As a basis for discussion, Jad has reviewed the scenarios and made the following observations: From the transformation scenarios reviewed, I can identify 3 kinds of transformations:
Transformations where the integration need focuses on automating the steps required to perform the transformation - control integration. typical example is the scripting of code generation of FPGA using TCL.
This ought to be best handled using the Automation Server.
model/data transformations where the integration need focuses on transforming data from 1 tool to another - data integration. Example, migrate from a UML model to a Simulink model.
Given that the focus is on data, should there be better direct support for this data integration in OSLC?
A specific example of the above data-focused transformation is when migrating the same type of artefacts from 1 tool to another - equivalent tool.
In this case, would be more reasonable to use OSLC services instead of a transformation engine?
quick summary and questions - based on a reading of the scenario by Jad
2 (new?) resources were defined in the scenario:
Entity Generator Resource - Entity Generator Resource is a resource that represents and manages a long running transformation of Entities to source code.
references a set of “oslc_am:EntityResource”s
Entity Resource - Entity Resource is a model element resource that represents an Entity. It has simple properties (name/type) and can be transformed into source code.
= an AM Resource? What is the relationship???
Scenario summary:
Loops through the oslc_am:EntityResource”s of a Entity Generator Resource
Produce a sourc file per EntityResource?
How do the defined resources relate to the Automation resources? (automation plan, automation request, automation result)
The looping through entities to produce source code - as defined in this scenario - does not seem to follow the Automation specification mechanism where a request is created to run the transformation, and producing a result.
