Hi all,
I have just recognized a strange behaviour of Aris Express which seems to be a software because as far as I know this is not a bpmn specification issue.
If you try to connect an artifact A (e.g. an activity or event) in pool 1 to antother artifact B in pool 2 this is no problem, unless the artifact B you want to connect to is idefined in an expanded subprocess. In this case ARIS does not allow you to connect to iartifacts inside the expanded subprocess. My workaround is to move the artifact B out of the subprocess, then connect it, and afterwards put it back into the subprocess.
Does anyone know if this is a software bug or maybe some kind of a feature?
Kind regards,
Christian
This is not a bug but the semantically correct BPMN implementation. You are not allowed to connect directly to objects in the subprocess according to the spec and therefore ARIS doesn't allow you to do this.
The correct way to do this is to connect to the subprocess object, because you should be able to create a valid BPMN diagram even when the subprocess is collapsed. When expanding a subprocess BPMN requires to include a complete, fully semantically correct BPMN diagram in that expanded subprocess (with start/end events and such *and yes, I know about implicit events, but I don't want to start this flamewar here*).
I addition to this ARIS will change your connection type to a "message" connection since you are crossing pools and then there are additional semantics/rules in using messages (message objects, message events, message tasks, message connections) in BPMN. Please have a look elsewhere in the BPMN group for a more detailed discussion on this -and a sane way to handle this complexity in real-life project situations.
I was going to comment, but Roland's already covered this point nicely.
The work around to move a connected box into a different pool usually pops up a window to ask whether you want to 'ignore constraints' and at this point your model is no longer semantically correct.
Regarding the specification this should be valid? Please correct me if i'm wrong.
BPMN 1.2 spec (p. 101)
"If there is an Expanded Sub-Process in one of the Pools, then the message flow can be connected to either the boundary of the Sub-Process or to objects within the Sub-Process."
BPMN 2.0 spec (p. 311, Figure 12.3)
This is interesting and an inconsistency in the 2.0 spec IMHO (maybe from updating form the 1.2 doc, since that sentence was removed below the picture below). The example in 12.3 is from the Choreography chapter and not the Collaboration specification (chapter 9).
In that definition Figure 9.4 shows only a message flow to a collapsed sub-process, which is consistent with the characteristics of level consistency - e.g. throwing an end event to the border of the superior sub-process task while becoming an intermediate event.
When you look in the sub-process chapter (10.2.5) there is no hint that this is allowed, Furthermore, for some sub-process types it is explicitly forbidden (event sub-processes, ad-hoc sub-processes).
I also played around -very unscientificly I must admit- with our competition (BizAgi) and this process modeler also does not allow to draw message flows to tasks which are part of a sub-process.
Hello,
It seems many people come across this issue using ARIS.
In this case, my interpretation of notation would be that, where it is not explicitly forbidden, it means it is allowed.
as taking into account some inheritance, that BPMN v1.2 should be fully compliant with BPMN 2.0... I would not like to believe if in v1.2 it was allowed to connect message flow with objects in sub-process, that now in v2.0 it would be restricted in all cases. Such modelling style I guess is really widely used practice.
BR,
Ilze
I have a similar problem, but in the reverse direction. I want to show messages from a task within a sub-process to another pool. I see an example of this in BPMN 2.0 By Example from the OMG website, but ARIS won't allow it. How can I show explicitly where a message comes from in the sub-process?