Re: eBTWG BPSS Project Proposal

[Maarten Steen <maarten.steen@telin.nl>]

David,

Thanks for this clarification. I'm fairly new to the ebXML scene, so I don't yet share the same collective memory. I'm doing my best to catch up.

I still don't understand though why transactions are not good enough to communicate business commitment. For me a transaction is an atomic interaction between two parties that results in a common understanding of the state in the shared business process. Transactions can therefore be used to
communicate commitments, like "by placing this order I promiss to pay X amount, and you promiss to deliver Y goods within Z days." If you don't want to rely on documents or messages being exchanged in transactions, then how else do you communicate commitment?

Cheers,
Maarten




"Welsh, David" wrote:

> >
> > Maarten Steen
> > > Sometimes, I get the feeling there is a hidden
> > > agenda in some of the project proposals (like the one on monitored
> > committments) to come
> > > up with a completely new language for business process analysis.
> >
> > Not really.  The monitored commitments project wants to move
> > in the direction of enabling business users to define
> > business processes
> > declaratively.  The declarations will need to map to one or more
> > implementation
> > languages, via predefined analysis components (for example,
> > commitment-
> > fulfillment relationships, which is the example of the idea
> > that the project
> > will focus on).
> >
>
> Maarten,
> The notion of 'commitment' and the concept of 'monitored commitments' really came up in the original ebXML BP work, atleast a  good half a year ago. (and other work such as Open-edi talks about 'commitment' too).
>
> 'commitment' is the desired business effect I think we're shooting to have from a technical B2B relationship, but the formalized structure of the context of a business commitment if it's only centered on a document payload falls far short of what's needed.
>
> As far as I can see, 'monitored commitments' is one of the main payback reasons why I'm willing to take what limited business time I have to formally model a business process; ie. so I can directly run the business process model against the actual business operation - and stop doing the same thing
> manually !
>
> BPSS started looking at things like REA, but due to lack of time (to complete the level of necessary detailed work) and an un-natural amount of design churn towards the end of the project, it was dropped from that final deliverable.
>
> This is really a continuation of that concept and work.
>
> -Dave
>
> > The idea is for business people to be able to say "I want to
> > use the XYZ
> > order-fulfillment pattern" with maybe a little customization,
> > rather than
> > needing to muck about with transitions, forks, joins, etc.
> > The patterns
> > could be implemented in different languages, as in current software
> > design patterns.
> >
> > Clear enough?  There are examples of patterns at the business
> > transaction
> > level in the UN/CEFACT UMM documents, for example, offer-acceptance.
> > We're just taking the idea up a level to longer business
> > collaborations.
> >
> > -Bob Haugen
> >
> >
> > ----------------------------------------------------------------
> > To subscribe or unsubscribe from this elist use the subscription
> > manager: <http://lists.ebtwg.org/ob/adm.pl>
> >

--
Dr. ir. Maarten W.A. Steen - Scientific Researcher
Telematica Instituut, Postbus 589, 7500 AN  Enschede, The Netherlands
http://www.telin.nl/
phone: +31(0)53 4850 321
fax: +31(0)53 4850 400