Just received a note from Phil Gilbert of Lombardi, a key contributor to the BPDM effort in OMG, that says:
[I] wanted to let you know that the OMG Architecture Board voted to approve the BPDM spec today. There are actually 2 more small hurdles instead of 1 more as I told you earlier. But these are 99.9% certain to approve specs that have passed the Architecture Board review. Apparently these take several weeks calendar time as the boards that approve aren’t on the TC calendar, they have their own.In any event, it appears that a major milestone for the industry has been passed: a specification for a business process metamodel (as opposed to UML-defined process) is poised to achieve standards status and have the backing (and implementation) of process platform and modeling vendors. This will insure a standards-based way in which BPMN models can be exchanged, and both standards are driven by the same organization, allowing for unprecedented alignment. In fact, at this meeting, the next major version of BPMN is being discussed and it is expected that the focus will be on using that effort to merge the BPMN and BPDM specs, so that there will be one modeling spec, and that spec will have it’s explicit notation and its explicit metamodel.
The first step was to define the notation, and now we’ve defined the metamodel.
I think end-user companies are the winners today. There is now clarity around the implementation of process at the business level, and that clarity is in the form of business-based process notions, not IT- and systems-based notions of process.
This message was an update amplifying a previous one that announced BPDM approval by the Domain Task Force:
… this morning in San Diego at the OMG Technical Meeting, the Domain Task Force voted FOR approval of BPDM! This is the second major hurdle of any spec. The first is the RFP which kicks off the submission team… that was done long ago for BPDM. The second is the DTF vote (which occurred today). The third is the review and vote by the Architecture Board, which will happen Friday of this week. Then the OMG Board ratifies the AB decision within the next 30 days or so.
While it is rare, the AB can reject or send back a spec for more work, but in general the vote of the DTF is the more difficult barrier because of the diversity of companies (and interests). The AB is more focused on the technical implementation of the specification and less concerned with the business issues that may be reflected in the votes of all the member companies.
Phil also replied to my questions with some answers from Lombardi and MEGA representatives on the BPDM submission team:
Q. How would you create serialized BPMN from BPDM?
A. We are updating the EMF plugins. This will allow creation of BPDM models using Java. Additionally there will be a very basic editor (not usable by business people) that will allow a developer to create BPDM models and to look at the produced xml. The goal is to have these plugins updated for Monday 26 March. This will also be discussed on Sunday by the submission team. Most likely we’ll do a presentation about the serialization of BPMN using BPDM and we’ll show the actual XML that is produced.
Q. Could xsd be derived from this xmi document?
A. Yes. EMF provides such capability. XSD most likely will not be part of the submission, but a separate document that will be distributed freely.
Q. Do you expect OMG to provide one?
A. Yes. During the meeting we will create and distribute the XSD.
Q. Can standard xml validation be used?
A. Yes. Using the provided XSD.
Q. Can I use XSLT [to transform and validate models]?
A. Yes. However XSLT is not very suitable for performing model transformations. For object models there are better languages that can transform models. One such example is QVT. On the Eclipse.org you can find other languages also.
Q. Is there some example that can be used to understand BPDM?
A. We are working on an example that demonstrates BPDM features. This is incorporated in the html documentation that is generated. Most likely it will need additional explanations, to be provided after the approvals
“The first step was to define the notation, and now we?ve defined the metamodel.”
Funny: they first define a notation without a proper syntax or semantics, then they go on to define a metamodel to properly capture its syntax, but the problem is that this meta-model does not match the notation… But patience! They’ll now start having meetings about how to revise the notation to match the meta-model…
Ah! And this weekend we’ll fix the XMI… And we’ll do a few twists here and there and we’ll get EMF to generate a proper XSD. But this XSD won’t be normative, so vendors will have the choice of using a different one?
Bruce: I think we’re in for another year of waiting for the next version of BPMN!
[…] BPDM Passes Important Hurdles (tags: bpmn bpdm bpm cx) […]
[…] BPMS Watch » BPDM Passes Important Hurdles The BPDM standard is moving ahead: this is a report from March on the status. I’m expecting to see a lot more about this at the BPM Think Tank next month. (tags: bpdm) […]