Re: multiparty (was) Fwd: Re: message routing

[bhaugen <linkage@interaccess.com>]

From: Stefano POGLIANI 

> I do not know if I am right, nor I am qualified to validate
> a mathematical model of this kind. But...

I don't have a mathematical model yet either (John Yunker might), 
but below I will try to explain briefly why I think this is important.

> ...my feeling is that trying to reconcile multi-party collaborations
> into a series/coordination/pattern-driven set of two-party
> collaboration is an academic effort at least from a modelling
> point of view.

This is not an academic effort for me.  My goal is to go as far
as I can to making business collaborations, including multi-party
ones, declarative instead of procedural.  In other words, so they
can be defined in web forms instead of UML or workflow diagrams.
I think that dialogs are simpler than multi-way conversations and
more suitable to declarative-ness.

> In some way, by applying some rigid practice I can "approach"
> OO-style of programming in any language; but is this "natural"?
> I mean, does this effort pay back?

There is a stack of natural dialogs in business:
e.g. offer-acceptance (contract formation), 
commitment-fulfillment (order-delivery), etc.

To some large extent I think that even multi-party collaborations
(as in the travel example) will fall naturally into a set of coordinated
two-party dialogs.

I am interested in pursuing this idea to the extent that it works,
but I would not apply it as a rigid practice or even try to
decompose anything automatically.

I'm approaching this more as a pattern language, in the sense
of Christopher Alexander and the software design patterns
community.

> I am looking
> for a way to express "simply" some concepts

Me too.

Regards,
Bob Haugen