Re: [soa-rm-ra] The Service Interface

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

Folks,

 

The key here is to separate the information needed by the service participants (i.e., service consumer and service provider)---which are typically humans---from their respective software or hardware agents that are used to perform the interaction.  In the latter case, we're talking about the communication between the consumer agent and the provider agent (or service or, if you prefer, service agent).  [Note that for the RA, we have agreed that the service (or service agent) is synonymous with provider agent.]  For that reason, I prefer the Item 7 articulation of Service Interface from Ken's latest message, which of course is just one of the many references to Service Interface from the RM.  Unfortunately, many of these articulations of Service Interface in the RM are misleading, inconsistent, or ambiguous.

 

Clearly, the Service Description artifact and all its linkages are necessary for consumers---again, typically humans---to determine whether or not a service offering provides the capabilities and qualities of services it's looking for, plus policy & contract information, etc.  But the meta-level aspects of the Service Interface also need to be captured in some manner in order for the consumer to either manually develop or use tools to automagically generate the consumer agent at design-time in order for the consumer agent to communicate with the service agent at run-time.

 

Looking forward to a spirited discussion during the telecon.  In any event, we need to nail this down soon!

 

Cheers...

 

 - Jeff, JPL