RE: SMEs : was RE: Party XML Schema Defintions

[James Bryce Clark <jbc@lawyer.com>]

Dave Welsh: 
> Without sounding like a devils advocate, how much is the real 
> busine$$ value of 'discovery' as described below.  In terms used 
> by some, "what's the customer experience?". 

My own informal view of the migration path requirements are as follows.
 
1.   Anything we do has to work for live transaction streams that have
already proven an intere$t in electronic tran$action$ by $howing up in EDI,
RosettaNet, etc.  [This is Dave's issue, I think.]

2.   Our work ought to leave a path to wider adoption.   Some of the likely
candidates who are not currently in the space are those 800-pound gorillas
you mentioned.   They want the paperfree savings, they need to rationalize
their supply chains, and they are sensitive (to greater and lesser degrees)
to the need to not impose disruptive or technically overreaching
requirements on their vendors.

3.   If the standard proves interoperable*, feasible** and popular***,
network effects will cause a marketplace to form.   We don't need to claim
that this will happen;  but should design so as to allow it   

Jamie

* The Y2K-type problem.  You don't have to pay someone $500,000 to migrate
your system  to it, and it doesn't require that you landlock your data in
another proprietary format.   [The ebXML requirements address this, and
Martin Bryan's comments and the BPE project are attempting to put it into
practice.]  

** The fish-bicycle problem.  The modelling sets map to your reality --
they allow you to do some of the high-volume stuff you actually do in
commerce.  [This is Bob Haugen's desk-test project.]  

*** The Gertrude Stein problem.  A combination of early adopters and
prospects for trustworthy resource discovery give you a basis for
confidence that there is a "there" there.   [This is a market effect, that
either happens or doesn't.]