Re: [ubl] Re: AW: UBL schema tests

["Stephen Green" <stephen_green@seventhproject.co.uk>]

Anne and All

Yes, you've emphasised that it was agreed that we'd just have 12 specialised
code DT's but at that stage I personally didn't anticipate all the
enumerations in the one
module as we have at the moment. This seems OK to me *But*:-
a) it means, in my view, virtually no need for the separate schema modules
to still exist
b) if we did insist on 12 separate schema modules I'd expect these to
contain the
enumerations rather than the one sdt module (I think this was the gist of
the diagram
Mark drew up in Washington).
c) I prefer a) with the enumerations altogether and a dropping altogether of
the separate one-module-per-codelist modules but there may be some good in
having b) with each
set of enumerations in a single module (it gives the opportunity for the
namespace
use for codelist details and perhaps version control and may help
maintenance of the
codelists - e.g. to update the version of the ISO currency codelist as
currencies change).
d) what we now have is worst of both worlds, IMHO.

Steve

----- Original Message -----
From: "Anne Hendry" <anne.hendry@sun.com>
To: "Lisa-Aeon" <lseaburg@aeon-llc.com>
Cc: <stephen_green@seventhproject.co.uk>; "Tim McGrath"
<tmcgrath@portcomm.com.au>; "David Kruppke" <kruppke@gefeg.com>;
<ubl@lists.oasis-open.org>; "Marty Burns" <Burnsmarty@aol.com>
Sent: Monday, March 08, 2004 8:02 PM
Subject: Re: [ubl] Re: AW: UBL schema tests


> Hi Lisa,
>
> Good question as to minutes.  I took some notes but expected the final
> definitive would come with the final version of the  ndr doc.  Is that
> available?  I haven't seen it.  The only one I have is dated 24
> February.  I did take some rudimentary notes on this part of the
> discussion, if you can glean from this:
>
> ...
> code lists shcmea: secton 3.7 of ndr defines rules for code list.
> develop ss with unspecialized and unspecialized dts.
> here would like to get concensus.
> need rep term qualifier - need 57 rep term qualifiers for sdts.
> no, only for code lists.
> St: need to make names from code lists, not from codes.
> replacing prefix with name that begins with rep term qualifier.
> J: in lcs bin?  Tim: yes.
> A: only for 12?  MD: must be done for all datatypes.
> ST; no, could just do for codes; only need one dt for ones [codes] we
> don't use.
> MD/TM/SG: work out in next 24 hours.
>
> MC: this was Sue and Stig.
> MD: ok, used namespce prefix as rep term qualifier
> MC: for specialized dt.
> ...
>
> So, in the end, what I now recall is that there was a bit of a standoff
> between whether or not we needed to have something for all the codes or
> just the 12, and Michael was to work this out with Tim and Stephen
> (lcsc).  I think Michael was on the edge of agreeing to go with
> Tim/Stephen's suggestion of only having 12 dts (the ones for the codes
> we specify) and using the same (only one) prefix for all the others.  Is
> this what would preclude the need for the specialized cl module?
>
> To hopefully help this process I've attached the graphic that we were
> using at the f2f.  May not be the final version after all the tweaks
> from Mark based on conversations, but pretty much the same, and we are
> currently talking about the two lower-right boxes contained in the lower
> (blue/green) half of the diagram.
>
> -Anne
>
> Lisa-Aeon wrote:
>
> >Stephen,
> >You were not able to see the drawing Mark had for us at the Face to FAce,
> >but it included a Code List Schema Module.  We need to be aligned here,
if
> >there are decisions made, do we have minutes or something from the
meetings
> >so that we can see what happened and why?
> >
> >What I have attached is just the external modules part of his drawing,
saved
> >as a BMP, hopefully you can open and see it.  I am confused, I guess
because
> >I missed so much of the meeting.
> >
> >Lisa
> >
> >----- Original Message -----
> >From: <stephen_green@seventhproject.co.uk>
> >To: "Lisa-Aeon" <lseaburg@aeon-llc.com>
> >Cc: "Tim McGrath" <tmcgrath@portcomm.com.au>; "David Kruppke"
> ><kruppke@gefeg.com>; <ubl@lists.oasis-open.org>; "Marty Burns"
> ><Burnsmarty@aol.com>
> >Sent: Monday, March 08, 2004 12:12 PM
> >Subject: Re: Re: [ubl] Re: AW: UBL schema tests
> >
> >
> >
> >
> >>Lisa and All
> >>
> >>Personally I'm not sure we now have any need of any Codelist Schema
> >>Modules. It was decided that Specialised DataTypes could be used for
> >>those Codelists which need XSD validation.
> >>Is there any need for anything other than the SDT Schema Module for
> >>these?
> >>If so then we would need a set of Codelist Schema Modules but I think
> >>it was decided, wasn't it, that there would be one prefix 'sdt:' for
> >>such codes.
> >>The other need for Codelist Schema Modules might have been for the other
> >>codes but these, it was decided wasn't it, would not have XSD validation
> >>so therefore they would surely not need a Schema -  they would have the
> >>'udt:'prefix, the CodeType being included in the Unspecialised
> >>Schema Module.
> >>If this is wrong could someone explain quickly.
> >>I *think* (not sure) that Codelist Schema Module may be a vestige from
> >>the previous (Montreal) codelist solution (along with the codelist
> >>folders and its subfolders). Someone please correct me again.
> >>
> >>Steve
> >>