State Model Subsystem Part 1!

As promised, I have just uploaded the State Model Subsystem metamodel diagram and illustrated model descriptions available in the usual formats on the Download page (near the top).  I'll get going on the other half.  In the meantime, feel free to post any questions or comments.

Comments

[Eric Mintz]

Lose R507 and add an Action to the Response subtypes. That way, an event can have a Response bound to the original state and a Transition bound to the new state.

Also, events can propagate to several Objects, as in hierarchies. So it looks like we need a many-to-many between Events and State Machines whose resolving object is bound to the response.

Further normalization is possible, and possibly desirable. A fully normalized logical data model contains only subtypes and one-to-many relationships. Is that a goal?

Thanks.

[Leon Starr]

Hi Eric! With regard to propagation, etc. That will be covered in the Polymorphism subsystem coming shortly. It will have some impact on the existing State subsystem. So I will do a double post with all the new stuff. I can assure you that a lot more than a 'many-many' will be required to sort all this out!

With regard to R507, keep in mind that Actions/Procedures are not yet modeled. Though that might not be what you are getting at. And I'm not at all sure what you mean there. In Executable UML, a Procedure is associated with a State. So the only way to trigger it is to transition into the destination State. Maybe you can clarify?