Will UN/CEFACT torpedo ebXML?

and mortally wound itself?

An e-mail recently graced my inbox that caused me to ponder these questions. For you to understand why I had this reaction, some background may be in order, especially if you're not familiar with this little corner of the arcane world of standards.

UN/CEFACT, with OASIS, sponsored the ebXML initiative and is continuing development of the Business Process and Core Components work that ebXML began. The CEFACT Techniques and Methodologies Workgroup is responsible for maintaining the UN/CEFACT Unified Modeling Methodology (the UMM) that is the Bible for how CEFACT and other groups following the ebXML approach are supposed to develop eBusiness standards. TMWG will be meeting in late May. According to the e-mail, one of the topics set for discussion the troubling open issue of how the ebXML Core Components fit into the UMM. That's the setup.

If the UMM is the Bible, then Klaus-Dieter Naujok is the chief prophet. Mr. Naujok has been chair of TMWG since its inception. He was also chair of the ebXML Working Group during its lifetime, and is serving as chair of CEFACT's Electronic Business Transitional Working Group. Regardless of whatever his actual opinions might be about the Core Components work, Mr. Naujok is widely perceived as being antagonistic to it. He wrote the e-mail that caught my attention. In it, while hoping for a favorable outcome, he raised the possibility that if TMWG doesn't "define what CC are in our UMM Meta-model, they will become no longer relevant in our BPI work and therefore have no standing in UN/CEFACT." In other words, CEFACT would drop one of the key ebXML products. It is extremely significant that Mr. Naujok should even raise this possibility, since it could lead to a very interesting chain of events.

Let us assume that TMWG doesn't figure out where the Core Components work fits and decides not to include it in their meta-model. Even assuming that the ebTWG approves the Core Components Technical Specification which is currently out for public review, it won't go anywhere. CEFACT last year adopted the then incomplete UMM in principle. In pragmatic terms if not in fact, TMWG pretty much has a blank check to finish it. If they decide that CCs don't fit, then CEFACT is not going to do anything with them. It will instead fill out the lower parts of the business process modeling methodology, which is the part that most directly leads to XML schemas, with the existing TMWG work.

So, why would this e-mail, this discussion about a somewhat obscure set of technical reports and draft specifications, and this possible chain of events lead me to pondering these questions? Let's consider the possible outcomes of such a scenario.

In the best case, everyone in CEFACT accepts TMWG's decision and rationale on their merits for dropping Core Components, and follows their lead. The current body of CC work, developed over 18 months by CEFACT's EDIFACT Working Group with assistance from ANSI ASC X12, is either picked over for what can be salvaged in compliance with the UMM, or discarded. Other standards bodies who have been basing their work on Core Components, such as the OASIS Universal Business Language Technical Committee, modify their approaches. OASIS modifies its ebXML related work to remove references to Core Components. Vendors, who have already built products to work with ebXML Core Components, modify their products.

In the worst case, TMWG's action causes a severe rift in CEFACT. Many of the EDIFACT Working Group business experts, who have invested a lot of time and effort in Core Components development, leave. The OASIS Board of Directors must already be somewhat upset with CEFACT over how the ebXML work has puttered along in CEFACT following the ebXML completion. They get royally ticked off that CEFACT would unilaterally discard one of the key ebXML products, and sever any further formal cooperation with CEFACT on ebXML. ebXML splits into CEFACT and OASIS variations. The OASIS UBL committee takes over development of Core Components and continues to build its UBL XML business standard using them. ANSI ASC X12, already very upset with CEFACT over the termination of the Joint Core Components work effort, drops any pretense of aligning with ebXML in developing its own XML schemas as American national standards. Vendors don't know which way to go. Some go with the UMM without Core Components, some ignore the UMM, some try to do both, and others give up entirely or go out of business because their ebXML investments aren't going to pay off.

In the middle ground, some CEFACT people support the decision, some don't care, some who don't agree with it stick around anyway, others are upset enough to leave, and most (but not all) of those who leave are replaced. The UBL committee sticks to Core Components as do some other bodies, but many go with the UMM. Others decide just to ignore the CEFACT work, since they can't be dependent on shifting foundations that take too long to build. Most serious vendors adapt to the changes, while many get out of the market. Though not as bad as the worst case, ebXML marketing momentum is dealt a blow and CEFACT's cachet is eroded.

So, how likely are any of these outcomes? To believe in the best case, you have to believe that decisions in standards bodies are made entirely on technical merits and that politics, business considerations, and emotions have nothing to do with it. If you believe that, then I have bad news for you, Virginia, there is no Santa Claus. The best case absolutely would not happen.

That leaves the middle and worst cases. In most situations offering best and worst cases, my gut usually directs me to a middle ground. However, knowing some of the dynamics and personalities involved with this situation, I lean more toward the worst case. If it happens, not only is ebXML torpedoed, but CEFACT puts itself firmly on a track of becoming an irrelevant, bit player in the future of electronic business. And, we dash whatever hope is left of having a single XML business standard.

How is this going to turn out? As I said, even though Mr. Naujok expressed his hope for a positive resolution, I find it very telling that someone with his leadership responsibilities and influence should even raise the possibility that the Core Components work would be discarded. Will TMWG resolve their problem with figuring out where Core Components fit into their UMM? Will they sweep their technical concerns under the rug and embrace Core Components for political expediency? Or, will they shoot themselves in their feet? However it plays out, this is going to be very interesting to watch!

May 1, 2002

© Michael C. Rawlins

About  |  Services  |  Resources  |  Contact

©Rawlins EC Consulting 2007