Hi
I'm first of all trying to establish a meaningful use object flow. For example, if a different message is sent to a user based on different flow, is it appropriate to have an activity node with an embedded object node, and then object flow to a send node - while having different control flow ?
So essentially, in the diagram below, what I want to say:
- if Activity1 was succesfull, send ObjectNode1 through Action1
- if Activity2 was not succesfull, do Activity2 and send ObjectNode2 via Action1
I also just want to confirm how to correctly model the details of a composite activity with an object node, below I have attempted this for Activity1 in the previous diagram.
If someone can also recommend a book which actually explains the complete semantics of UML activity diagrams. I know UML 2.0 activity diagrams are supposedly based on Petri Nets, but this in itself is not helpful, and for some reason there is significantly more useful information on the syntax and semantics of Petri Nets than UML, where mostly for UML you find maybe some syntax, but mostly just the enumeration of the input alphabet, if we were to treat it as a formal language.
