RE: Slight mod to service model

 

> -----Original Message-----
> From: Cutler, Roger (RogerCutler) 
> [mailto:RogerCutler@chevrontexaco.com] 
> Sent: Monday, January 12, 2004 7:11 PM
> To: Richard.Chennault@kp.org; fgm@fla.fujitsu.com
> Cc: www-ws-arch@w3.org
> Subject: RE: Slight mod to service model
> 
> 
> Fair enough.  And you said in your other message, "My 
> questions were meant to enlighten myself and my own endeavors 
> and should not be construed as an attempt to alter the course 
> of the document".

Right. On the other hand, this question is at the very core of the WSA and
the question that *I* get asked constantly, i.e. "what's the relationship
between WS and SOA".  So, we have to have a good answer for how the
definition of a "service" relates to the definition of a "message", and we
have to have a clear picture -- if we are to declare victory.

Informally, a "service" is invoked by one or more messages, but the
definition of the service must talk about the semantics of what the service
does as well as the syntax of the message.   Also, the implementation of the
service may entail a number of other messages to
orchestrate/choreograph/compose/whatever the service out of others.  That
needs to be reflected somehow.

This is exactly what we are discussing vis a vis the SOM diagrams, the
definition of SOA vs DistObj, the word we are going to use instead of
"orchestrate/choreograph/compose/whatever", and so on.

Received on Monday, 12 January 2004 19:23:02 UTC