W3C home > Mailing lists > Public > www-ws-arch@w3.org > August 2003

RE: Issue: Synch/Asynch Web services

From: Ugo Corda <UCorda@SeeBeyond.com>
Date: Mon, 4 Aug 2003 13:10:34 -0700
Message-ID: <EDDE2977F3D216428E903370E3EBDDC9081269@MAIL01.stc.com>
To: <jdart@tibco.com>
Cc: <www-ws-arch@w3.org>

My point was that originally the WSD group seemed to at least mention the sunc/async distinction at the MEP level (In-Out vs. Request-Response). So if I see a WSDL interface that defines an In-Out MEP I have the expectation that the service will behave asynchronously (otherwise the Request-Response MEP would have been used) even though, as you say, the WSDL does not define any correlation mechanism.

But if Request-Response is dropped from the spec, then evidently In-Out must work for both synchronous and asynchronous behaviors, so the WSDL interface gives me no clue of what the service expects.

Ugo

> -----Original Message-----
> From: Jon Dart [mailto:jdart@tibco.com]
> Sent: Monday, August 04, 2003 12:55 PM
> To: Ugo Corda
> Cc: www-ws-arch@w3.org
> Subject: Re: Issue: Synch/Asynch Web services
> 
> 
> Actually "same channel" is mentioned in the WSDL Message 
> Patterns Draft 
> only in the context of request-response (where it makes 
> sense). That is 
> a means of correlation, but it is generally N/A for asynchronous 
> interactions.
> 
> --Jon
> 
> Ugo Corda wrote:
> >>The problem is that WSDL MEPs are not connected with any concept of 
> >>correlation, even an abstract one.
> > 
> > 
> > Well, the latest WSDL Message Patterns draft mentions a 
> transport-level correlation mechanism (that's what "the same 
> channel" seems to mean). Are you saying that the WSD group 
> has decided not to touch that area at all (which could 
> explain dropping the Request-Response MEP in favor of just In-Out)?
> > 
> > That might be a proper decision for the WSD group, 
> particularly in light of the time pressure it is under. On 
> the other hand, if that group does not address it, other 
> groups will, as you say (in addition to the ones you mention, 
> BPEL has also its own way of dealing with it). The problem 
> is, as usual, that those groups might introduce different 
> concepts and mechanisms that, in the end, will not easily 
> interoperate.
> > 
> > Ugo
> >  
> > 
> > 
> 
> 
> 
> 
Received on Monday, 4 August 2003 16:10:41 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:22 GMT