I already mentioned in my earlier post “BPMN 2.0 takes dancing lessons – do we need choreographies?” that BPMN 2.0 will come with an additional way of modeling choreographies. The main novelty is that complex interactions and their refinement to message exchanges really move into the center of attention. In the earlier post I already argued why this style of modeling is actually a very good idea. It will likely bring a number of advantages over the old modeling style: choreographies can be created faster, can be understood more easily and modeling errors due to incompatibilities are avoided to a large extent.

On the other hand, this new choreography modeling style is not fully understood yet in terms of the new anomalies it will bring. Contributors to the BPMN 2.0 standard approached me already quite a while ago to have a look at race conditions that might arise in BPMN 2.0 choreographies. In the choregraphy space we are heavily dealing with asynchronous communication and we carefully have to consider what can go wrong if the partners act independently from each other.
Read the rest of this entry »