W3C home > Mailing lists > Public > www-ws-desc@w3.org > September 2003

Re: On WSDL "operation"

From: Mark Baker <distobj@acm.org>
Date: Tue, 16 Sep 2003 11:35:35 -0400
To: Jim Webber <jim.webber@arjuna.com>
Cc: "Champion, Mike" <Mike.Champion@SoftwareAG-USA.com>, WS Description List <www-ws-desc@w3.org>
Message-ID: <20030916113535.T4344@www.markbaker.ca>

On Tue, Sep 16, 2003 at 09:55:36AM -0400, Jim Webber wrote:
> 
> Mike:
> 
> > Yeah, but a WSDL "operation" presumably DOES something.  It's 
> > not just a
> > message for the sake of sending bits around.  How about 
> > "serviceInvocation"
> > or "serviceRequest" or something along those lines? If a WSDL 
> > message does
> > anything, it does request that a service be performed by the 
> > provider on
> > behalf of the requester [the current WSA terminology, IIRC].  
> 
> As you point out, it presumably does something (though we can't be sure
> about that). However the fact that a service probably does something with
> that message is not important at this level. What is important is describing
> the messages going in and coming out and describing that. Trying to apply
> "invocation" semantics to this violates encapsulation, and encourages
> developers to view services as invokable objects rather than entities which
> merely exchange messages.

I don't think "objects" is implied, but invokable interfaces, sure,
'cause that what all this SOA stuff is about, no?

Otherwise you're just talking bit transport.

What's important from a WSDL POV is the application semantics; what
does it *mean* for those bits to get there, and what does the sender
of those bits know if a successful response is returned, for example.

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
Received on Tuesday, 16 September 2003 11:34:32 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:26 GMT