UNSPSC - whose in charge anyway?

[David RR Webber <Gnosis_@compuserve.com>]

These industry classification codes http://www.unspsc.org were
originally designed for governmental statistical gathering on 
GNP and productivity measuring - stemming from the US NAICS 
codes and similar.

As with other things they are being pressed into service for XML
based business transactions.  This appears to be a classic case
of 'close enough for government use' - but needs more for really 
robust business use.

Being a 2 digit pair means everything has to fall into one of 
99 value slots in the bucket.  This is just not reality.

I'm seeing switching to alphanumeric pairs at least buys you
a ton more bandwidth.

But who makes this decision?  Do implementors just go there?
RosettaNet now has Rosetta UNSPSC codes, and then they
map those back to UNSPSC codes.  Clearly there's a better way.

The whole issue here that is not being addressed is code tables
generally.  Plenty of focus on Schema - but now we moving closer
to a use model right there, we need to solve this bottom layer 
question definitively - particularly so industry groups evolving
XML based exchanges have a reference point of contact.

Maybe we need an ebXML WG on code tables?  Liaison to 
OAG and ISO would make sense - or maybe a joint effort 
between sich bodies?  Its not really a question of 
re-inventing anything - there are plenty of ISO and other 
standards out there - but we do need a central point - and also
XML'ized versions of these reference tables.  ebXML will provide
the API - but how is the secretariat?

Thoughts?

DW.