Re: [Fwd: RE: "information resource"]


Hey again,

On Mon, Oct 18, 2004 at 03:00:14PM +0100, Williams, Stuart (HP Labs, Bristol) wrote:
> > > >Consider that if I'm provided two URIs, one which identifies "an
> XML 
> > > >representation of my medical record at time T" (where T is fixed)
> and 
> > > >the other which identifies "an XML representation of my current 
> > > >medical record", then if I invoke GET on each at time T, I get two 
> > > >equivalent representations back.
> > > >I suggest that what makes those two resources different is an 
> > > >"essential characteristic", and because it isn't reflected in the 
> > > >messages,
> > > >
> > > 
> > > Small point... but that would depend on whether you regarded the 
> > > request URI (or URI returned in header fields) part of any of the 
> > > messages exchanged.
> > 
> > I don't believe so.  The response messages would certainly be 
> > different if the URIs were included, but not different in a 
> > way which licenses the recipient to determine whether the 
> > essential characteristic I'm describing is present or not.
> You stated that a distinguishing "essential charateristic" isn't
> reflected in the messsages. I was remarking that the distinguishing
> characteristic is 'reflected' in the use of (at least) different request
> URI... (the request message being one of the messages exchanged to
> obtain a response). 

We must be talking past one another...

I'm talking about the "essential characterisitic" (per the Basel
definition of "information resource") of a commitment to reflect the
future states of a resource via its representations.  That commitment,
or lack of it, isn't reflected in any messages even though the
messages may differ.

I think if there's any argument to be had here, it would be with my
assertion that this commitment is an "essential characterisitic" of a

Mark Baker.   Ottawa, Ontario, CANADA.

Received on Monday, 18 October 2004 14:27:16 UTC