We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

Anne Stuart’s BPM in Action

Dennis Byron

BPM Viewpoint: The 2008 BPM Standards Debate--Looking Ahead

user-pic
Vote 0 Votes

In my last BPM Viewpoint, I pointed out my biases against de jure standards in general, which are colored by among many things my experiences surrounding the founding and early efforts of the Object Management Group--OMG--in the 1989-1992 timeframe. I also reviewed some business process management (BPM) research that I had done on the BPML/BPEL standards in 2004, while an analyst at IDC.

Combining the two strains of thought, I expressed my surprise that after 4 years--during which I had not been researching BPM in any detailed manner--that there was still a BPML/BPEL controversy.

Well the above paragraph is not entirely accurate: the current debate is between BPEL (Business Process Execution Language) and the Business Process Modeling Notation (BPMN) standard, which emerged as a remnant of BPML. It surprised me because I thought, and many of you agreed, that one was a notation language with no execution engine and the other was an execution engine with a very fundamental non-notational language.  

So what's the debate?

Well it turns out that the debate is not really about the BPMN that I had looked at 4 years ago. The debate rages in my opinion because there is a new standard, BPMN 2.0 (or 1.2--but who cares), under consideration. That's because the BPMN I knew in 2004 was turned over to the OMG in 2005. Apparently the OMG had hoped to peacefully merge BPMN with a standard it had been developing at the time, called the Business Process Definition Metamodel (BPDM). (Among my other biases against de jure standards, I hate keeping track of all the acronyms. Wintel is a standard you don't have to look up. VHS is a standard whose meaning you understood even if you never saw a Betamax tape.)

According to Wikipedia:

"BPDM is an alternative to the existing process interchange format XPDL (XML Process Definition Language) from the WfMC (Workflow Management Coalition). The two specifications are similar in that they can be used by process design tools to exchange business process definitions. They are different in that BPDM provides a specification of semantics integrated in a metamodel, and it includes additional modeling capabilities such as choreography... In addition, XPDL has many implementations, though only some support for XPDL 2.x, needed for interchanging BPMN. BPDM implementations are in preparation, including support for BPMN, and translation to XPDL."

So the issue in my opinion is that unable for whatever reason to combine BPMN and BPDM, and not really caring about the Business Process Management Initiative's decision in 2004 to drop BPML in favor of BPEL (see earlier post), OMG is off to the races with a new competing standard. It's because we have these multiple standards groups, that we have competing standards.

I am not even going to figure out which supplier is for which standard. I know the answer if I ask them will be: "We're for both!"

In addition the new BPMN/BPDM combo was originally supposed to be approved in September 2006. Instead--surprise/surprise--the new standard is still working its way through the OMG standards process. I say that based on conversations with OMG members. I can't find any exact information about the progress on the OMG web site but my real message is: It doesn't matter!

According to those participating, the question is deciding where the handoff between analyst and developer takes place from a standards point of view, and how much programming work there is left to do after the handoff. The right answer should be there is no handoff and no further work to do after the model is done. But in the standards-debate view, there are still two roles--analyst and developer--and the roles need to talk to each other better. In addition the new hybrid role, modeler, gets thrown into the discussion.

Among the vendor community, Active Endpoints has been most vocal about this debate at least on the record. (An early Oasis BPEL TC member, Active Endpoints is also actively pushing the standard down the open-source road via a related organization called ActiveBPEL.) Mike Rowley of Active Endpoints tells me the extent of the conflict depends on whether the eventually approved version of BPMN 2.0 (or is it 1.2?) allows things like a WSDL specification of services or permits expressions the way Xpath did/does. Active Endpoints believes the market would be best served by the two standards remaining complementary. The company believes notations must be first serialized into an execution-focused language like BPEL because no notation system can specify everything that needs to be done. Active Endpoints believes that's why most modeling-oriented approaches end up frustrating business analysts and forcing developers to write lots of Java code to implement simple processes.

So it appears the market is no closer to providing a simple seamless standard method of executing a model. And more important a simple seamless method of changing a model in response to customer or supplier demands, or a merger, or a government regulation (TARP anyone; how are you going to model and measure your CEO's excessive compensation?)

Now the market is going to decide. Which is the way I prefer it anyways.

[As an aside, BPEL is also evolving under a movement called BPEL4People that is intended to make it more human-centric-workflow based. OASIS doesn't appear to want to get involved. So that effort is even slower--maybe even stalled--than the BPMN/BPDM exercise within OMG.]

3 Comments

| Leave a comment
user-pic

OASIS started its BPEL4People effort in March 2008!

http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=bpel4people

Thanks for the info, Martin. See my related response to Bernd.
Dennis

BPEL4People is a Oasis TC, your "does not want to be involved" seems incorrect, can you elaborate? (In fact your Blog article even triggered a publicity offer from Oasis towards the TC).

Besides that, I agree very much, the OMG really seems to disturb the status quo. And some vendors (like SAP) seem to abandon BPEL activities in favor or BPMN, already.

Greetings
Bernd

Thanks for the comment Bernd. On BPEL4People I was
primarily referring to the slowness of getting any movement
but I also recall a FAQ on the OASIS web site (see
http://www.oasis-open.org/committees/wsbpel/faq.php). I guess
it would be more accurate to day, the BPEL TC doesn't
"want to be involved" with BPEL4People.
Dennis

Hello Dennis,

the reason for the BPEL-TC not commenting on the BPEL4People activities within OASIS is the simple fact that the BPEL Standard is released by Oasis, which was the Objective of this TC, and therefore it has been shut down. So the FAQ you quote is actually outdated and will not be maintained. Lets see if we can do something about that. Otherwise I can asure you as a member of both TCs that Oasis is/was supporting both movements equally well, and a lot of the original contributors to BPEL2.0 now work in BPEL4People TC.

Hopefully there is quicker process as with BPEL standard, but thats one of the disadvantages of committee driven work.

Greetings
Bernd

Bernd, thanks for the clarification. Dennis

Leave a comment

Business process management and optimization -- philosophies, policies, practices, and punditry.

Anne Stuart

I am the editor of ebizQ.

Recently Commented On

Monthly Archives

Blogs

ADVERTISEMENT