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

Re: Summary of the QName to URI Mapping Problem

From: Sean B. Palmer <sean@mysterylights.com>
Date: Fri, 17 Aug 2001 16:11:02 +0100
Message-ID: <002201c1273d$2b571460$d4d993c3@Palmer>
To: <Patrick.Stickler@nokia.com>, <connolly@w3.org>, <drew.mcdermott@yale.edu>
Cc: <www-rdf-interest@w3.org>
[-cc RDF Logic to placate DanBri!]

> I.e. if 'ns1:' = "urn:x:abc" and 'ns2:' = "urn:x:abcd"
>      then both 'ns1:defg' and 'ns2:efg' are mapped to
>      the same URI "urn:x:abcdefg"! Yet these are clearly
>      separate resources per their disjunct QName identities

This is not the case in RDF, because it doesn't have a notion of QNames,
only URIs. In XML-other-than-RDF, yes, the QNames in your example above are
treated as separate "things", but in RDF they resolve to the same URI.
That's no problem because it's in the RDF specification, and every RDF tool
uses the QNames to derive URIs.

The "problem" is that therefore the original QNames are "impossible" to
address, because RDF applications always try to resolve the QNames to URIs.
The answer is to just create a model for them. I already provided an
example of how to do this, and yet you seem to be dissatisfied for some
reason, and I still haven't been able to gague why that is.

Let's say that the controllers of the XHTML namespace wanted to assert that
people should use the URI:-

   http://www.w3.org/1999/xhtml#title

to refer to the ExpEType QName:-

   {http://www.w3.org/1999/xhtml}title

they don't have to choose the URI with the added #, but it seems logical.
All they have to do is to issue some RDF, preferably at the namespace
location in the first place, and/or the URI that are being used to
represent the QNames, such that:-

<http://www.w3.org/1999/xhtml#title> a :QName;
   :ns <http://www.w3.org/1999/xhtml>;
   :ExpEType "title" .

This complies to the RDF and XML Names specifications, doesn't break any
current implementations, and doesn't use anonymous nodes. It solves your
problem, and reduces the confusion to nil. I can't see how you can continue
to believe and state that this is a "problem". It's an annoyance, but there
is nothing particularly broken. RDF can't refer to QNames in syntax
(there's lots of things that it can't do in the syntax...), but it can in
the model. Simple :-)

--
Kindest Regards,
Sean B. Palmer
@prefix : <http://webns.net/roughterms/> .
:Sean :hasHomepage <http://purl.org/net/sbp/> .
Received on Friday, 17 August 2001 12:53:00 GMT

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