W3C home > Mailing lists > Public > www-rdf-logic@w3.org > April 2001

RE: Dirt simple RDF was: Re: A plea for peace.

From: Jonathan Borden <jborden@mediaone.net>
Date: Thu, 5 Apr 2001 23:09:23 -0400
To: "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>
Cc: <www-rdf-logic@w3.org>
Message-ID: <000c01c0be46$fb4fe800$0201a8c0@ne.mediaone.net>
Peter F. Patel-Schneider wrote:
>
> > Why is string ordering needed (for the moment)?
>
> It is not, it is just that the RDB model includes the string ordering,
> which is not valid for URIs, and thus the two models do not line up.
>

The XML Namespaces recommendation directs namespace names (which are URIs -
properly 'URI references') to be compared as literal strings. That's one
option, the other is to compare canonicalized URIs. Either way URIs can
essentially be treated as strings for string comparison purposes.

Providing string comparison/ordering does introduce the useful property of
defining a collection as set of triples, ordered collections being ordered
sets of triples:

Canonical order could be defined as alphabetic order by subject, then
predicate then object. Collections may be defined as sets bounded by string
order e.g.

all triples having the subject http://example.org/foo
all triples having the subject http://example.org/foo and the predicate
http://example.org/memberOf

a sequence can be defined by canonical order where the predicate URI order
increases over the sequence.

-Jonathan
Received on Thursday, 5 April 2001 23:08:47 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:38 GMT