Re: Comments triggered by Tom's comments

[Jean.Kubler@unece.org]

Hey Tom,
It is always with great fun that I read your mails!. They are definitely written
in your unique style. Am I the only one appreciating that style?

Your comments triggered some points below. I am not shure where the dicussion in
the group stands on them as i try to catch up.

I will try to be on the conf call this evening.
Regards, Jean. PS in case you come to Switzerland, a fondue is waiting for you.

1  ========================================================================

> Most of the consortium stuff I've reviewed is conceptually great but does
> not have the cross industry / global scope needed so it never gets further
> than its own niche.  And the main reason is that they just cannot
> normalize (and manage) the business semantics issues under one umbrella in
> a usable time frame. That is the reason for all the different pockets of
> activity...everyone thinks that they can do it better until they get into
> the actual work.  That is the reason for all the competition of solutions.

THIS POINT IS VERY INTERESTING. WE LEAVE IN A VERY OPTIMIZED WORLD WERE
EVERYTHING HAS TO BE DONE BY YESTERDAY. BUT WHEN WE DEAL WITH ACCEPTANCE BY THE
MAJORITY OF A CONCEPT, WE STILL FACE THE FACT THAT WE ARE ALL HUMAN BEINGS AND
THAT IT TAKES SOME TIME TO ALLOW PEOPLE TO BUY INTO. YOU HAVE THEN THREE
SOLUTIONS TO THAT ISSUE: THE DICTATORSHIP APPROACH, WITH OR WITHOUT AN ARMY; THE
CONSENSUS APPROACH IN THE STANDARDIZATION DOMAIN, GIVING PEOPLE TIME AND
OPPORTUNITY TO REACT AND THE NEW APPORACH BY INTERNET: LET THE PEOPLE TAKE THE
STUFF AND SEE IF THEY USE IT. I AM IN FAVOR OF THE LAST TWO OBVIOUSLY OTHERWISE
I WOULD NOT WORK I UN.

THIS BRINGS ME TO THE FOLLOWING POINT YOU MENTION:

A single flavor of XML--I can think of nothing so damming. 
XML was created (I wrote the mission statement) to overcome
the "wrong thinking" that any single semantic language could
be "the solution". At that time, HTML was thought to be within
one or two revisions from being the universal markup language.

I MUST NOTE HERE THAT, AS PER TODAY, TIME PROVED THAT COMMENT WRONG: HTML IS THE
UNIVERSAL MARKUP LANGUAGE, AND EVEN OUR SECRETARIES KNOW HOW TO USE AND USE IT.
IF WE WOULD ACHIEVE THE SAME RESULT WITH XML, BINGO WE HAD THE JACKPOT. AND I
CAN DREAM OF OUR SECRETARIES CAPTURING DATA AUTOMATICALLY FROM THERE E-MAIL
DOCUMENTS OR WORDPROCESSOR ETC.


2  =======================================================================
WHICH APPROACH?

> Ray Walker told me in San Jose that we have got to win over the EDI crowd
> to make this work.  Yet Ray is very adamant about forwarding his CEFACT
> UML /OOedi work under the banner of ebXML.
> How do you pull that off?  That has been my dilemma.  UN/ebXML will not
> fly any better.  If you build it will they come.  All I ask is that you be
> objective when addressing this problem and not come to pre-conclusions
> that you cannot do both.

FROM MY EXPERIENCE - COMING FROM STATISTICS, I KNOW THAT YOU CAN DO THIS IN
STATISTICS EASILY BECAUSE THEY - THE STATISTICAL OFFICIES DID SINCE AGES WHAT
THE SIMAC (SIMPLE EDI GROUP) PROPOSES: ALIGNEMENT OF MASTER DATA, ELIMINATION OF
NON-CORE DATA, USE OF UNIFORM AND UNAMBIGUOUS STANDARD INTERNATIONAL CODES AND
CLASSIFICATION SCHEMAS. THIS IS WHY THEY CAN DO MODELLING SO EASILY! AS LONG AS
THIS IS NOT DONE ACROSS INDUSTRY SECTORS, THERE IS NO HOPE THAT UML - TECHIE
FREAK INTERESTING- CAN DRIVE THE BUSINESS AND YOU END UP IN BEING ABLE TO BUY
ONLY TOILET PAPER.....HOWEVER IN A VERY SOPHISTICATED WAY.

FROM THE ABOVE, IT BECOMES CLEAR TO ME THAT YOU CAN HAVE A MODELING APPROACH
WITHIN VERTICAL BUSINESS SECTORS (GIVEN THAT YOU HAVE THE RESOURCES, EXPERTISE
AND MANAGEMENT DECISION TO DO SO) OR WITHIN WELL STRUCTURED SECTORS SUCH AS
STATISTICS. HOWEVER, WHEN IT COMES TO THE 80/20, THEN THIS APPROACH WON'T TO MY
BELIEF BE OPERATIONAL IN THE COMING TWO YEARS. SO I WOULD CONCLUDE THAT WE HAVE
TO HAVE A TWO WAY APPROACH: THE REAL 6-18 MONTH OUTPUT BASED ON THE CONCEPT OF
'DOCUMENT' AND THE OTHER.

