W3C home > Mailing lists > Public > public-ws-chor@w3.org > May 2003

RE: Implementation of choreography interface

From: Stanislaw Ambroszkiewicz <sambrosz@ipipan.waw.pl>
Date: Mon, 12 May 2003 14:04:59 +0200 (CEST)
Message-Id: <200305121204.h4CC4xt03482@ns.ipipan.waw.pl>
To: public-ws-chor@w3.org, sanjay.patil@iona.com

Sanjay Patil wrote:
"I think it should be possible for someone else to implement 
 the client-side choreography interfaces. 
 It should also be possible to some one else to implement 
 the provider-side interfaces and readily interoperate with the clients. 
 IMO, these possibilities mainly define what interoperability 
 means for our working group." 

This looks very promising. If I understood this and the picture correctly, 
it is possible that the implementation of 
client-side and provider-side choreography interfaces 
may serve as a middleware between client and provider. 

Provider registers its application as a service to the 
middleware, actually, to the component that implements 
the provider-side choreography interface. 
In this case the component may serve for other applications 
providing service of the same type. 
This may be of particular interest because of 
failure of a single application. 

Client sends requests (tasks) to the middleware, actually, to 
the component implementing client-side choreography interface. 
It is clear that this component may serve for other clients as well. 

It looks nice at least IMO. However, IMHO there is a problem with 
interoperability if there is no uniform way (language) 
to express service types and clients' requests. 

I could be wrong.
Any criticism is welcome. 

Best regards,
Stanislaw
     http://www.ipipan.waw.pl/mas/
Received on Monday, 12 May 2003 08:05:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 18 December 2010 01:00:17 GMT