I don't think that we need to say anything about
this at all. As Marty has indicated, control over
the conversation is within the application/process management
domain, not the MSH. If anything, this would be something
in the realm of the infamous Service Interface layer
and not part of this spec.
We're writing a protocol spec, not an implementation design
specification.
Cheers,
Chris
David Fischer wrote:
> I'm sorry, I was not clear enough.
>
> Conversations may not end, from the MSH perspective, because there is no
> notification to the MSH that a Conversation has ended -- at least we have not
> identified such a thing. If we would specify such a thing, the notification
> must be made to both ends of the Conversation.
>
> This sounds like new functionality.
>
> Regards,
>
> David.
>
>