Re: What does RDF consider a namespace to be?

From: Brian McBride <bwm@hplb.hpl.hp.com>
Subject: Re: What does RDF consider a namespace to be?
Date: Mon, 24 Feb 2003 17:19:16 +0000

> At 10:03 30/01/2003 -0500, Peter F. Patel-Schneider wrote:
> 
> 
> >What does RDF consider a namespace to be?
> >
> >It appears to me that the XML namespaces document makes XML namespace
> >simply be the set of URI references that share a common prefix.
> 
> Really?  I just checked, and I found nothing like that in the XML 
> namespaces doc
> 
>    http://www.w3.org/TR/REC-xml-names/#sec-intro
> 
> I did find:
> 
> [[
> [Definition:] An XML namespace is a collection of names, identified by a 
> URI reference [RFC2396], which are used in XML documents as element types 
> and attribute names.
> ]]

True, the connection to URI references is made by RDF, not XML.

> I found nothing that said that the names in an XML namespace were URI 
> references and a specific statement that an XML namespace is not a set:
> 
> [[
> XML namespaces differ from the "namespaces" conventionally used in 
> computing disciplines in that the XML version has internal structure and is 
> not, mathematically speaking, a set.
> ]]
> 
> >   Therefore
> >all XML namespaces contain an infinite and unchanging set of URI
> >references.
> 
> Your premise being false, this conclusion is not justified.  I found 
> nothing in xml namespaces doc that said an xml namespace could not be finite.

Yes, my understanding of XML namespaces was flawed.  I now see that they
are even further from the term ``namespace'' as it is used in some places in
the RDF documents.

> >However, Concepts says that ''Some terms in these namespaces have been
> >deprecated, some have been added, ...''  which appears to indicate that the
> >names in the namespace can be changed.  Does RDF actually use a different
> >meaning of a namespace than is used in XML?
> 
> Checking through the specs, and in the light of other responses to your 
> comment, I found some cases where the term 'namespace' is used where it 
> might be omitted or the term 'vocabulary' might be better.
> 
> I am recording this an issue against primer, concepts, syntax and schema (a 
> search of test cases and semantics for 'namespace' found no matches) to 
> review the use of the term 'namespace'.  

I found 5 matches for ``namespace'' in RDF Semantics.  I don't know if any
of them need any changes, but they are there.

These are recorded as:
> 
>    http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-17
>    http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-18
>    http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-19
>    http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-20
> 
> The WG will respond in due course.
> 
> Is that satisfactory for now?
> 
> Brian

Hmm.  For now, I suppose so, but this is in no sense a final resolution of
the issue.

Also, there is a general issue here, namely how should RDF use the term
``namespace'', that is unlikely to be satisfactorily addressed by
independant editorial action.  I believe that the RDF document editors are
trying to further harmonize the use of terms in the RDF documents - perhaps
a better disposition would be to expand this harmonization to include
harmonization with other W3C document and put ``namespace'' on a list of
terms to investigate.

peter

Received on Tuesday, 25 February 2003 05:37:36 UTC