Minutes of CPPA Negotiation Conference Call June 12, 2002

Attendees:  Marty Sachs, Neelakantan Kartha, Monica Martin, Brian Hayes

 

Brian reviewed the status of the BPSS instance document for negotiation.  He tried to set up the model so that the same transaction is reused and both parties have the same role.  The biggest problem was reuse of the transaction. He discovered problems, especially, on how to put in the preconditions on the transactions. He prepared notes on what changes are needed in the BPSS specification and started a discussion in the BPSS team.  The specific problem is that the current BPSS specification allows reasoning only about documents.  We must be able also to reason about transactions.  An example is that when a party issues an offer, it can’t start another transaction until the other party has responded.  In other words, a precondition on the transaction has to be that the initiator of the last transaction can’t initiate another one.

 

Brian said that BPSS needs the following:

 

·        Ability to point to a business information object (the schema for the previous transaction

·        Ability to point to the current transaction

 

He is trying to get these changes into BPSS version 1.05 during the next two weeks so they will be in the approved version 2 specification.  Once the BPSS changes are done, the BPSS instance document for negotiation can be completed in a couple of weeks.

 

Kartha asked if the negotiation algorithm will be described by a BPSS instance document. Marty replied that the BPSS instance describes only choreography.  The negotiation algorithms are really at the application (business process) level. Brian said that the UMM provides support for use cases for the collaborative business process but we also have to define the enterprise business process (the private process at each party).  The UMM has some support for the enterprise business process but doesn’t go into much detail.  The designer of the enterpriser business process has to do some OO analysis based on the UMM use cases.  We can think of the enterprise business process as a UMM model from one party’s perspective.

 

We reviewed the element-attribute spreadsheet that Dale Moberg distributed today. After filling in a few items as examples, we, at Kartha’s suggestion, decided to partition the work as discussed at the Reston F2F. We agreed to limit the discussion on the conference calls to “controversial” items.  We divided the rows that have nothing in the negotiation range column as follows:

 

105-150:  Kartha

151-196:  Marty

197-250:  Dale

 

It was suggested that questions on the packaging and security rows be referred to Dale.

 

We decided to fill in our own private copies of the spreadsheet and worry about merging them later.  We will each try to get half of our assignment done for next week’s call.

 

 

Respectfully submitted,

Marty Sachs