Is there any such thing as common horizontal vocabulary (CHV)

[Todd Boyle <tboyle@rosehill.net>]

There are several problems faced by SMEs (and independent software 
vendors in SME space).

1. Lack of a common horizontal vocabulary (hereinafter, CHV) for simple 
things like orders, payments, and billing.  A common dictionary of data 
elements and small aggregates such as product, party, contact, and 
address are needed.  Furthermore, a common set of basic documents are 
needed.

2. Lack of choreography.  In my inexpert opinion, ebXML TRP, TP and BP 
are suitable common messaging and business process solutions for SMEs if 
they are reconciled to work together as a whole.   Other choreography 
solutions exist, both proprietary and within standards universes.  Without
a business process solution, SME integration is dead.  It won't happen.
http://lists.ebxml.org/archives/ebxml-bp/200103/msg00022.html et seq.

3. Lack of any secure platform.   Privacy, authentication, signing, and 
nonrepudiability are integral to choreography solutions.  But they are 
meaningless on a platform like Windows which intruders can penetrate, 
and will certainly penetrate when money is at stake.  SMEs cannot 
maintain perimeter firewalls and "perimeter solutions" are anyways prone 
to catastrophic failure if the workstation is insecure.

If you want SMEs conducting business online, you're going to require a 
separate device laying on the table outside the PC, such as a smartcard 
reader or MeT enabled cellphone, having an integrated lockbox for 
signing keys, and display and keypad for a PIN which are in silicon and 
outside the possibility of being commandeered by hackers.   All of the 
interactions in the PC would be sent for approval to the device, signed, 
and forwarded to destination.

Problems 2, 3, ...n are outside the scope of this email.  

Let's discuss the problem of business vocabulary for SMEs which is 
stuck, right now.  ebXML does not provide a common horizontal 
vocabulary.  Now, EWG has taken over the role of defining data 
elements.  EDI community has no track record or natural constituency as 
a provider of common horizontal vocabulary.  UN/CEFACT delegated 
its process to ebXML for 18 months and ebXML created no CHV. 

The ebXML Core Components design, based on vocabulary whose meaning 
depends on context is adverse to the interests of SMEs.  The ebXML 
CC design  is unusable by SMEs for the immediate future, particularly,
at runtime. 

SMEs will converge around some other CHV and it will become entrenched 
long before ebXML context runtimes are available for SMEs at favorable 
prices and commercial circumstances (access to underlying markets of 
suppliers and customers on an equal basis with the global 500 companies.)

The abdication of ebXML from context-neutral vocabulary has removed 
pressure from the existing providers of CHVs such as OAGI, RosettaNet, 
and xCBL to converge their vocabularies.  As a result they continue full 
speed to advocate their own vocabularies over others and to sign up 
users as fast as they can.  That's fine--these groups are the best hope 
of SMEs right now, for common vocabulary.  SMEs recognize on some
level their vocabular must fit harmoniously within larger vocabularies.
But while the Midrange CHV providers rearrange deck chairs, neglecting
SMEs has certain risks. 

Just to focus the mind, realize that the first battle is already over. 
Account aggregation won.  It has *millions* of users.  For example, read
http://www.google.com [ yodlee umonitor verticalOne corillian 2001 ]. 
These are a disaster for any data integration.  Anything the user does 
at these portals stays there.  The financial services companies only do 
payments and other basic services, but they're here today, so they got 
all the restless adopters and they are VERY sticky.

The next biggest winner of SMEs may be the monolithic webledgers 
(business suites) such as Intuit, ePeachtree, Microsoft and Netledger. 
These will all have XML interfaces and the trend is very clear they will 
all have desktop programs interfacing with the portal in the sky (ONLY 
THEIR OWN PORTAL probably.  It does not matter, that they have "open XML 
interfaces" if the vocabulary and the whole choreography is proprietary, 
their security is bound into it, it requires a VAR etc. )

SMEs will not use a webledger until it makes bookkeeping automatic, as 
Biztone, bAPort, eLedger, etc. have found to their chagrin.  SMEs will 
sign up in large numbers only when their payables, receivables, bank and 
inventory movements are synchronized with th trading partner. The 
'aggregators' won't succeed at this. But the monolithic webledgers will 
succeed.  The danger right now is that Intuit, alone, will end up 
running the show. http://developer.intuit.com and see the video by Scott 
Cook, and of course, look at the QBXML schemas.  Intuit has 
approximately 90% market share in the US SME market bwa ha ha! Nobody 
else is even close.  This game is theirs to lose and they are not fat 
and dumb like the telcos or Pan Am airways etc.  They are very sharp.

Nevertheless, the emergence of SMBXML, QBXML, Intacct XML and eLedger XML 
offers a potential CHV for SMEs.  These XML vocabularies are almost exact 
replicas of the user interfaces for modules in those proprietary small 
business software.  But because SME software has a natural centrifugal 
forces around common business gestures like composing an invoice, 
ordering goods, receiving a payment, etc., their XML schemas emerged as 
having much more similarity than what you see in the midrange XMLs 
(OAGI, RosettaNet and xCBL) or the EDI community. 

Because the new XMLs for the webledgers are single vendor solutions, 
they each reflect their own unified models.  Not only sharing a common 
data dictionary, but sharing common assumptions for user semantics, 
interface behaviors, data storages and so on.  You guys are always looking
for a unified model, and telling me there's no such thing as a unified
core of XML vocabulary.  These new XMLs prove otherwise. 

There is great potential, at this hour, for the midrange XMLs and EDI 
community (OAGI, RosettaNet, and xCBL) to embrace these valuable new 
XML vocabularies as the natural CHV for small business, i.e. the 
non-context-sensitive, common layer within ebXML.  

Let's get working.  Let's populate the ebXML registry with a CHV.

I could be wrong, I often am.  Tell me where I'm wrong!

Respectfully,
TOdd