Bob,
I agree with your observations.
I will put a TimeToPerform at the BinaryBusinessCollaboration which is the
equivalent of the BusinessCollaborationProtocol. The REA based constraints
will be addressed in the next release.
-karsten
><Karsten Riemer>
>We have decided that none of the timing and security parameters go above the
>transaction level, i.e. we have decided not to have any timing and security
>defaults at the collaboration level.
></Karsten Riemer>
>
>As Dave Welsh says, I think there are needs for time constraints
>at the business collaboration level.
>
>In the metamodel, there is a class on the BOM level called
>"BusinessCollaborationTask" which contains a timeToPerform
>parameter. I think it would be good to reified that parameter
>down to the BusinessCollaborationProtocol.
>
>Other internal collaboration constraints could be modeled
>(correctly, I think) with Commitments in my "economic
>relationships in collaborations" proposals. Commitments
>have time constraints; if not fulfilled in time, remedies
>could be specified in the collaboration protocol.
>
>Between those ideas, I think any requirements for
>time constraints in collaborations could be handled.
>Needless to say, they will require collaboration-level
>software, too.
>
>However, this discussion does bring to mind that
>the metamodel from which we are extracting all
>of this infrastructure stuff needs more work.
>The metamodel as adopted by UN/CEFACT
>TMWG is subtly different from the one on the
>ebXML site, and neither has been updated
>officially in the last flurry of infrastructure
>activity.
>
>Respectfully,
>Bob Haugen
