RE: What WSDL defines - the diagram!

At 01:43 PM 11/11/2003 +0000, Savas Parastatidis wrote:
>. . .
>
>A consumer (the client in your diagram) may interact with many services.

Right, and the service may interact with many clients.

>Suggesting that a WSDL document defines the behaviour of such a consumer
>may give the wrong impression to the community.

How so?  A WSDL document *does* partially define the behavior of both 
parties.  This is the reality.  I think it would be misleading to suggest 
otherwise.

Of course, it is important to understand that the WSDL document only 
PARTIALLY defines the behavior.  Maybe this is what you were trying to 
suggest.  Both the client and the service may have behaviors that are 
beyond what the WSDL specifies -- including the ability to send and receive 
messages that are not described in the WSDL.  This may not be obvious at 
first glance, but it is (again) fundamental to the meaning of a WSDL 
document.  It is essentially an "open world" assumption: the WSDL document 
does not define EVERYTHING about either the service or its clients.  This 
idea was explained previously in a discussion[1] about the recommendations 
of the patterns task force[2].

>A WSDL document defines the contract to which a service is prepared to
>adhere. If a consumer agrees to the contract, then it has to respect it
>through its participation to the defined message exchanges and through
>the use of correct syntax/format for the documents being exchanged (the
>messages). So, we talk about a consumer's participation in the
>interaction patterns with respect to a particular contract. (If this is
>what you were suggesting, then we are in agreement).

Yes.

>  . . . .

1. http://lists.w3.org/Archives/Public/www-ws-desc/2003Sep/0062.html

2. http://tinyurl.com/upbs , which forwards to
http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/wsdl12/meps-vs-iops/recommendations_clean.htm


-- 
David Booth
W3C Fellow / Hewlett-Packard
Telephone: +1.617.253.1273

Received on Wednesday, 12 November 2003 10:28:18 UTC