Minutes of Atlantic UBL TC call 30 September 2004

[Anne Hendry <anne.hendry@sun.com>]

The UBL TC met at 15H00 - 17H00 UTC WEDNESDAY 6 OCTOBER 2004.

Present:
  Jon Bosak (Chair)
  Mikkel Hippe Brun
  Mark Crawford
  Michael Dill
  Stephen Green
  Anne Hendry
  Bill Meadows
  Lisa Seaburg
  Sylvia Webb

Regrets:
  Mavis Cournane


LOCALIZATION

  See CNLSC, KRLSC, JPLSC, and ESLSC reports sent to list this week.

WORK LIST UPDATE (http://ibiblio.org/bosak/ubl/worklist.xls)

STANDING ITEMS

  Calendar review: http://ibiblio.org/bosak/ubl/calendar.htm

    See calendar.

  Event reports (see calendar)

    - Mikkel gave presenation at Brussels meeting on UBL, NDR, and CCs.

    - Mark attended OTA on 21 Sep
      International Managers Colloquium on 22 Sep
      Dept of Navy (to discuss NDRs) on 27 Sep

    - IDA eprocurement expert group meets next week

      Speaking of ...
      The UK Office of Government Commerce (?) posted a message to IDO,
      OGC, ISSS, saying 'good input but don't have resources' and shouldn't
      be brokering requirments of all EU stakeholders.  Told them 1.1
      timeframe and invited them to give us input that we can align
      within that timefream - would have to be in COO format (doc types,
      how created from library, etc).  This will be on their agenda for
      next week.  So the invitation is out there, and have been clear on
      timeframes - said if we could get them by April/May we would 
include them.

  Liaison reports

    Report on discussions with OASIS Tax XML TC;
    call for participation in joint team to work on XBRL in UBL

      Jon has been in diccussions with the Tax XML TC since November,
      but scheduling issues interfered with their frequency.  Now that
      we have appointed a formal liason, it would be appropriate to discuss
      Tax XML TC requirements.  To date their requests have been for us to
      embed XBRL data in UBL messagaes.  We can see the sense in creating
      accounting reports using XBRL from data encountered after a UBL
      message has been received, but it's hard to imagine a need for
      embedding it within the UBL document as they seem to be asking.

      Sylvia mentioned that Tax XML group was intersted in two new items:
      Certificate of Residence and Indirect Taxation (rather than supply
      chain components).

      We need a cross-TC technical team to address all these things,
      incluing IP issues.

      AI: Jon will propose to the team the formation of a group to
      investigate.  Initially the team will be composed of Sylvia
      and Mark.  Others who are interested should voice their interest
      to Sylvia and Mark.

    eBSC representation after October

      Jon has had informal discussions with eBSC; they are putting
      together a matrix of ecommerce specifications.  Mark will handle
      the next meeting.  Monica has been attending, but is finding
      scheduling problems.  Mark mentioned that Jim Wilson is attending.

      AI: Jon will check with Jim.

    Other liaison reports

      Mikkel: At eGov meeting this week; colleague presented proposal
      to start work on NDRs starting with ATG NDRs and look at the
      requirements of localization of CCs.  This could make use of
      the localiztion paper from the UBL TC.  We could provide
      valuable input if UBL is willing.  Mikkel will lead the
      subcommittee that will make recommendations.  Agreed.

      Mark: Calls for participation in next rounds of CCTS.
      The right approach is to do the next round of CCTS work as
      a partnership between TC154 and UN/CEFACT so CCTS retains
      a broad flavor.  Would like OASIS at the table.  This falls
      under the broader question of an OASIS - CEFACT relationship.
      Jon mentioned that the Chair of the Technical Managment group
      will meet with Patrick and Jamie to discuss issues, which is
      moving the relationship in a productive direction.

  NDR

    We need  a plan for resolving NDR comments during the comment period
    so we can finish the document.  We need someone to compile a comments
    list.  Comments period ends on 17 September, so need someone after this.

    AI: Jon will post an announcement to the list about the formation of
    the team, which will be led by Mavis Cournane.

  HISC report

    None at this time.

  SSC report: see below under "EDIFIX proposal"

  Code List Team report

    Tony has talked to Marty who will review Tony's latest document,
    but no other comments have been received since it was put out for
    review.

    Mark Crawford mentioned that he and Mark Palmer have set an agenda
    for the next eBSC meeting and have agreed to devote a couple of hours
    on the second day to code lists.  Would like to make Tony's XML 2004
    paper (posted to the UBL a week or so ago) part of the reading material.
    Tony agreed to this.  Mark will make sure we get minutes of the meeting.

  TBG17 Liaison Team report

    No Liaison report on the Atlantic call.

FOR DISCUSSION THIS MEETING

  EDIFIX proposal

    The SSC discussed this straw proposal last week:

      http://lists.oasis-open.org/archives/ubl-ssc/200409/msg00009.html

    as a result of which, Sylvia Webb sent this diagram to the TC:

      http://lists.oasis-open.org/archives/ubl/200410/msg00003.html

    and Anne sent additional material from this week's SSC meeting
    which can be found at:

      http://lists.oasis-open.org/archives/ubl/200410/msg00013.html

    Mark voiced concerns about using priorietary solutions - will be
    a black box, keeping data and transformation information inaccessible.
    Another concern is that we may give the appearance of closed market -
    then other tool developers will not even deem an interest.
 
