-
1. Re: Why must we call executionContext.leaveNode() in Node's
fatbeat Oct 12, 2006 10:55 AM (in response to youngliuus)from my point of view you can have more options after an action handler, since an action handler can be on a transition or a node.
at the end of the action handler you can tell the instance which way you want to go after the action handler, since a node can have more than one transition. -
2. Re: Why must we call executionContext.leaveNode() in Node's
youngliuus Oct 12, 2006 11:04 PM (in response to youngliuus)Thank you for your response. I agree with what you said. But we can add calling leaveNode in class Node, after calling ActionHandler.
The case I feel really unconvenient is when the ActionHandler is used by a node, and there is only one leaving transition, or we just want to go after the default transition. And I believe this is true for many cases.
Moreover, it's a good idea that let the end-user write script for a node's action. Currently, we must tell the end-user to call leaveNode after the very end, and setVariable to processVariable before leaveNode if she/he updated some of processsVariables; if she/he didn't do these, the process doesn't work.
Of cause, I can add my node type to jBPM, and implement this in my node type.