So the Action Model is the set of actions a service can be asked to do in order to realize RWE. It is potential. Turning that potential into reality is accomplished through Joint Action between the Consumer and the Service. The consumer is asking for the service to carry out the advertised action – likely some business function – expecting the advertised RWE.
In order for Joint Action to occur, the Execution Context needs to be established. The RM defines the EC. Much of the EC derives from information in the service description. However, the service description may indicate alternatives, e.g. alternate restrictions on sharing private information or the set of vocabularies the service is prepared to understand. Let’s look at the vocabulary example: the service description indicates the service can understand both NIEM and UCore schemas; the consumer chooses to use UCore schemas and so indicates in its messages. The agreement on vocabulary is set because the EC as established allows for consistent choices among the participants and, indeed, one of those choices will be used. If the Consumer wanted to use a schema other than NIEM or UCore, some vocabulary mediator would need to be identified and available for use in order for the EC to be successfully established.
It is possible that in establishing the EC, there will be resulting limitations that preclude using some of the Actions in a service’s Action Model. This will likely affect the RWE that can be realized or the conditions under which the RWE can be realized. For example, there can be a limit on the cost that can be incurred and so a data set will only go through one level of processing instead of a second level the service might also offer.
Now, there may be numerous interactions required to establish the EC. This is most notable in the idea of establishing willingness and trust. Our current discussion limits Joint Action to realizing actions in *an* Action Model and not every interaction that may go into establishing the EC. Note, this does not preclude some of those interactions being done using services and that there are Joint Actions with those services. In any case, we need to explicitly acknowledge and account – but not necessarily fully expand the social aspects in excruciating detail – for those interactions that are essential to making the ecosystem work.
Now action as “application of intent to achieve an effect” seems to more directly support the ecosystem context of interaction to establish EC and then engage in Joint Action than it does the idea of Action in the Action Model. I mentioned during the last call that it seems section 3 is more what leads up to using a service and then dealing with the RWE while section 4.3 is more what happens after the joint action begins and the magic happens producing the RWE. I’m not requesting a change in the definition but noting there is a critical need to consistently connect all uses of Action with the definition. That is the final piece that I think will tie up discussions – I’ve been reading RA emails from 2007 and 2008 – that have gone on for years.
Ken
---------------------------------------------------------------------------
Dr. Kenneth Laskey
MITRE Corporation, M/S H305 phone: 703-983-7934
7515 Colshire Drive fax: 703-983-1379
McLean VA 22102-7508
Attached is the one-pager we talked about at today’s conference call. For those who didn’t attend, the minutes should be posted shortly. The purpose of this is for us to come to agreement on these concepts and their relationship to each other before next week’s call.