RE: GREAT ebXML ARTICLE

[Todd Boyle <tboyle@rosehill.net>]

Alan said,

> You have identified a key problem.  The public affairs challenge 
> we face is to break through the enertia and skepticism.  The 
> succession of public POCs since December has chipped away at some 
> of the skepticism.  But to break through the enertia takes 
> something that makes the kids say 'Cool' or us boomers say 'Groovy'.  
> 
> For what its worth, my suggestion to break through the enertia is to get 
> end-users on board with ebXML.  So far those identified with 
> ebXML have been mainly vendors, consultants, and standards junkies.

Here look at this. 
http://www.xml.com/pub/a/2001/04/04/webservices/index.html

My mailbox is stuffed with things like this every single day.  
Whole magazines full of this stuff. The mags include these
http://www.advisor.com , http://www.xmlmag.com/
http://msdn.microsoft.com/msdnmag/default.asp , www.entmag.com/ ,
http://www.winntmag.com/

Their stuff works.  Because they have 1,000 people working on it
50 hours a week.  That takes money.  Why do they spend so much 
money?  Because they will own the category, just like they own
the desktop and the wordprocessor and spreadsheet.  Show me any 
product that remotely competes with Biztalk server for $1000.

If the rest of the industry (Sun, IBM, et al) want ebXML to happen,
they should get together, and hire everybody who built ebXML, and 
call it a joint venture.  Hire a bunch of credible, experienced 
people to do nothing but explain what ebXML 2.0 is doing, and 
hire a bunch more people to test it, etc.  Whoever you can't hire, 
you can almost certainly borrow for 3 months.

I honestly think that if you put all these people into an office
building someplace for about 90 days, you would have something 
usable because unlike 18 months ago there is a shared vision.

You need a Manhattan project.  With some urgency.  Forget about the
UN/CEFACT process.  Leave it behind.  Honor EDI, and other legacy.
but uncouple ebXML dev. from the sponsoring organizations.  Take
it to them in weekly reports.  Give them 30 days to respond then
move forward.  If Europeans are so worried about fairness, then they
can buck up some money and send people to work on the project.

For their part, I think UN/CEFACT, DISA X.12, OAGIS, Bolero/Swift,
CommerceOne, and everybody else who is competing to build basic,
horizontal vocabularies should agree to sunset all of them, at a 
point certain, 12 months from now, and commit to Core Components.

Jump off a cliff together like Robert Redford and Paul Newman.  

Do the work, to reconcile the models.  They are not that incompatible.  
They are just different.  And reconcilation is a lot of work.  In
other words, resolving the models is just a matter of money. 

The reconciliation job is a classic matrix management problem--two 
dimensions of teams are needed, obviously the industry teams D1, D2, etc. 
and also, the base classes teams.  So the job is impossible without a 
groupware infrastructure.  Our infra. at ebXML was bad the first 6-9 mos,
still *mediocre.*  I hope we never have another major ebXML conference
without an 802.11 LAN for coordinating the work and changes, of the 
teams, and providing timely views to the attendees.  This would cost
what?  $10,000 for a fileserver, 4 NAPs and 50 cards.  Peanuts.

Honestly I think ebXML needs a Tim Berners Lee, your Linus Torvalds.
Somebody who would keep it all in their head and steer the ship.
I would even settle for a Lee Kwan Yew.  But it would have to be 
somebody who really knows where we're going technically, who knows
why we're going there.

Otherwise get used to biztalk and other proprietary solutions because
that's where I think Small and Medium Business is going, Americans 
and Europeans alike. What other choice do they have?  None, at this 
point.  You will have EDI at the enterprise an supply chain, and
you will have 28 million small businesses in proprietary desktop
and web services applications tied together with hundreds of millions
of discrete, customized connections-- all different -- meaning, you
can't do business without a Biztalk server and a WHOLE lot of 
translation and mapping going on,

Hope I haven't made a complete fool of myself, again,
respectfully,

TOdd