Title: RE: [ubl-ndrsc] Bill&Eve's proposal on local/global elements
Mark:
If you
look at the CCTs, you will see things like "Identifier".
Take
xCBL NameAddress as an example. We have two child elements that are
"identifiers". One is a simple string type, called "external address
identifier." The other is a more typical identifier that is used to supply a
coded, official identifer for the addresee - a DUNS number, for example, using
fields to specify code lists, etc.
This
may not even be the best example, but what it points out is this: we have a
choice between making everything be represented by a single XSD structure,
designed to agree ith our list of CCTs, in which case we get bloated structures
with lots of unneccessary optional elements, and a consequent lack of ability to
automatically validate, or we have more than one structure for certain of the
CCTs.
This
is what I was trying to say with the phrase "variety of structures." Still
nothing that couldn't be handled with a mapping table,
however.
Cheers,
Arofan
-----Original Message-----
From: CRAWFORD, Mark
[mailto:]
Sent: Thursday, January 31, 2002 8:00
AM
To:
Subject: RE:
[ubl-ndrsc] Bill&Eve's proposal on local/global
elements
Arofan wrote -
> I anticipate
having a lot of discussion with Tim and other
>
members of the
> LCSC about this issue. This will
focus on the mapping between
> CCTs and the
> information needed to fully document schema. In some
cases,
> there will be a
> clear mapping between the CCTs as described by ebXML and what
> appears in UBL
> - in
other areas we may need a variety of structures. This
> should be brought
> clearly to light as
we work up the next "harmonization" example.
Agree in principle. What do you mean by a "variety of
structures"?
Mark
