OK, this wish is a bit more for discussing how to model human-centric workflow features which are mostly needed in the context of forwarding an process from on human task to the next one.
I choosed an example where some receivers ot the human task named “approve Credit Enquiry” should parallely check and forward “manually” (see wish two) selecting from the context menu “reject” or “approve”. Judgements will be written into workflow comments before. The receivers may be determined using a Rules Engine or by the forwarding user of activity “Enter Credit Enquiry Data”.
But, if there are for example 10 receivers, when should be decided to go for a rejection or an approval?
- Should every receiver vote, so the decision should be carried out when the last forwarding was done?
- Or should we go if the the simple majority is reached so the outstanding votes can be removed from the inboxes?
- Or should we go if the percentage of e.g. 30 is reached?
- What, if there is a parity, should the owner decide what to go for?
So we implemented all of these votings as attributes on the sequnece flow (her a manually one that is not BPMN 2.0 conform :-) some years ago and there are all used out there.
And we implemented some more attributes to support the users during forwarding the task:
- Ask the user to enter his password again which is procotolled for audit purposes
- Ask for the reason he selected this forwarding and not the other one which is again protocolled
- And as whether he wanted to go for this forward or not. If not cancel the selection.
I think there may be other reasonable features in the context of human-centric workflow situation. Let us speak about and decide which of them should be part of BPMN 2.1