mid-course correction on abstract model for module processing

Abstract Model Group:

I had a lengthy call with Henrik on Monday to clarify some things about
the SOAP execution model, the semantics of actor URI's, etc.  It turns
out there are a number of unadvertised mechanisms that address some of
the concerns that motivated my former proposal.  It also appears
possible to have an extensibility mechanism that would allow more
elaborate "manifest" processing models within a SOAP-like framework.

As a result, I have tried to capture a set of "Concepts" which are much
closer to the existing SOAP framework.  I recommend shelving the former
proposal for the time being; it can be considered later as fodder for a
more elaborate processing model that uses the extensibility mechanisms.
There are still some outstanding issues with the SOAP model that need to
be addressed and cleaned up.  I will use the alloted agenda time
tomorrow to discuss these issues.

Mark Jones
AT&T Labs