Hi,
I noticed collapsed sub processes are allowed in 8.4 and this works.
Some of our modelers use collapsed event sub processes (created in another tool) and while those are accepted in modeler, the engine does not execute them.
Do you intend to support collapsed event sub processes, or is this a bug and the modeler should warn about collapsed event sub processes?
Could you please clarify? Thanks.
Great question @DatyEighty
I have a quick follow-up.
Why do you prefer to use collapsed event sub-proceses?
Are you building processes for execution only or for documentation as well?
Let me find out and i’ll get back to you.
Thanks for reaching out. I was mostly curious about the new feature.
We’re building processes for both documentation and execution. We’re modeling using another vendor’s tool that supports additional BPM scenarios. I’d say models in that tool feel more spacy, i.e. they seem to require more space than in Camunda Modeler. Therefore some modelers prefer to use collapsed event sub processes. However, it’s not a common or preferred scenario per se.
Based on previous Camunda versions, we will probably require such models to be adjusted once they go from documentation to execution. Maybe we forbid collapsed event sub processes entirely. We just haven’t put a lot of thought into that detail yet.
There is a ticket lingering in the backlog for this, but personally i don’t think it’s likely to happen. We don’t get a lot of people asking for this to be implemented.
On another note, I don’t think it’s a good idea to use collapsed event sub processes - I think it hides quite important information about the execution of main process. So even if it was implemented i’d probably still suggest not to use it.
@Niall Thanks, I thought so.
I’d expect the modeler to warn about that feature. Or do you think the current behavior is okay, since the modeler can’t model collapsed event sub processes itself?
So, the modeler has a bunch of linting rules that are independent of the errors that the engine throws. Seems in this case we’re simply missing a linting rule for the modeler. I think this should be quite easy to add, so I’ll suggest that we warn users that isn’t a supported construct.
Thanks a lot for the feedback. 
I agree with @Niall this feature is not needed as we really dont use collapsed subprocesses. It is good to have feature but not really a must.
Luckily one of the befits of being an open source product is that the community can sometimes step in when we can’t prioritize certain features.
So, it turns out someone from the community has implemented this feature already. Who knows, it might get included soon
Check out these links for detials:
Thanks for the heads up. Unfortunately, this is only about modeling. But 'm not complaining. More features for bpmn-io are good
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.