THE 80/20 APPORACH BASED ON THE CONCEPT OF DOCUMENT IS VERY APPEALING TO ME,
GIVEN ALL THE WORK THAT WAS DONE BY OUR ANCESTORS IN TRADE FACILITATION: VERY
VALUABLE WORK THAT SHOULD BE PUT UP TO DATE WHEN NEEDED. I STILL DREAM OF A NEW
UNITED NATIONS RECOMMENDATION ON XML DOCUMENTS EQUIVALENT TO OUR RECOMMENDATION
1 ON STANDARD PAPER LAYOUT FOR TRADE DOCUMENTS. ALSO, IF WE CAN PROVIDE SUCH
THING, WE WOULD CATER FOR ALL THE PEOPLE OUT THERE WHICH STILL USE PAPER
DOCUMENTS FOR THEIR DAILY BUSINESS. THE USE OF A DIFFERENT TECHNOLOGY WOULD BE
VERY STRAITFORWARD. ALSO, DON'T FORGET THAT A LOT OF THIS PAPER WORK CAN'T BE
CHANGED SO EASILY AS A PIECE OF SOFTWARE. i GIVE YOUR FOR EXAMPLE THE
INTERNATIONAL CONVENTION ON TRANSIT WHERE THE DATA REQUIREMENTS ARE VERY DEFINED
BY THE VARIOUS CUSTOM ENTITIES. TRYING TO SELL THEM UML AND MODELLING BEFORE
IMPLEMENTING IT WITH XML WOULD BRING US ALL TO OUR RETIREMENT AGE, EVEN THE
YOUNGEST ONES!. IN THAT CONTEXT, I UNDERSTAND YOUR WISH  TO SEE THINGS DONE
BEFORE YOU RETIRE - ONLY TWO YEARS TO GO, I NEVER THOUGHT YOU ARE SO OLD - AT
LEAST YOU DON'T LOOK LIKE! 

AS A LAST NOTE ON THESE TWO APPROACHES YOU SHOULD NOTICE THAT AS A FRENCHMAN
USED TO DESCARTES AND HAVING SUCK AS CHILD THE MODELLING MILK OF MOTHER FRANCE,
I LEARNED A LOT IN THE CONTACT OF MY ANLGOSAXON FRIENDS WHO ARE FAR MORE
EMPIRICAL THAN I AM. I AM CONVINCED THAT THE RIGHT WAY IS SOMEWHERE IN BETWEEN
AND AM ASTONISHED BY SEEING ANGLOSAXON'S BECOMING MORE FRENCH THAN THE FRENCH
THEMSELVES. 


HOWEVER CONCERNING THE 80/20 APPROACH BASED ON THE CONCEPT OF DOCUMENT WE NEED
TO SPELL OUT HOW THIS WOULD/SHOULD BE DONE. I AM DEFINITELY AGAINST THE
MIGRATION PATH TO ALLOW LEGAY STUFF. THE WORK LEGAY COVERS GOOD THINGS AND ALSO
A LOT OF RUBISH, I HAD TO DEAL WITH SOME CODE SETS THAT CLERALY SHOW THE
BULLSHIT THEY CONTAIN AND I WOULD NOT LIKE TO FORWARD THIS IN XML. I AM
THEREFORE REQUESTING THAT THIS APPROACH TAKES ON BOARD THE PRINCIPLES OF
FACILITATION, DATA MODELING, SIMPLE EDI BY SIMAC, THE EXPERIENCE GAINED BY THE
T8 HARMONIYATION GROUP IN UN/EDIFACT AND THE WORK OF ISO JTC1 SC32. WITHOUT
THAT, WE WOULD PROVIDE THE WORLD WITH GARBIGE.

  
> Consolidating the existing EDI crowd before moving forward would be a
> coupe.  I think it can be done conceptually in a weeks time if we do not
> waste time normalizing and restructuring the semantics but focus on the
> method and framework and coordination. (XML is just a vocabulary and
> syntax.) This provides us with a single migration path to the long term
> ebXML solution if and when we get one. 

I FULLY AGREE WITH YOUR POSITION HERE EXCEPT THAT THE VOCABULARY IS MISSING AND
THAT IT IS A POWERFUL SYNTAX - I REPEAT MYSELF: IF EDIFACT SYNTAX WAS DESIGNED
AS XML, WE WOULD NOT BE HERE PRESENTLY.