- From: Williams, Stuart <skw@hplb.hpl.hp.com>
- Date: Wed, 24 Oct 2001 13:21:46 +0100
- To: "'Marwan Sabbouh'" <ms@mitre.org>, "'Kumeda'" <kumeda@atc.yamatake.co.jp>
- Cc: "'xml-dist-app@w3.org'" <xml-dist-app@w3.org>
Yasuo, Marwan,
I do intend to reply to both your messages on this thread. I'm on a course
this week, so finding moments to puts some focussed time into responding is
problematic. Also, I shall be on vacation from Friday until after the
following weekend.
I think that there is much that we agree on. BTW, in terms of defining the
services that SOAP provides to it's users, that's a place that the Abstract
Model (http://www.w3.org/TR/xmlp-am.html) was going. It was (is) not to
everybodies taste... and we will need to do some maintainence to it in due
course.
One of the other differences of viewpoint that exists across the WG might be
characterised (hopefully neutrally) as follows:
1) The Chamelon View (aka 'bottom-up'): Basically, SOAP takes on the
character of the thing that it is bound to. So 'Contract 2' and 'Contract 4'
are very similar, and certainly 'Contract 2' varies with 'Contract 4'.
2) The Platform View (aka 'top-down'): SOAP defines a service abstraction at
'Contract 2' that is largely 'invariant' across different underlying
protocols ('Contract 4's). The SOAP layer works to present a consistent
abstraction of its services to the application entities above it.
I think that we have folks trying to pull/push us to both of these
viewpoints and I for one can only hold one at a time. I'm not sure that a
view that accomodates both exists and we have never driven it 'hard' to on
or the other.
FWIW there are also folks who would not subscribe the the view that even in
the abstract there is a boundary at 'Contract 2'. They see the Application
and the SOAP layer as somewhat monolithic.
I'll respond properly when I have more time... hopefully tomorrow.
Best regards
Stuart
Received on Wednesday, 24 October 2001 08:24:16 UTC