Collaborations and nesting (TP teleconf 2-28-01)

[James Bryce Clark <jbc@lawyer.com>]

For those of you who are following the CPA-BP collaboration dependencies, I
wanted to pass along these matters from the 2/28 TP conference call. 

This is one of two messages, one e-mail per topic.   Words in quotes below
are loosely rather than strictly employed.  Sorry, but I am trying to grope
towards simple understandings of complex things.  

Nesting collaborations
------------------------------

The conversations in Vancouver that generated the previously circulated
2/15/01 "Simple Negotiation Pattern" also gave us a chance to think more
about the following two assertions.   A week of discussing them over hotel
food and beer has increased our confidence level.  (I'm tentatively
confident it was the discussion, not the beer, that achieved this, but you
tell me.)  

(1)  Most multi-party e-business arrangements of current interest to us can
be logically built up from atomistic bilateral transactions, as long as we
can nest collaborations.  

      (Comment:  I was particularly pleased to see that Bill McCarthy's REA
constraints, particularly duality, are parallel to applicable legal
enforcement and interpretation constraints.  Both seem to validate the
foregoing conclusion.)

(2)  The necessary conditional relationships and transitions between
transactions are fairly simple, and can be included in some form in 1.0.  

     (Comment:   Just exactly how "simple" is still up for grabs, but we
seem likely to get some level of functionality included.)

Given (1) and (2), I wanted to confirm whether the CPA/CPP Specification
would have any problem with:

*  Infinite nesting of collaborations
*  A nested collaboration where all the CPAs are at the "low" bilateral
level, and no CPA directly associated with the "high" level composite
collaboration.  
*  A nested collaboration where the only CPAs is at the "high" composite
level, so there's no separate CPA associated with each "low" level binary
transaction, other than by reference to the "high" level one.

So I asked the TP group.  Marty's team was patient in talking this over
with me yesterday, and confirmed that they see no problems with the first
two, under their expected ver 1.0 spec.   The third may take some more
discussion.  
 
Obviously there are multi-party limitations, as ver 1.0 CPAs can only be
bilateral;  and there are some logical issues, such as the possibility of
overlapping CPAs in a composite collaboration.    Jamie

James Bryce Clark
Spolin Silverman & Cohen LLP 
310 586 2404    jbc@lawyer.com