Lists Home |
Date Index |
- From: David Brownell <email@example.com>
- To: Paul Prescod <firstname.lastname@example.org>
- Date: Mon, 21 Jun 1999 15:46:36 -0700
Paul Prescod wrote:
> David Brownell wrote:
> > Yes. I'll commend you to some archives at the OMG for this topic;
> > I recall something like a four year ongoing discussion about the
> > notion of identity, which concluded that there are enough different
> > notions that picking one to use everywhere was wholly impractical.
> I find this hard to believe. Are you saying that in the OMG universe I can
> have two references to an object and there is *no way* of asking if they
> refer to the same object?
There's a loose notion of "equivalence"; you can ask if
an object reference is equivalent to another. Primarily
that's a syntactical operation designed to let you create
tables keyed by object reference, without duplication; it's
explicitly understood that the objects two references refer
to may in fact be "the same" without having the refences be
"equivalent". (Some replication schemes work like that, as
do many proxying and load balancing schemes ... large systems
use such techniques all over the place.)
Generally, identity is an application level notion and not
part of the infrastructure. Servers can provide such notions
for the objects they implement, as can collections of servers,
but it all requires pre-arrangement and is of limited scope.
> > (So if folk on this list are confused by the notion of "identity",
> > it's understandable and you're in fine company. It's not at all a
> > simple notion.)
> I'm not so concerned about the folks on this list as the folks who are
> writing specs. They write as if the concept of identity were well-defined
> in the Web context (see, e.g. RDF and XLink) when clearly it is not.
I've queued up that WCA-Terms URL for later consumption ... could
be I'll agree with you on that!
xml-dev: A list for W3C XML Developers. To post, mailto:email@example.com
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:firstname.lastname@example.org the following message;
To subscribe to the digests, mailto:email@example.com the following message;
List coordinator, Henry Rzepa (mailto:firstname.lastname@example.org)