Re: UBL Beta -> 1.0 schema approach

[Tim McGrath <tmcgrath@portcomm.com.au>]

At the plenary yesterday it was discussed about the Sept17th NDR 
checklist and how much 1.0-beta and hence 1.0 production is away from these.

I have tried to apply my comments from Draft L of the NDR document to 
the Sept 17th checklist and came up with the following.
[NB bear in mind my original comments were based on the document 
descriptions not just the rules and this cross-reference may not have 
captured all issues]

 From Sept17th NDR checklist [NOT including any rules marked "NEW"]...

GXS1 (sept 17th) is now GXS2 (Draft L) - we have never done this and 
still dont. I seem to remember some discussion about how stylesheets 
could strip these out anyway.
MDC1, MDC2,MDC3, MDC4 - are all out of scope - they are modeling 
guidelines not mandatory XML rules.  i do believe we practice most of 
them.  the only one that is problematic for 1.0 is MDC3 which is 
unenforcable anyway.
GNR5 and GNR6  - are wrong according to UBL agreement in May 2003 (my 
comment marked NDR50)
ELN2 - does not  agree with current practice in 1.0-Beta (my comment 
marked NDR57)
ELN3 and ELN4 - were part of  a discussion with Mike Grimley and I think 
there are plans to change this - currently it is ambiguous but i think 
1.0-Beta does not comply with it as written in the checklist.(my comment 
marked NDR57)
ATN1 - the CCT schemas do not comply with this rule.(my comment marked 
NDR60)
CTN2 - we dont do this and it is the basis of many follow on rules (my 
comment marked NDR55)
NB. The Schema Structure rules (SSM1-15) have been re-numbered in draft 
L so it is hard to reference them. but SSM4 (in draft L) was SSM8 (on 
Sept 17th) and relates to my issue NDR36 - i suspect the CCT schemas 
dont comply with this.
RED2  - this is not an enforcable rule nor is it an NDR (similar to MDC 
rules above) - i am also not sure we can claim compliance with 1.0-Beta 
(my comment marked NDR84)
DOC1- we dont comply (my comment marked NDR40)
DOC2 - (sept 17th) has disappeared. this is OK, it related to code list 
supplementary components - which has been superceded by CLSC (but i 
guess it means we dont comply with it)
CDL2, CDL3, CDL4  and CDL8 - are LC issues for dealing with code lists 
and hence out of scope (my comment marked NDR 77 and NDR 79).  it does 
not address the stock,standard placebo concept and so i suspect we will 
not complywhen we come to implement CLSCs architecture.
CDL5 - i now realize this is also not an NDR issue and not practical 
either - we may want to not enumerate code list values. (e.g. UN/LOCODE 
has more than 38,000 entries and may be used several times in a document)
ELD1 (sept 17th) is now EDL2 (Draft L) -  I dont think we comply to this 
and dont think we should (my comment marked NDR28)
ELD2 (sept 17th) is now ELD1 (Draft L) - out of scope for NDR and i dont 
think we comply (my comment marked NDR21)
ELD3 - needs rewording to fit UBL terminology(my comment marked NDR68)
ELD4 - needs rewording to fit UBL terminology(my comment marked NDR70)
ELD6 (first one) - should be confirmed with CLSC (my comment marked NDR73)
NB the checklist has two rules marked ELD6
ATD1 - may prohibit us using qualified data types and code lists (my 
comment marked NDR74)
ATD6 - needs clarification we may not comply (my comment marked NDR75)
CTD1- needs rewording to fit UBL terminology(my comment marked NDR68)
CTD5, CTD6, CTD7 and CTD8 - are follow on from CTN2 and we are not 
compliant (my comment marked NDR64)
CTD16 - appears to have been dropped from Draft L except it still 
appears in the Appendix (table A12). no explanation is available but i 
dont think we comply for the same reason as CCT element names do not use 
Dictionary Entry Names.
IND5 and IND6 - i dont think instance rules are in scope for NDR. they 
are either part of implementation 'best practice' recommendations or 
they are not necessary, and certainly cannot be mandated.  but as this 
has no bearing on 1.0 compliance we can ignore them for now.

I hope this helps us keep on track for "what we need to do to have a 1.0 
release with matching NDRs".

-- 
regards
tim mcgrath
phone: +618 93352228  
postal: po box 1289   fremantle    western australia 6160