Re: [soa-rm] An SOA-RM policy map

[Don Flinn <flinn@alum.mit.edu>]

Frank

I like the diagram as it brings new dimensions to the problem.  A few
comments:
I believe:
- Policy is an instantiation of (some/all?) the Semantics 
- Policy models Real World Effects
- There are two classes of Participants 
    i) The runtime participant and 
    ii) the participant that defines the service (Defining Agent??).  
- The runtime participant is constrained by the Semantics
- The Defining Agent creates the Semantics.  
- Similarly, the runtime participant desires real world effects (no
change)  
- The Defining Agent uses real world desired effects 

The Defining Agent is probably too concrete for a RM but there is a
duality to the participant that should be made clear.

Don


On Wed, 2005-06-01 at 11:37 -0700, Francis McCabe wrote:
> This is not done in the IHMC tool set -- I have problems with it at  
> the moment.
> 
> This is another diagram expressing some relationships around  
> semantics, policies etc.
> 
> Note that it includes a participant concept -- that means that there  
> are participants in service oriented systems.
> 
> Some of these relationships are precise, (projection of), and some  
> are fuzzy (is about). The latter is really a relation schema.
> 
> 
> 
>