Re: [soa-rm-ra] reaching closure on Action

["Jeffrey A. Estefan" <jeffrey.a.estefan@jpl.nasa.gov>]



Hi Michael,

 

One of the questions we asked Frank this morning (and this example drives it home) is: How is Service Action different from Action?  Particularly if Action is defined to be the application of intent by Participants (or Agents) in order to achieve a RWE.  Based on this definition, isn't Service Action an example of Action?  Conversely, what about the Consumer Action?  If Service Action is to be called out separately, why not Consumer [or Consumer Agent] Action?

 

A key point we've been iterating is that a Service does not act until acted upon.  The introduction of the concept of Joint Action helps drive this notion home.  There's a fuzzy boundary that we're still wrestling with, i.e., what constitutes separate Action vs. Joint Action.  It's not all that clear cut, particular in SOA-based systems where we are trying match needs with capabilities and those entities with needs are not (or at least we assume are not) the same entities with capabilities.  It would seem to be silly to assume otherwise, at least in terms of the scope of the RM and the RA.

 

In any event, I know you were unable to attend the call today, but we decided toward the end of the day that there should probably be some loosely defined notion of separate Action in Section 3 of the RA, but that for modeling SOA-based systems, will be focusing primarily on Joint Action.  Frank has the "action" (no pun intended) to write something up that helps illustrate this based on a real world example.  He said he probably won't be able to get that out for a couple of weeks, but in the meantime, we can press on with Joint Action.

 

Bye for now...

 

 - Jeff