- From: Assaf Arkin <arkin@intalio.com>
- Date: Sat, 15 Mar 2003 13:43:16 -0800
- To: "Geoff Arnold" <Geoff.Arnold@Sun.COM>, <www-ws-arch@w3.org>
> -----Original Message----- > From: www-ws-arch-request@w3.org [mailto:www-ws-arch-request@w3.org]On > Behalf Of Geoff Arnold > Sent: Friday, March 14, 2003 11:55 PM > To: www-ws-arch@w3.org > Subject: Re: Friendly amendment #2c [Re: Straw poll on "synchronous" > definitions] > > > > Two quick questions: > > (1) Do people feel that we're converging on language which > addresses both two-party and multi-party interactions? > If not, does that matter? I feel there's a bit too much emphasis on two-party interactions. I can see the general interest if that's all of what people are doing right now. But if multi-party becomes more interesting down the road (I suspect that when tools advance it would), would we have to junk everything we're doing right now and do it all over again? Or can we talk about multi-party with two-party being a specific and more commonly used case? > > (2) Are we confident that our definition is robust > enough to be adopted by the choreography folks? I think it's a good definition as far as choreography of request/response patterns goes. But it still does not answer two important questions (in my opinion): a) can we say that an operation is synchronous or asynchronous based on it's interface independent of the protocol we use? can we use that to apply different semantics (e.g. transactions, queuing, protocol, features)? b) can we say that something that is not a simple request/response pattern (WSDL or choreography) is synchronous or asynchronous independent of the protocol? can we propose a generalized rule that covers more ground?
Received on Saturday, 15 March 2003 16:43:47 UTC