OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

ubl message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: Minutes of the Manhattan UBL TC meeting 23-27 January 2006

Hello UBL TC,

Here are the minutes of the Eighteenth UBL TC Meeting, held at Sun
Microsystems offices in Manhattan 23-27 January 2006.  As usual,
plenaries with phone access were held every morning, but in a
departure from our usual format, the closing plenary was held
Thursday, and Friday morning was used for an additional work
session.  Only the plenaries are minuted.

Participants at the meeting are requested to review these minutes
at their earliest opportunity and to respond with corrections if
any are seen to be necessary.

Jon Bosak



   Note that attendance during any plenary call counts as
   attendance at the meeting for membership purposes.

      Members attending in person:

         Peter Borresen    MTuWTh
         Jon Bosak (chair) MTuWTh
         Mavis Cournane    MTuWTh
         Stephen Green     MTuWTh
         Mike Grimley      MTuWTh
         Betty Harvey      MTuWTh
         Kim, Sung Hyuk    MTuWTh
         Mark Leitch       MTuWTh
         Andy Schoka       MTuWTh
         Paul Thorpe       MTuWTh
         Sylvia Webb       MTuWTh

      Observers attending in person:

         Michael Onder     MTu

      Members attending by phone:

         Anne Hendry       MTuW
         Mikkel Brun       MTu
         Ken Holman           WTh
         Kumar Sivaraman   M
         Tim McGrath       M
         Bryan Rasmussen    TuW

      Observers attending by phone:

         Tim Benson        M

   The group was quorate (or as we like to say, "dangerous") for
   all of the plenary sessions, which means that the agreements
   noted below constitute formal resolutions of the TC.


   SBSC report

      AGREED that we are targeting CS as end state for 1.0 SBS.

      ACTION: JB to rezip the UBL 1.0 SBS and submit for public
      review next week.

   UK report

   KRLSC report

   Danish report

   Appointment of UBL TC Secretary

      AGREED to appoint Zarella Rendon UBL TC Secretary for site
      maintenance, membership tracking, and license management.


      AGREED that in the future we will generate the stripped
      schemas in the xsdrt directory as part of EDIFIX schema
      creation rather than producing the xsdrt versions from the
      full schemas using a perl script.


   Customization and profiling

      ACTION: BR to draft a proposal for an "extension" element,
      including namespace management, recommended practices, and
      guidelines for what should be validated in .xsd and what
      should be validated in .sch; target for proposed .xsd ANY
      and extensions: prior to the Atlantic TC call 8 February.

      AGREED to follow up on the idea expressed as "standardize in
      .xsd, customize in .sch".

      [Note below that JB has been assigned the task of drafting a
      customization document later on based on this discussion and
      the extension proposal.]



      We note the issue regarding the explosion of XPaths in the
      2.0 PRD.  We need to look for inadvertent nesting.  Perhaps
      Cheekai has code that can help find the problem.

      ACTION: JB to point Cheekai to Ken's message:


      ACTION: GKH to upload the XPath analysis of
      OrderCancellation to help demonstrate the problem.

   HISC issue: UNLK surrogates

      ACTION: SG to provide HISC with pointers to a few good books
      containing samples of paper documents from which we can
      synthesize UNLK-like layouts for the UBL 2.0 documents that
      don't have official UNLK equivalents.

   Code lists

      We reviewed Ken's proposal for code list value validation:


      AGREED to accept the proposal.

      We note that Section 7.1 of Ken's proposal exposes some
      difficulties regarding the naming of code list metadata info

      ACTION: MartyB and TonyC to review the issues raised in 7.1
      and propose a revised set of metadata names, keeping within
      the spirit of the current naming conventions and best

      ACTION: GKH to write another post on this question as input
      to MartyB and TonyC.

      AGREED to adopt the general principle that structure and
      vocabulary are schema-validated in a first pass, while
      values and business rules are schematron-validated in a
      second pass.  We note that the second (schematron) pass
      won't work without the first (schema) pass.

      Advantages: Cleanly externalizes code lists; streamlines
      code list management; does a much better job of
      value-checking than schema validation alone; recognizes that
      real-world applications always have to do further value
      checking beyond schema validation anyway; provides a
      powerful and flexible approach to meeting the user
      requirements for code list customization; provides a
      standards-based platform for additional user-specified
      business rule checking using schematron.

      Disadvantage: Current tools (XMLSpy, Tibco, WebMethods, ...)
      are not currently set up for two-pass validation.  However,
      the schemas we supply will continue to work with these
      tools, and we are committed to supplying scripts that will
      perform both levels of validation using free software.

