Just found out that we've used the groupId in the jPDL file, for example, in the TaskCandidateTest.class, we've had following jpdl file. | <?xml version="1.0" encoding="UTF-8"?> | | <process name="TaskCandidates"> | | <start g="20,20,48,48"> | <transition to="review" /> | </start> | | <task name="review" | candidate-groups="jbpid_group_id_._._unit_._._sales-dept" | g="96,16,127,52"> | | <transition to="wait" /> | </task> | | <state name="wait" g="255,16,88,52"/> | | </process> | For the candiate-groups, we need to specify the groupId, because the groupId is identity component specific, which means it might differ in different identity implementation. So I don't think it is good to use the id in this case. Because in the process of doing jboss identity integration, I've found if I want to make the test suite passed, I need to update the jpdl file. And if I want to switch to use the jbpm-built-in identity, then need to update the jpdl file again due to the change of groupId difference. My suggestion is: for groups, we used: | candidate-groups = "groupName/groupType" | if the groupType == null or a default jBPM groupType (unit), it can be just specified as only groupName, like: | candidate-groups="groupName" | While, for the users, we've used username as the value. | user="username" | Got any thoughts?? thanks Jeff View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4237698#... Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&a...