I'm interested in the same issue. Assume you have a common set of tasks, called activity C, which you execute several times in a process. It was necessary to place multiple invocation or execution of C (called C1 and C2) on a diagram. The sequence flow heading from the execution depends on the execution itself, that is, sequence 1 is A > C1 > B, while sequence 2 is P > C2 > Q, just as C1 and C2 were subprocess invocation.
I can't find a suitable way to do this in EA, 'cuse the model doesn't really links the called activity and the calls, especially not visually. Therefore the readers of the diagram may have hard times to find what C1 and C2 are (what's more, the activities are documented in linked documents, and there does not seem to be a convenient or straightforward way to access the documentation of C from C1 and C2).
Alternatives:
1) I can instantiate C (CTRL-dragging to the diagram), but this is not BPMN-compliant, since activity instances will be created instead of activities. EA does not recognize any link which the users could follow to identify the called activity.
2) I can create a new activity for C1, and then set the calledActivityRef to C. Yet, there is no visual relationship between the two. Even worse, no matter if I set C1 to task or subprocess, when double-clicking on it, EA displays its properties pages instead of opening the diagram of the called activity. Set C1 to composite activity to get the related diagram, and your model will not be BPMN compliant, again.
How do you get around in such situations?
