Baffled by rake symbol intricacies

[art702]

Coming from years of modelling with other tools I came back from exile to EnterpriseArchitect (EA) to find that I'm dumbfounded by the way CallBehaviours are handled in EA. The reason for this may be the "dumb" part of dumbfounded, though, so please bear with me.

My way of handling things: Add activities to an Activity Diagram (AD) one by one. Later on, I want to refine on one of the activities by making it a "CallBehaviour" with the small rake symbol in the lower right corner. I found the command at "ContextMenu > Advanced > Convert to Invocation" which makes the rake symbol appear.

From that point on I am at loss what to do. I cannot "double click" to move to the called behaviour, i.e. another activity diagram. I can associate an existing AD with the activity only by walking through a series of steps to add "references" to an existing AD in the "Project Browser", provided that the activity I want to link is currently visible in the main edit panel. I cannot read up on the different ways that call behaviours for activities and call operations on actions are supposed to be used, as there is no documentation regarding this aspect (that I can locate). I find the symbols in the Project Browser only partly helpful. I find that Composite Activities work the way I expected CallBehaviours to work ("Make Composite", double click).

Questions: I would be most thankful to helpful comments regarding

(1) how to model a CallBehaviour using EA
(2) how to differentiate CallBehaviours and CallOperations and when to use what
(3) how to read the Project Browser symbols
(4) how to navigate around my model between the CallBehaviour and it's associated detailed AD (if there is such a way) using the mouse or the keyboard
(5) how to find a UML specification root for "composite activites" which I didn't learn about at my UML class in primary school (was I just missing that day?)
(6) how to decide upon whether CallBehaviours/CallOperations or Composite Activites are the way to go.

A recent post on tracing from Activities to Actions was somehow helpful. Still I'd appreciate your comments.

[Helmut Ortmann]

Hello,

Call Operation:
Just drag and drop an operation on the activity Diagram.
Use Call Operation if you want to have the relationship to your class model.

Call Beavior:
Just drag and drop an activity on the activity diagram. Choose Invocation.
Use Call Behavior if you want to call another activity or reuse beahavior.

Re Use of Activities:
You may drag an Activity(Composite Activity) by link on the diagram. You may drag an Activity as Invocation (Call Behavior) on the diagram. With a Call Behavior you can send parameters to an activity. For reuse of Beavior I only use Call Behaviors.  

Efficiently navigating and working:
Have a look on Geerts Navigator. I have my own Addin and a slightly different strategy.

Helmut

[Geert Bellekens]

The way I do it is:

- I create an action on my diagram
- Go to advanced and set the subtype of my action
- Close the properties dialog.
- Reopen the properties dialog and set the called Behavior/Operation,...

This gets a lot easier if you stop using Activities on Activity diagrams.
I've written a blog post about that: UML Best Practice: There are no Activities on an Activity diagram

Geert