I’m trying to decipher Cordys chief strategist Jon Pyke’s post today on the case management proposal at OMG. It’s hard to tell what he’s saying, but I gather things did not go well in Costa Rica. I could have told him that, based on the bmi thread beforehand. He casts as the villain “analysts and consultants
I hope he’s not talking about me, but I don’t know of any other analysts who are even thinking about case management or discussing any OMG efforts in that area. If so, I think the blame is misplaced, since if anything I am a supporter of standardizing a notation for case management and integrating it with BPMN. Probably one of the few friends Cordys has in that regard…
I do think the OMG process is an unlikely one for case management. But just because I’ve given up on it before it starts doesn’t mean I am trying to “scupper” it (whatever that is… sounds bad). Anyway, BPM analysts are hardly the ones driving the boat at OMG. I wish! My BPMN 2.0 experience has taught me that the real battle there is between the “architects,” UML/metamodel people stuck in the 20th-century paradigm of models that can be compiled into any programming language, and the “engine vendors,” who want models that roundtrip with metadata-driven (code-less) implementation design. Preferably consistent with their existing engine architecture. I have more sympathy with the latter, but I have to say that neither one is especially concerned with notation per se, i.e., diagrams that are expressive but lack executable detail.
I wish the Cordys guys well in their quixotic OMG adventure, but I think an informal process in which participants share the basic objective and ground rules – notation-centric, non-executable, xsd-not-UML, linked to BPMN 2.0 – is more likely to achieve something than a formal process in which half the participants never liked BPMN in the first place. If Henk and Jon get tired of spinning their wheels in OMG, I would welcome their ideas in another forum.