2008-11-04

LinkedIn: Is BPMN portability important to you?

<question group="BPM Guru" >
Is BPMN portability important to you?
There have been a number of blog posts recently on how to provide the underpinnings to allow one to prove correctness of a BPMN diagram. This is necessary to allow a diagram to be carried from one vendor tool to another. If you can't prove that it is correct, then different vendors may have different interpretations, and you fail to achieve portability. Interesting posts:

http://www.brsilver.com/wordpress/2008/09/21/model-portability-in-bpmn-20/

http://kswenson.wordpress.com/2008/09/29/will-bpmn-20-have-model-portability/

The BPMN standards committee has not responded in this discussion for compliance criteria. WfMC is trying to provide compliance criteria, including levels of compliance, for BPMN inside the XPDL work. As the OMG officially distances itself from WfMC, this work might be contradicted by OMG group. That does not help anyone. We need the official compliance criteria to be in the official BPMN spec.

My question is: Is it important for the BPMN standards committee to focus on Model Portability?


</question>

I think that model portability in BPMN is very critical.
Ideally, I would like that BPMN quickly and without repeating errors will reach the existing level of XHTML -- a clear set of elements, power of XML, control of interpretation by CSS and extra flexibility by DOM+API.

"What you model is what you execute" is the goal. How the execution is implemented - via BPEL or not - is another question (for example, an HTML document can be rendered with PostScript). Of course, the execution semantic of BPMN must be well-defined and being validatable.

Thanks,
AS
Post a Comment