Hmm, that's a pity! The tool I was planning to use is not working as I would like it to work. To explain it, this will become a rather technical story for some of you. For others however this will be a piece of cake!
Here we go:
Here we go:
I modeled a Recruitment process in modelworld.nl and while running through the process I ran into the error that it is possible to ‘fire’ both the processes in the lanes while only one should be able to fire. This because of the fact that the recruiter lane should only start after an application has been submitted. Let me explain what I did, so anybody can reproduce it. In prototype it is possible to run through a process by selecting the “Recruitment” process and click ‘run this model’.
- Click Show tasks
- Here you see both the starts of the process
- Click Start next to “Accept create new application”. Now you see a Short case description (Order #...) and some details (blank)
- Click OK
- Now you can continue using the same case under “My current tasks” or you can create a new case. I will continue using the same case.
- Click Start and OK (repeat this until you are in “Called for interview?”)
- Here you see a choice: Yes (application received and invited for an interview) or No (application is received but declined)
- Click NO and you’re done for this example.
According to the above described steps, it is possible to finish the process without receiving a response from the recruiter, while the recruiter did not invited you for the interview. This should not be possible because the bottom lane is dependent on the lane above. The error already occurred in the second step; it should not be possible to start with “Application received” because no application has been sent yet. This means that it is possible to have two complete separate cases running through the model during one simulation.
So, either the tool should be fixed or I should find a new tool. In the paper by Van Gorp and Dijkman (2011) a solution is provided to fix this problem. They use the reEnterSubprocess rewrite rule, as depicted below. See section 3.2 in the paper by Van Gorp et al. In the coming weeks I will focus on this a bit more and I will continue with reading for my literature study.
Rule enterMessageCatchIntermediateEvent and enterMessageStartEvent by Van Gorp and Dijkman (2011) |
For my literature study I will dive into the pathways, as described by the European Pathway Association (EPA, http://www.e-p-a.org/) and during the holiday I will try to read some articles from this journal, which looks very promising.
So, I have plenty to do besides the upcoming exams.
Tweet
Hi Willem,
ReplyDeleteI am aware that the BPMN prototyping function on modelworld still has a lot of shortcommings.
Regards,
Bart-Jan