Re: [ubl] A Codelist Issue

[jon.bosak@sun.com]

A note to put this discussion in perspective.

We did in fact discuss the use of substitution groups and abstract
types for code lists in Washington, and we did come to an
agreement among the people attending that we could live with
making an exception to the NDRs for code lists to allow these
constructs.  But we did not have a quorum at that meeting, so that
agreement cannot be considered a real consensus.  Arriving at that
consensus is what we're doing now, out on the TC list where
everyone can see it.

My own position at this point leans toward the views expressed by
Tony and Anne: this approach does look promising, but if we can
get UBL 1.0 done without modifying the rule about substitution
groups to do it and make that decision when we understand its
implications better, then that is what we should do.  (As people
who have worked on projects like this with me before will
remember, I consider this always to be the right course of action
in circumstances like these.)

The two big questions remaining for me right now regarding
substitution groups are these:

 - Is it really true that a future solution that uses substitution
   groups will be backward-incompatible with one that does not?

 - Are we introducing a feature that will make the construction of
   RNG and ASN.1 versions of the schemas impossible?  We chose to
   make XSD the normative form for political reasons, but we are
   under an obligation to respect ISO and ITU standards as well as
   W3C recommendations, and this means not doing anything that
   would make those alternatives impossible to construct.

Jon