W3C home > Mailing lists > Public > www-tag@w3.org > July 2002

Re: resource equivalence was: Re: fragment identifiers

From: Michael Mealling <michael@neonym.net>
Date: Wed, 24 Jul 2002 11:39:13 -0400
To: Jonathan Borden <jonathan@openhealth.org>
Cc: Michael Mealling <michael@neonym.net>, www-tag@w3.org
Message-ID: <20020724113913.B1825@bailey.dscga.com>

On Wed, Jul 24, 2002 at 11:16:34AM -0400, Jonathan Borden wrote:
> Michael Mealing wrote:
> [[
> On Wed, Jul 24, 2002 at 09:40:23AM -0400, Jonathan Borden wrote:
> ...
> >
> > BTW: I've suggested equating resources by the equality of their
> > representation sets as axioms [10,11] in
> > http://www.openhealth.org/RDDL/SchemaAlgebra
> >
> > [[
> > [10] equivalent(URIa,URIb) := Entities(URIa) = Entities(URIb) and
> > cardinality(Entities(URIa)) > 0
> >
> > Two URIs are equivalent when they map to the same set of entities.
> >
> > [11] equivalent(A,B) <=> exists URIa such that A = resource(URIa) and
> exists
> > URIb such that B = resource(URIb) and equivalent(URIa,URIb)
> >
> > Two resources a and b are equivalent if the set of entities given the URIa
> > and URIb are equal where URIa identifies a and URIb identifies b.
> > ]]
> >
> > where Entities(URI) is the set of entities which the URI maps to across
> > media-types, time, and other e.g. HTTP request parameters.
> 
> Wrong. Two URIs are equivalent when they match each other according to
> the canonicalization rules in RFC 2396.
> ]]
> 
> The terms "equal" and "equivalent" are subtly different, and I did not use
> the phrase "URI equivalence" randomly, nor did I intend it to be conflated
> with "URI equality"
>
> There is often more than one way to compare things, e.g.
> http://lists.w3.org/Archives/Public/www-archive/2001Nov/0044.html

Sure. But we have to be very careful to state that equivalence other
than syntactic URI equivalence is system dependent. You're statements
implicetly made the suggestion that they were universal and they aren't.

> [[
> The only uniform statement you can make about the equality of two or more
> Resources is whether or not the URIs in question are syntactically
> equal according to RFC 2396's normalization rules. Period.
> ]]
> 
> Yes, well this definition of "URI _equality_" appears a priori true given a
> 1:1 relationship between URI and resource. I am intending to capture the
> situation where one may desire a number of URIs to refer to the same
> resource, and so while the resources identified by such a set of URIs are
> not exactly _equal_ we can say they are _equivalent_ in some sense of this
> term.

But please be careful to qualify you're statements of 'equivalence' so that
it is clear that it is only _your_ definition that you are using. There
is no uniform or universal concept of equivalence for URIs and Resources
other than what RFC 2396 specifies.

> If there were no need to consider such issues the DAML/OWL terms:
> isEquivalentTo, sameClassAs, samePropertyAs wouldn't be necessary, but they
> are actually quite useful when trying to have a logical description of
> various things.

I didn't say they weren't necessary or useful. Just that they are
system specific. The semantics associated with DAML/OWL's isEquivalenceTo
is scoped to that specification only....

-MM

-- 
--------------------------------------------------------------------------------
Michael Mealling	|      Vote Libertarian!       | urn:pin:1
michael@neonym.net      |                              | http://www.neonym.net
Received on Wednesday, 24 July 2002 11:41:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:10 GMT