RE: The role of context in the re-usability of Core ComponentsandBusiness Processes - OR Say What???

[Andrzej Bialecki <abial@webgiro.com>]

On Thu, 15 Mar 2001, Bob Haugen wrote:

> Martin Bryan:
> >The CEN/ISSS project group on Defining and Managing Semantics and Datatypes,
> >at their meeting on Tuesday [...] expressed a wish that precanned sets of business 
> >processes be made easily available to SMEs as part of off-the-shelf office suites.
> 
> Good!  They are correct! I hope they express their wish to the ebXML
> steering and executive committees.
> 
> The ccbp-analysis group has collected a catalog of common business
> processes that would be a good starting point. The official version will be
> send to QR by March 19.  The catalog does not contain fleshed-out
> business processes, it is mostly a survey and cross-reference of
> commonly used business processes from a variety of standards
> organizations.

I was lurking on this list for some time now. Let me share with you my
impressions about the CC (I also participated in that CEN/ISSS meeting as
a representative of an SME).

First of all, I had an impression (gotten from the "Methodology for the
Disc. and Analysis of CC") that this group is convinced and dedicated to
the use of UML to present models. However, from reading the final
documents I can see it's not the case, and I can't understand why. That
Excel spreadsheet is mighty unclear, and it looks like just a heap of
everything and anything thrown together from various existing standards.
The UML model recreated from this (that someone sent a week or two ago),
seems to support this impression...

Next, I thought that I perhaps missed some unofficial documents that
present MORE of the core components than just the Party related ones. I
was very surprised to find out that this is the only catalog of CC
available. Where is the rest of the core components then, for those who
want to exchange something more than the Party related information?

Now, please bear with me one more moment. The SMEs, and not only, are
waiting with great eagerness for the ebXML project to produce something
that can be implemented, and which would ensure interoperability and lower
costs for them to enter the e-commerce arena. Many ebXML spec drafts have
already been used as a basis for early adopters, as well as strategic
guidelines, but not the CC, which (as the name suggests) form the core
needed for exchanging the business information. So, IMHO, many of these
SMEs will be very disappointed if the CC specification will be accepted in
its vague and incomplete form, as it basically is today.

What are my expectations then? Maybe I'm just an isolated case, but I
think that in order for this specification to be accepted and deployed,
and if you really have the SME market in mind, these specs must be much
more concrete. Giving too much freedom is not always good, especially for
interoperability. They should present a consistent, well normalized model
that is readable (i.e. a UML representation). And, they should present a
COMPLETE model of the core components needed to perform the most common
business exchanges, with the guidelines for extending it. Otherwise, what
is the practical meaning of this spec if it doesn't guarantee a decent
level of interoperability? Why should I accept this vague model (that in
fact places on me the burden of creating my own components) instead of any
other one (RosettaNet or xCBL comes to mind)?

As a system architect for an SME, I'm certainly not going to embark on the
ebXML adventure without much more clear and precise definitions for such
central concepts. Which is sad, cosidering how many excellent specs this
project already produced...

Andrzej Bialecki

//  <abial@webgiro.com> WebGiro AB, Sweden (http://www.webgiro.com)
// -------------------------------------------------------------------
// ------ FreeBSD: The Power to Serve. http://www.freebsd.org --------
// --- Small & Embedded FreeBSD: http://www.freebsd.org/~picobsd/ ----