    Jon noted, though, that we do have a need to expose the model in
    some formalism and Sylvia added that EDIFIX has agreed to support
    the spreadsheet as an input format.

    Sylvia added that EDIFIX can help to produce other outputs required
    by UBL and UBL users and, in terms of poulating a registry, macros
    or scripts could be written to popluate a registry from the ss.
 
    Stephen mentioned that he has been testing a new version of UBLish
    for 1.0 and it is more consistent with the spreadsheets than what
    we have now.  It is an excellent tool to use as a second round of
    verification on the schemas.  The schemas UBLish is now producing
    are very clean. w.r.t conformance to the spec and conformance to
    the ss model.  See comments Stephen has sent to the list on results
    of the testing.  There seem to be about 5 diffs.  These are not major
    issues, but they may cause some confusion in understanding for
    localization, for instance.  They should be addressed, minimally,
    with documentation.  Even though this information has come in too
    late to address for 1.0, it allows us to document and shows that
    there is enough information being generated by EDFIX to allow others
    with tools to do the same thing.

    Jon mentioned how great it would be to get this information earlier
    in the release cycle.  If anyone was willing to run XPS then this
    could happen, but we also need the person who knows the tool to help.

    We discussed how many licenses might be needed.  Mark brought up the
    question of how people without licenses will be able to see the
    information.  Others need access to the information, not to change
    it, necessarily, but to see it.

    Sylvia stated that GEFEG is working to provide the ability to export
    any portion of the model or a complete model as a ubl compliant ss 
or xsd.
    Although the Reader provides the ability to view the EDIFIX model and
    it's components, the Reader is configured by GEFEG to point to a
    a specific set of files, it can't just be pointed to any files.
    So each person needing to view or work with EDIFIX would be provided
    access.  It would not be ok to use this license for private work, 
though.

    It's imperative to make a decision on this by next week's meeting.
    We need these questions answered in a proposal and sent to the TC
    well in advance of the next meeting so a decision can be reached then.

    AI: All send questions to Anne (or the SSC) within the next 24 hours
        so they can be reviewed in the next SSC meeting which occurs
        24 hours from now.

  Process for working on proposed additions to UBL

    Jon will propose some specific procedures for working with requests.

    Mark mentioned that there are some bounds set by NDR for 1.1 -
    can add and delete optional items, etc., but not make changes
    to existing schemas that would break polymorphic processing.  

    In terms of direction, we will handle each submission on a
    case-by-case basis.  We will want teams to review for technical
    correctness and sanity check on the way out.  Probalby want the
    whole TC to apporve the final inclusion of something in 1.1.

    Bill suggested we have a screening committee.
    Anne agreed that the entire TC should approve the final changes,
    but that, as Bill suggested, there would be preliminary steps where
    screening would be needed.  Suggested taking a look at the proposed
    process shown in the diagrams sent yesterday as they show a generic
    process such as this and are a good starting point for describing
    our process.  It would be good to capture the entire process, not
    simply EDIFIXs part - they are not that easy to separate out and
    we need to know where the interfaces are between EDIFIX and UBL.

       - How do we decide which new items will be included?  (Form
         an input team? Ask submitters to write proposals and have
         the TC decide? Let the chair decide?)

       - How do we get the submissions into shape for inclusion?
         (Form a team to work on each? Make the submitter do it?
         Call for participation from submitting organizations?)

       - Who reviews the submissions for correctness once they are
         in the proper form?  (The whole TC? A submission Q/A
         team?)

       - How does this tie in with the proposed EDIFIX-based
         development methodology for 1.1?

    Further discussion of questions above will be taken up again next
    week after the EDIFIX and UBL process has been put forward in more
    detail.
 
  Review of Atlantic call minutes (any items not previously discussed)

    None.

  Addition of new items to the work list

    None.

  Other business

    TAB activity re rfc 3121

      Mark announced that TAB actions may change RFC 3121 in a way that
      might impact UBL URN naming.  Mark thought UBL should participate
      in the TAB activity to ensure we are not blindsided by this and
      have an opportunity to give out input.  Jon noted that w.r.t. 1.1
      it is too late to make those types of changes, but if people want
      to participate they can do this on their own.

      AI: Mark send to the list information on the TAB activity and ask
      those that have opinions to participate in giving input.

    Stephen asked for some of the differences that were found by UBLish
      to be noted in release notes.  Jon suggested doing some type of
      errata page that we could post on the web site,  Jon will ask
      OASIS about this as a way of handling these discrepancies, and,
      more generallky about some way to push updates onto the site.

    Stephen described a request for one other thing for 1.1 worklist:
      a batch schema.  This would be used if people typically doing b2b
      don't want to send an invoice or order separately they would
      combine several in one message - as a batch, once a week or so.
      Then there would be this batch schema that would allow you to
      present some of the data, such as tax, as a total of the included
      documents.   Mark stated that this is too far in the realm of ebMS.
      Stephen disagreed, and Mark then suggested looking into the
      Generic Header.  If messaging doesn't provide the functionality,
      and the Generic Header doesn't either, then Mark's view is that
      this is a proprietary, application-specific implementation request
      and should not be addressed by UBL.

Meeting adjouned: 17H10 UTC