- From: Savas Parastatidis <Savas.Parastatidis@newcastle.ac.uk>
- Date: Tue, 9 Sep 2003 21:03:11 +0100
- To: "Cutler, Roger (RogerCutler)" <RogerCutler@chevrontexaco.com>, "Martin Chapman" <martin.chapman@oracle.com>, "He, Hao" <Hao.He@thomson.com.au>
- Cc: <www-ws-arch@w3.org>, "Jim Webber" <jim.webber@arjuna.com>
> > Seems to me that it would be friendly toward the reader if one explained > this pretty clearly up front. If you just say that SOA services are > stateless without explaining how one then layers stateful interactions > on top that it might be a bit confusing. > I totally agree that a reader should be told that stateful interactions are very much part of the Web Services Architecture _but_ that this is achieved through mechanisms sitting on top of services and not because of those services' semantics. Stateful or contextualised interactions should be explained and defined at the same level as coordination, transactions, choreography, security, etc. It is the case that some of these concepts can be built or require contextualised interactions, as it is the case, for example, with WS-Coordination, WS-Transaction, WSBPEL, etc. .savas.
Received on Tuesday, 9 September 2003 16:05:41 UTC