RE: What to do about namespace derived URI refs... (long)

> -----Original Message-----
> From: ext Sean B. Palmer [mailto:sean@mysterylights.com]
> Sent: 06 June, 2001 20:27
> To: Seth Russell; Patrick.Stickler@nokia.com; www-rdf-interest@w3.org
> Cc: Ora.Lassila@nokia.com
> Subject: Re: What to do about namespace derived URI refs... (long)
> 
> 
> > But if the web page is describing an abstract entity [see
> > below], and some author decides to coin the URI for
> > that abstract entity to be his web page URL; then the
> > RDF cannot tell the difference between descriptions of
> > the abstract thing and descriptions of web page.
> 
> But then the author has assigned a URI to two different resources,
> which is prohibited by the URI specification. It's either a Web Page,
> or it's the concept of some abstract thing, like love or hatred, or it
> represents some physical object like a brick, but not all three. Not
> any more than one.
> 
> So you have http://robustai.net/seth/ and it's your homepage. Fine.
> You can put it in your RDF bookmark system. Now, you want to coin some
> abstract entity that is defined by a view of that resource, for
> example "Truth", so you use the URI http://robustai.net/seth/#Truth or
> even http://robustai.net/seth/Truth as long as the latter is never
> referred to as a Web Page. Even if it is, there isn't a problem,
> thanks to contextualization - someone can call it a Web Page if they
> want to - it won't break any of your programs.
> 
> --
> Kindest Regards,
> Sean B. Palmer
> @prefix : <http://webns.net/roughterms/> .
> :Sean :hasHomepage <http://purl.org/net/sbp/> .
> 


To me this is wrong in the same sense that taking company pens
home is wrong. You'll probably always get away with it, but I
wouldn't expect you to brag about it ;-)

Even though clever folks who grok the RDF and the semantic web
might be able to differentiate between the two "interpretations"
of web page versus abstract concept -- the average content
producer who might be wanting to create content and rich
metadata to benefit from the semantic web likely will not, so
it will hinder adoption of semantic web technologies and
standards. 

I hear time and again how *confusing* RDF is, and
I believe that one of the roots of that problem is the fact
that folks are looking at the serialization of instances
and cannot see how they relate to triples. Yes, yes, the
general details are there in the spec, but still it takes
a while to understand the relations, and most folks give
up long before they look at the data model and come to 
appreciate that that is where 99% of the value of RDF
lies, and the serialization is just a means to an end
(personally, I'd love to see all serialization aspects
simply left out of the revised specs, with just guidelines
about how serialization should map to triples).

If we cannot provide content creation (i.e. serialization)
models that make the specification of knowledge sufficiently
convenient for the "common Joe" and the relationship of that
serialization to the underlying data model clear and intuitive,
then folks won't use RDF. I'm not talking about XML vs. GUI.
I'm talking about schemas which have a clear mapping to
the RDF data model, including and especially a consistent
representation for identity. A rose is a rose is a rose
and we *won't* call it by any other name, even if it *would*
smell as sweet ;-)

Though I don't believe that "my grandmother" should be able
to grok everything, it should nevertheless be reasonable clear
and obvious, and above all *consistent*.

The current situation with using URLs to identify abstract concepts
is just begging for trouble. It is non-obvious. It is unclear to
users (e.g. the common complaint "Why can't I get the schema from
this namespace URL, it's broken!), it prevents a clear distinction
between an abstract concept and some concrete realization of that
concept, etc. etc. yada yada yada... It's simply wrong. IMHO ;-)

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 Thursday, 7 June 2001 03:47:03 UTC