- From: Hugo Haas <hugo@w3.org>
- Date: Wed, 2 Feb 2005 09:12:27 +0100
- To: Greg Truty <gtruty@us.ibm.com>
- Cc: public-ws-async-tf@w3.org
- Message-ID: <20050202081227.GB14233@w3.org>
Hi Greg. * Greg Truty <gtruty@us.ibm.com> [2005-01-28 10:45-0700] > I had taken the todo of putting together some thoughts on the 3 forms of > wire level messages: > > 1) one-way > 2) bindings for req/resp (async) > 3) bindings for req/resp (async w/different transports). > > I put it in the form of a presentation (to make it easier to walk > through). There are speaker notes (and/or clouds w/text), to try and get > various points across. Nice analysis. Reading it, I realized that, I think that you can't do scenario #3 with a single interface with WSDL 2.0, as the underlying protocol is expressed on the binding component. Could you please expand on your comments in slide 10 about whether it makes sense from the POV of the requester? I'm having issues to see why it wouldn't, although you need to see the binding to have the whole picture. Putting the WSDL 2.0 issue aside, for your scenarios to come to life, it seems that the main issues are one-way SOAP MEP & binding. In slide 6, are you saying that SOAP 1.1 (WS-I enhanced) has such a binding? Cheers, Hugo -- Hugo Haas - W3C mailto:hugo@w3.org - http://www.w3.org/People/Hugo/
Received on Wednesday, 2 February 2005 08:12:28 UTC