Re: [ebxml-bp] BPSS evolution

["David RR Webber" <david@drrw.info>]

Title: Message

Martin,

 

Definately want to discuss this today.  My immediate thought is that the BCM work

here could be a perfect fit.  Especially since BCM has clearly detailed how UMM

fits into the overall production stack and the role it provides.

 

I don't want to rush into anything though - so perhaps it makes sense to go back

to the BCM group and ask for a presentation on how BCM model synergizes the

BPSS - and also what new pieces this brings to the table - such as the BCM

Choice Points for context driven linking and switching control.

 

We can refine the needs I'm sure and then see what the BCM can provide

back to us.  I'm trying not to just throw this ball to myself here!  The BCM team

has a great mix of people who deeply understand BP at the architectural level

and we definately need all that great input.

 

Thanks, DW.

----- Original Message -----

From: martin.me.roberts@bt.com

To: jeanjadu@Attachmate.com ; ebxml-bp@lists.oasis-open.org

Sent: Monday, November 03, 2003 6:05 AM

Subject: RE: [ebxml-bp] BPSS evolution

Dear all,

    The questions JJ asks indicate to me that the move to OASIS has removed one of the key conceptual models behind the BPPSS namely the UMM model.  Before deciding on the XML content for the OASIS BPSS there will be a need to understand the concepts that need to be implemented.  UMM and UML offer some good features, but removing the link will undoubtedly give the team more degrees of freedom.  Is this a good thing?  I know from my experience that not having constraints usually wastes time and energy as those boundaries are defined.

 

    So it looks like this group had better go on concept hunt!

 

 

Martin Roberts
xml designer,
BT Exact
e-mail: martin.me.roberts@bt.com
tel: +44(0) 1473 609785  clickdial
fax: +44(0) 1473 609834
Intranet Site :http://twiki.btlabs.bt.co.uk/twiki

-----Original Message-----
From: Jean-Jacques Dubray [mailto:jeanjadu@Attachmate.com]
Sent: 03 November 2003 04:26
To: ebxml-bp@lists.oasis-open.org
Subject: [ebxml-bp] BPSS evolution

hi:

 

here is a list of possible directions to evolve ebXML BPSS:

 

1. Control flow: 1.01 / 1.1 BPSS control is based on modified UML  activity diagram semantics. This is a bit limiting and UML is changing these semantics in UML 2.0

    a. we could liaise with ws-chor and see if we could reference a ws-chor definition for a business transaction, binary and multiparty collaboration definitions.

    b. we could look at the work of Will Van Der Aalst if liaison with ws-chor is not possible

2. Business Transactions: 1.01 / 1.1 BPSS BT Message Exchange Pattern is not general enough

    We could extend the definition of BT (or other concepts) via different MEPs

3. Business Transaction Protocol: 1.01/1.1 Embedded in the BPSS spec

    a. we could liaise with ws-caf or other relevant BTP groups and specify their usage in the context of BPSS rather than

4. Specify a WSDL binding for the BSI: 1.01/1.1 currently no work has been done in this direction.

5. Allow for a mixed mode: business transaction / web service within a collaboration definition. 

6. Orchestration / Choreography / Collaboration architecture and bindings

    Today there is still a lot of confusion on these concepts. We need to provide a better picture on how these concepts play together in a common architecture.

 

7. I have a few more but I would like for other to propose them

 

It would be good also if we could surface the issue of web services alignement to all ebXML working groups.

 

Best regards,

Jean-Jacques Dubray
attachmate
3617 131st Ave
Bellevue, WA 98006
tel: 425-649-6584
Cell: 508-333-7634