Minutes of Pacific UBL TC call 29 July 2004

[jon.bosak@sun.com]

MINUTES OF PACIFIC UBL TC MEETING
00H30 - 02H30 UTC THURSDAY 29 JULY 2004

==================================================================
ATTENDANCE
==================================================================

   Jon Bosak
   Thomas Lee
   Tim McGrath
   Yukinori Saito
   Shin Takagi
   Sylvia Webb
   Patrick Yee

==================================================================
STANDING ITEMS
==================================================================

{Standing_Agenda:6}  Calendar review

   TM: Clarification: the eAC meeting in Seoul is 27-29 October.

{Standing_Agenda:7}  Event reports

   TM: Continuing last week's report from Bangkok

    - An interesting aspect of the APEC symposium was the presence
      of delegates from Chile and Peru.

    - Tim met with a group of about 25-30 people representing the
      Thai government and private sector to brief them on UBL and
      the LSCs.  Somnuk has put out expressions of interest to the
      relevant organizations.

    - Tim talked further with Jean Kubler of UNECE regarding the
      relationship between UBL, CEFACT, and UNeDocs.  Jean
      emphasized the desirability of harmonizing UBL with UNeDocs
      and the UNTDED on its way to UN/CEFACT, the Certificate of
      Origin and the Sanitary/Phytosanitary Certificate being
      given as examples that relate to the APEC initiative.  A
      revision of UNTDED is currently in progress.

  JB: We should make UNTDED alignment an objective of 1.1.

==================================================================
ATLANTIC REVIEW
==================================================================

We spent the meeting reviewing results from the Atlantic call
regarding the issues list.

a.4 (Atlantic conclusion was: Need Mike and others to comment on
    this... Get Tim's input in Pacific call)

    TM: It seems to be agreed that attribute naming prior to 1.0 is
    incorrect.  The attribute in the example should simply be
    code... According to NDR ATN1, we never use the object class,
    only the property term and representation term.  The problem is
    that we then sometimes end up with two attributes on the same
    element having the same name, so the rule is unenforceable.
    This was raised three months ago, and we said that the only way
    to resolve it is to include the object class in order to
    guarantee a unique attribute name.  This does disagree with
    ATN1, but not with the general naming principle that we always
    use only the property term and representation term when that
    does give us a unique name.  This was my argument to ATG2 as
    well.  Before 1.0, we solved this by making up a name to create
    one that was unique.  I am proposing to make this more formal
    and reliable by using object class; all we need to do is add a
    clause to ATN1 that says "except where the property term and
    representation term are not unique, and then we use object
    class as well."  This is in the submission to ATG2.

    JB: Why is there still disagreement on this?

    TM: Not sure that Gunther has read the material.  He is not
    proposing a different solution, just the rule and examples of
    where our practice differs from the rule.  We should change the
    rule.  We should be in agreement with ATG2 on this -- these
    names appear in the CC schemas -- but we haven't gotten
    clarification.

    JB: It would seem that ATG2 should either accept our proposal
    or just define unique names and say that's what they are.

    We agree that JB will post this summary and then follow up in
    the Atlantic call 4 August.

a.5 Agree with finding of Atlantic call.

a.8 We think that the difference here is exactly the difference
    between local and global, and that revisiting this would be
    revisiting the local/global debate.  Done.

e.1 Agree with Atlantic.

f.1 Agree with Atlantic.

g.1 Agree with Atlantic.

h.1 Agree with Atlantic.

h.2 Agree with Atlantic.  The cardinality was originally 1..n, but
    then cases like the one Sylvia raised were brought to our
    attention, so we changed the cardinality to 0..n but forgot to
    change the definition.  So agree that this should be changed in
    both spreadsheet and schemas.

    There might be other instances where cardinality doesn't align
    with the definition; this is a Q/A exercise.

h.3 Agree with Atlantic.

h.4 Agree with Atlantic.  "Department" comes from the OASIS CIQ TC.

h.5 Agree with Atlantic.  We note that providing examples is a big
    work item, but we hope to have have real examples to draw on by
    1.1.

==================================================================
PRIORITY ITEMS FOR THIS MEETING
==================================================================

Our main priority was to review the discussion of the 1.0
issues covered in the Atlantic call; see above.

We also had scheduled for this call a status check on the
following action items:

   TM: Prepare a discussion paper for a meeting with Crimson Logic
      in Singapore 20040718 or 0719

   KT: Align with 1.0 CD and have revised version of COML ready
      for consideration in Copenhagen

   KT: Send scanned images of typical CO paper documents.

   It would be good to get a status report on these items.

TM: Did meet with Crimson Logic [CL].  Prepared a discussion
paper, made the CL spreadsheets into something like the 1.0 ss,
and found that aside from the root of the new document type,
needed only 5-6 new ABIEs, the most significant being Consignment,
because this has not earlier been needed in our basic procurement
scenario.  Adding these new pieces moves UBL into the area of
transport.  Spent three hours with the CL development team in
Singapore, explaining how the ss work and how the CCTS rules
work.  The team sees that Tim's interpretation is still not
complete, so they will put further work into the ss to completely
reflect their requirements.  The plan is to bring the submission
back to the whole TC for discussion once the CL team has had
another pass at it.

Formerly Singapore Network Services, CL has been doing EDI
software for many years.  It used to be owned by the government of
Singapore and now is jointly owned by the government of Singapore
and the national telecom provider.  CL has actually developed
three different document models for Certificates of Origin (one
for a portal, one for UNeDocs, and one for UBL), so we need to
make clear which one we're talking about.  The CL team is aiming
to have the revised submission ready for review in Copenhagen.

TM will follow up on the paper versions of the CO documents and
notes that the CL technical people brought up the need for digital
signatures again.  [This is on the work list for UBL 1.1.]

==================================================================
WORK LIST ADDITIONS
==================================================================

   None from this call.

==================================================================
OTHER BUSINESS
==================================================================

   TM will be available for the Pacific call 5 August.

   YS will not, but will ask Noboruh Itoh to send a JPLSC report.

   TL and PY will not be able to attend next week, but will remind
   William Chan to send a report.

   It was noted that the Japanese translation was announced on the
   ubl-dev and ebxml-dev lists as previously agreed.