NDR 2 observations

["G. Ken Holman" <gkholman@CraneSoftwrights.com>]

Hi folks, I have three questions in this note for 
which I would like to get feedback from committee members.

During the Atlantic call I brought up two issues of the NDR 2 document:

    http://www.oasis-open.org/committees/download.php/22992/UBL-NDR-2-20070207.pdf

(1) Code lists - chapter 6

This chapter does not describe the meta data 
information items we have used for CodeType and 
IdentifierType.  Someone using the NDRs to 
convert spreadsheets to schemas would not be guided in the meta data to use.

Now this might be okay ... do we want to leave 
open to implementers of UBL NDRs the chance for 
them to use their own choices of attribute meta 
data for code list and identifer information 
items, or do we want to document those 
information items as we have included them in our 
schemas (which might lead implementers to feel the need to copy their use)?

The comments currently in the chapter are 
accurate with respect to the *valiation* of 
information items based on code lists and 
identifiers using enumerated lists of values, but 
given the role of the NDR to direct the synthesis 
of schemas from spreadsheets, I think two questions need to be asked:

(Q-1a) - should we document the meta data 
components we are declaring for code lists and 
identifiers in our schemas? (I think we should)

(Q-1b) - should we indicate the meta data 
components we are using necessarily must be 
implemented or are they merely an exemplar that 
implementers can choose to follow or choose to 
craft for themselves with their own choices of 
meta data components? (I think I can go either 
way ... mandate it if we want some level of 
document conformance or don't mandate it if all 
we are providing is schema design guidance)

(2) Minor versioning - chapter 3.6

I put out some new ideas on the table in August 
of last year in Montréal regarding minor 
versioning and I would like to discuss these in 
Manhattan.  These ideas were very well received 
by Mikkel's staff during a briefing in Denmark 
and I think are worthy of consideration for us to 
follow.  It is my goal to write these up in detail before Manhattan.

Since we haven't yet released a minor version, 
and the NDR 2.0 document is supposed to reflect 
what we *have* done (not necessarily what we are 
going to do since we can revise the NDR when we 
revise UBL), should we remove some of the 
description of minor versioning in the document 
and replace them with something along the lines 
of "a revision to these NDRs will include minor versioning rules"?

Jon pointed out that we cannot change the list of 
rules as the list of rules was published for UBL 
2.0.  A quick review of the existing [VER*] rules 
indicates they are all being followed for UBL 2.0 
(good thing!) but might have to be modified 
slightly in an NDR 2.1 for UBL 2.1 ... not sure 
... it depends on whether one equates the version 
of a schema with the version of the document 
element of that schema ... don't worry, I hope to 
make this all clear in my upcoming contribution to the Manhattan meeting.

My concern is that other implementers of NDR 2.0 
may publish minor revisions of their work and 
their approaches may differ from what we decide to do in NDR 2.1 for UBL 2.1.

(Q-2) - should we touch chapter 3.6 at all or 
leave it as is until an NDR 2.1 accurately 
reflects what we eventually decide to do for UBL 2.1?

Thank you for your opinions on these questions!

. . . . . . . . . . . . Ken

--
World-wide corporate, govt. & user group XML, XSL and UBL training
RSS feeds:     publicly-available developer resources and training
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/o/
Box 266, Kars, Ontario CANADA K0A-2E0    +1(613)489-0999 (F:-0995)
Male Cancer Awareness Aug'05  http://www.CraneSoftwrights.com/o/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal