Re: [namespaceDocument-8] Proposal: "Namespace Document" = RDF

On 2002-03-08 18:33, "ext Jonathan Borden" <> wrote:

> Patrick Stickler wrote:
>> See above. RDDL as presently defined does not appear to
>> be able to describe arbitrary resources denoted by URIs,
>> but only to describe a "namespace", which is odd, since
>> there's not much one can actually say about punctuation...
> a) This discussion is _about_ "Namespace Documents". Any inability to
> describe arbitrary resources is not within the scope of this discussion.

I hardly see how this is out of scope. It is directly concerned with
what a namespace document is intended to contain and the utility it
is intended to provide.

I suggested that a namespace document is a placeholder for descriptions of
*any* arbitrary resource having some *relation* to the namespace, whatever
that relation might be.

Such resources may be vocabulary terms, functional vocabularies, document
models, schemas, stylesheets, software components, etc. Statements about
those resources are not statements about the namespace.

If a namespace document only describes the namespace, then it will
be a very small, boring document. In fact, I would
not expect there to ever be any statements about the namespace itself
in any namespace document. The namespace is simply a point of focus.

(of course, that's taking the view that a namespace is punctuation, not a
vocabulary, or a document model, or a schema, or a ....)

> b) to quote the RDDL spec:
> [[
> This document describes the Resource Directory Description Language (RDDL).
> A RDDL document, called a Resource Directory, provides a package of
> information about some target, including:
> ]]
> Nowhere does it say that RDDL is limited to the description of namespaces (I
> can assure you of that). Nonetheless RDDL was not designed as a replacement
> for RDF.

Fair enough. But perhaps RDF, in addition to XHTML for humans, is what
we really need.

A "package of information about some target" sounds like your describing
the namespace itself, if that package is the namespace document.

> c) You have asked about the ability to describe an "arbitrary" resource in a
> RDDL document.
> ... 
> So you see that by incorporating xml:base (which RDDL does), the RDDL syntax
> is very close to the RDF N-Triples syntax. This was not an accident.

OK, I can see that RDDL's use of XLink can be coerced into working
like RDF, but 

(a) why would I want to use something other than RDF to say things like
    one says in RDF?

(b) how can I be sure that others will use RDDL/XLink in a way that is
    mappable in a similar fashion to RDF?

> Indeed one can argue, I think successfully, that the XLink syntax is closer
> to RDF N-Triples than the RDF XML syntax.

No argument there ;-)

> From an architectural point of view, it would be good to more
> officially/explicitly specify the relationship between XLink and RDF (the
> current interpretation is from Ron Daniel's W3C Note)

One problem with the XLink-like-RDF approach is that it only works
if folks intend to express their knowledge in an RDF like manner,
and thus, true machine accessiblility of namespace document defined
knowledge as RDF statements remains a wildcard.

If RDDL combined XHTML and RDF rather than XHTML and XLink, I
think things would be both clearer and consistent for machines
while still meaningful for humans.



Patrick Stickler              Phone: +358 50 483 9453
Senior Research Scientist     Fax:   +358 7180 35409
Nokia Research Center         Email:

Received on Monday, 11 March 2002 11:27:13 UTC