[
Lists Home |
Date Index |
Thread Index
]
(snipping the CC list for mercy's sake)
The problem with this question is in agreeing on a
representation for any thing in the set of all things.
We can devise representations of a black hole,
the singularity, but that is as far as one can go
because the physical entity never returns an event passed to it that
enables one to verify that representation. That is
the limit of representative state. We know where it is,
and we know some properties of the space around it, and
it radiates, but internally, we have to guess. We
can only address the location; that is why it is a singularity.
The only use we can logically make of the information
we have is to avoid it (transbounded state engines aside:
I'm not folding information space today without a lot
more coffee).
It is an absurd case but we are talking about system
boundaries. In many of the papers from the sources
you cite, the word "useful" appears. A "useful representation"
is what is wanted. One could say without error, I believe,
the Web system provides a means to identify and access
useful representations. Past that, we are back to abstractions.
We can go far too deep into epistemology with this thread.
All we have to do is agree that identifiable representations
bound the things that are on the web (identifiable because
of the serial issue Jonathan points out: by definition,
state is a set even if it has a single member). Then it is an issue
of implementations of means. WSDL is on the Web. That
makes SOAP a web technology. It is possible and I think
useful to bound that that way and to warranty the transaction
only to the endpoint. Again, what they choose to hide behind
that endpoint is their responsibility and the due dilligence
of the user/customer. This is why other circles have
CONOPS as part of the contract of use.
len
-----Original Message-----
From: Mark Baker [mailto:distobj@acm.org]
On Fri, Apr 26, 2002 at 12:37:24PM -0500, Bullard, Claude L (Len) wrote:
> All things representable can be on the web. That is key
> and bounds the set.
Agreed. But do you not agree that all things that are identifiable are
representable?
TimBL and DanC use the phrase "Identity, State, and GET";
http://www.w3.org/DesignIssues/Axioms.html#state
http://www.w3.org/Talks/2002/02/27-wa/slide9-0.html
|
