RE: What do people really expect from ebXML? - spinning mirrors?

["William J. Kammerer" <wkammerer@foresightcorp.com>]

Back when I was in high school, just a few short years ago, I took a
Physics class and one of our assignments was to figure out the speed of
light using Foucault's method based on spinning mirrors or some such
nonsense.  Though we were provided with the materials and detailed
instructions, I - having no patience or desire to get my hands dirty -
proceeded to figure out how to do the least amount of work.   So instead
of laboriously screwing around with mirrors, motors and lasers, and
making repeated measurements, I figured if I started with the (known)
speed of light I could just work backwards and calculate (on a computer,
which they had in those days) the correct position and rotation of the
geegaws.  Needless to say, I came up with a beautiful set of
experimental observations.

Unfortunately, I was the only one in the class who came up with anything
near the correct answer - apparently the teacher had left out a step or
two in his instructions, which would have made arriving at the actual
speed of light impossible.  I was found out, and swore off the hard
sciences from that day forward.

Lack of patience is a family trait, as witness my great-uncle Paul
Kammerer, from Vienna.    Uncle Paul was a Commie, and was intent on
proving the inheritability of acquired characteristics, which would
somehow buttress Marxist ideology. See
http://www.mbl.edu/publications/Ciona/Kammerer/.  Anyway, instead of
waiting to see if his experiments proving Lamarckian inheritance would
be borne out, he supposedly counterfeited results which, when
discovered, ended in disgrace and suicide in 1926.  Fortunately, the
Cincinnati Kammerers came through relatively unscathed by the scandal,
though we had for a short time considered changing the family name to
Goebbels.

Despite these unfortunate histories, I would still advocate reverse
engineering when it comes to core components.  Instead of analyzing a
"process," which is messy, - and besides, the process may not give us
any idea of what the data requirements are - we should just work
backwards and mine core components from already existing materials like
EDI guidelines and EDIFACT directories.

David Lyon wants simple transactions, like POs.  Instead of agonizing
over the "process," why not look and see what's been done before? -
where presumably the authors thought long and hard into what goes into a
simple PO.  But where should we start?

Simpl-EDI sounds, well - simple. See
http://www.e-centre.org.uk/products_edi_simpledi.htm  But simplicity
belies its name - the EDIFACT messages are simple only because
assumptions are made on the use of sophisticated databases keyed on EAN
(UPC) codes and EAN GLNs.  Simple - yes, but for LEs (Large
Enterprises)!  Maybe a more fruitful model is that of the Open Buying on
the Internet (OBI) Consortium, whose specification provides for a simple
ASC X12 850 PO;  see http://www.openbuy.org/.

If my arguments aren't convincing, then have a look at the ebXML Catalog
of Common Business Processes (bpproc_v0.99.pdf), whose section 6.4 -
Discovery of Core Components - notes the following:

   The catalog of common business processes is useful for
   discovery and analysis of core components that will be
   used as the building blocks for deriving business documents
   within a given context. This can be done by checking all
   sources of documents listed and cross-referenced on the
   Common Business Process Catalog to identify a document that
   may have the information needed (which may be EDIFACT, X12,
   xCBL, RosettaNet PIPs, CII, OAG BODs). There may be an
   existing document, which is similar and could be evaluated.
   Next identify if the document components meet the business
   requirements. If so,  then these components can be reused.

Sounds reasonable. I'm sold.

William J. Kammerer
FORESIGHT Corp.
4950 Blazer Pkwy.
Dublin, OH USA 43017-3305
+1 614 791-1600

Visit FORESIGHT Corp. at http://www.foresightcorp.com/
"accelerating time-to-trade"