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

Re: Summary of the QName to URI Mapping Problem

From: Piotr Kaminski <pkaminsk@home.com>
Date: Fri, 17 Aug 2001 10:16:24 -0700
Message-ID: <027f01c12741$18201dc0$35254d18@gv.shawcable.net>
To: "Narahari, Sateesh" <Sateesh_Narahari@jdedwards.com>, <www-rdf-interest@w3.org>
Cc: <SCranefield@infoscience.otago.ac.nz>
Sateesh Narahari wrote:
> I am confused guys. When exactly did a URI become a Unique Resource
> Identifier from being a Uniform Resource Identifier?.

I don't think anybody's claiming that URIs are unique, in that a resource
might be identified by more than one URI.  There are, of course,
advantages to using only one URI per resource, since that makes it trivial
to distinguish when two URIs identify the same resource.  However, I don't
think anybody's debating this right now.  :-)

> If there is no requirement for being unique for a derived URI, whats big
> deal if two QNames derive same URI?.

I think you're confusing uniqueness with distinction.  Think of the
mapping as a relation between QNames and sets of URIs.  Uniqueness
requires that each URI destination set be of size 1.  Distinction requires
that all the sets must be disjoint.  They're quite independent
requirements.  (If I was more awake right now, I'd put this in more
mathematical terms...)

Uniqueness is nice, but its lack won't produce incorrect results -- at
worst, it'll fail to produce some useful results (because we can't match
resources as being identical).  Distinction is critical; if you don't have
that property, you can derive incorrect conclusions based on the mistaken
assumption that two different QNames are the same.

        -- P.

--
  Piotr Kaminski <piotr@ideanest.com>  http://www.ideanest.com/
  "It's the heart afraid of breaking that never learns to dance."
Received on Friday, 17 August 2001 13:18:13 GMT

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