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

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

From: <Patrick.Stickler@nokia.com>
Date: Thu, 7 Jun 2001 17:34:51 +0300
Message-ID: <6D1A8E7871B9D211B3B00008C7490AA50795874D@treis03nok>
To: sean@mysterylights.com, Patrick.Stickler@nokia.com, seth@robustai.net, www-rdf-interest@w3.org
Cc: Ora.Lassila@nokia.com


> -----Original Message-----
> From: ext Sean B. Palmer [mailto:sean@mysterylights.com]
> Sent: 07 June, 2001 16:40
> To: Patrick.Stickler@nokia.com; Seth Russell; www-rdf-interest@w3.org
> Cc: Ora.Lassila@nokia.com
> Subject: Re: What to do about namespace derived URI refs... (long)
> 
> 
> > Perhaps Sean would be interested in a combined effort for a 'name'
> > URN scheme that would correspond to the hierarchical path syntax
> > defined by my HRN example along with an optional temporal field for
> > those applications which require names that are temporally as well
> > as globally unique.
> >
> > Eh?
> 
> Fine by me. However, I don't really see people starting to take it up
> all that fast... for as long as URLs are around and people can use
> them in RDF, they will be used - you can't stop that from happening,
> and rightfully so. 

Well, I'll agree that it's going to be darn hard if not impossible
to stop, but not necessarily "rightfully so" ;-)

> The RDF applications that I use work just fine with
> URLs, and yes, most of them I use strictly as "names" rather than
> retrievable resources. But they work for the kind of stuff I'm doing.

But will they work for everything they have to work for in the SW?

> There is the problem with referring to concepts cross-media, and for
> that I support stuff like the tag URI scheme, and/or the name URN if
> it ever gets off of the ground. 

But references to concepts cross-media is essential to the SW! How do
folks expect intellegent agents to trully communicate if there are 
no reliable mechanisms for cross-media references to shared concepts?!

If my agent requests data from another and it is provided as an XML
instance with an XML Schema yet it cannot reliably map the property
element <locale>fi</locale> to the abstract definition of the ISO 3166-1 
country "Finland", then it's just noise and wasted bandwidth.

And since my agent normally does not use XML Schema, it *certainly* isn't
going to be reifying the country "Finland" with a URL ref to that
XML Schema using XML Schema MIME content type fragment syntax! Eh?

How does the mapping occur? Where is it defined? 

Furthermore, who says what the "official" standard URI ref for the ISO
3166-1
language "Finland" is? The only official ISO specification is in print only.
All online (URL referencable) definitions are unofficial -- and they vary
from HTML to text to CSV, etc.

For the SW to exist, there must be an intersection of shared knowledge,
and for shared knowledge there must be shared names, and for there to
be shared names they must be persistent and immutable across arbitrary
interchange mechanisms. I just don't see the necessary mechanisms for
this kind of reliable syncronization between the current family of 
standards XML, XML Namespaces, RDF, XML Schema, etc.

The issue of standardized ontologies fortunately can be addressed
irrespective of the above problem, but unless those ontologies can
traverse the SW infrastructure intact, then there is no SW.

> But if you really want to get going,
> you'd better do it fast: URLs as namespaces is becoming a heavily
> established practice.

Whether Joe public will ever get weaned away from incorrect use of
URLs as abstract concept names is of secondary concern to whether the
infrastructure of the SW supports the interoperability between agents
that must occur for any true benefit to be realized.

And hopefully, tools and GUIs will eventually help the general 
user to "do the right thing" insofar as the correct use of 
various URI schemes is concerned.

> Does this mean that the RDF, RDF Schema, DAML, Dublin Core, FOAF,
> DCTypes, EARL, Annotea and so on namespaces are broken? 

IMHO, yes. And the fact that this problem is so pervasive is rather
frightening. The reason why they work is probably because the URIs
employed are unique strings within those particular (closed) systems
and thus they have worked sufficiently well as names -- but I would
argue that they are not unambiguous, consistently portable, nor 
universal names for resources
which remain constant across all media encodings or MIME content
type interpretations. We're about to hit a scalability barrier once
all these small isolated systems with "ontologies in a vacume" have
to actually interact on the web with a consistent and explicit 
intersection of identifiers. If the identifiers cannot survive
interchange via serialization because of differing interpretations
as to how to derive identity of resources from the namespace and
the serialization encoding, and various systems and standards cannot
agree how to construct the same set of URIs from QNames, then there's 
no interchange of knowledge at all and the SW is still born.

Perhaps we should forego XML serialization completely and just 
encode and exchange triples... ;-)

Cheers,

Patrick
Received on Thursday, 7 June 2001 10:35:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:49 GMT