is an issue about this on the list, but I agree, it is a 
> >> problem.
> > Thank you. Perhaps you can then explain why it is a problem to
> > Dan and the others who don't seem to understand what I'm talking
> > about...
> 
> DanC and others understand this issue well -- I've spoken to 
> them about it. 

I'll try to take your word for it, though comments in the recent threads
don't have me fully convinced that that is so.

> It's when you start claiming that different 
> QNames must be disjunct 

Yup, they do. And actually it's the XML NS spec that says it.

They don't necessarily have to correspond to two different
RDF resources (as there can be intentional synonymy between
different vocabularies which are defined as equivalent)
but it is not for RDF to say IMHO that they cannot correspond to
differt RDF resources -- and I would argue that in most cases they
are intended to be different, because of the IMO proper interpretation
of the XML NS spec.

> and that URIs are really made up of two 
> parts that you go off the deep end.

Uhhh, no, sorry, never said that. 

In fact, I said explicitly the opposite, that QNames are NOT equal to URIs.
QNames are made up of two parts. URIs are not. No disagreement there I 
think.

Cheers,

Patrick

--
Patrick Stickler                      Phone:  +358 3 356 0209
Senior Research Scientist             Mobile: +358 50 483 9453
Software Technology Laboratory        Fax:    +358 7180 35409
Nokia Research Center                 Video:  +358 3 356 0209 / 4227
Visiokatu 1, 33720 Tampere, Finland   Email:  patrick.stickler@nokia.com
 
Received on Friday, 17 August 2001 14:26:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:51 GMT