      Issue: We have agreed to use the ATG2 UDT schema, and this
      commits us to including the ATG2 enumerated code list values
      for four standard code lists.  We now believe that schema
      validation is the wrong place to do value checking, and
      continuing to use the ATG2 UDT will deprive our users of the
      ability to extend the four code lists enumerated therein.

      We note that the problem is conformance to ATG2 UDT, not
      conformance to CCTS.

      (At this point we recessed for a couple of hours to allow
      the NDR team to evaluate possible ways around this problem.
      The team reported back that they could not see any workable
      alternatives beyond either (a) accepting the ATG2 UDT and
      documenting the restriction this puts on users or (b)
      reversing the decision to use the ATG2 UDT.)

      AGREED that in the interests of convergence we will stand by
      our decision to import the ATG2 UDT, despite the fact that
      we believe it to be the wrong approach to code list
      management, and will document the negative consequences of
      adopting the ATG2 UDT schema for the information of users.

      AGREED that we will not provide any other value enumerations
      in our schemas.

      AGREED that all code lists with known values will be
      provided as XML instances using the genericode schema.

      AGREED that we will continue to use UBL 1.0 names for code
      list metadata information items and will adopt the mapping
      of genericode metadata to UBL 1.0 metadata already worked
      out by GKH in his proposal.

      AGREED that these decisions will be reflected in the NDRs.

      AGREED that no code list supplementary component attributes
      will use FIXED.


   Support for UBL 2.0 implementers

      We arrived at the following preliminary outline for UBL 2.0
      support materials.  This is still subject to revision.  Note
      the embedded action items.

         Standalone supporting specifications

               Owners: MG and MC
               Checklist by end of Feb (note schema gen dependency)
               Doc done by end of March
               Schema/NDR Q/A by end of initial public review
            Code list methodology specification (for 2.0)
                 Owner: GKH
                 Dependency on a sample instance
                    ACTION: SG to have sample done by 13 Feb
                 Done (then) by 20 Feb
               Association transformation script
            Genericode (subset) spec
               ACTION: GKH and TonyC to tell us timeline
            Customization guidelines
               Owners: A group that includes MG, MC, BR
               Start customization in April, finish in May (before
               Would help to have a strawman before April
               ACTION: JB to prepare a strawman by end of March
            2.0 SBS
               Owner: SG (with help from GKH)
                  Dependency 1: revised content and schemas
                  Dependency 2: ebBP and CPP/A
                  Initial pkg using current content for an initial review:
                    end of Feb
                  Second (complete) pkg: mid-June (12 June); changes to 
                     2.0 may force further revision

         The Support Package

            Code list support
               Owner: GKH
               Genericode versions of all the referenced 2.0 code lists
                  Timing dependency on result of initial public review
                     (need to know the code values and related metadata)
               Default code list association file(s)
               Default .sch assertion file(s)
               Default XSLT validation file(s)
               Timing: completion of values/metadata + three weeks
            Pointers to UML and EF data models
            Danish use cases and example instances
               Owner: PB
               Finished by CEFACT meeting in Vancouver (mid-March)
            Sample business rules expressed in .sch
               Owner: PB
               Finished by end of May
            Output formatting specifications (may not include all doctypes)
               Owners: HISC (GHK and ZR, possible DK consultant)
               Timeline: 8 doc types, end of Feb; cutoff end of May
               Dependency: mid-April cutoff for all layouts;
                  sample instances from SBS now; examples from DK assumed
                  by mid-March
               Stylesheets at same time as completion of formatting specs
            Input mapping and XForms
               Owner: BR
               Timeline: don't know; will start right away;
                  cutoff end of may
            Model UBL contract (U.S.)... including CL associations
               Owner: JB
               Cutoff: end of May
            Fiscal accreditation of UBL 2.0 (UK) (an endorsement)
               Owner: ML will find out by 15 Feb whether is possible

         Later standalone products (after October 2006)

            NDR (as a Standard)
            User's Guide to UBL practice
            Implementer's Guide to UBL practice
            Taxation guidelines for invoice (a piece of the above)
               First draft after Feb Tax XML TC meeting in London
                  (including possible content review feedback)
               Possible second draft in June (if many questions from TC)
               Timing: content dependency
               Guidelines for other documents that have tax implications
                  (credit note, debit note, self-billed invoice,
                  self-billed credit note, purchase order)
            Other subsets

   Issues list access and maintenance

      AGREED that we will password-protect the issues input form
      (TC members can use the form, but all others must use the
      OASIS public comment form).  There will be a column/field
      for category (NDR, PSC, TSC, HISC, SBS1, SBS2, Code lists)
      and another one for priority (bug, missing functionality,
      new requirements (RFEs)).  We will try to process issues as
      they come in, with those clearly belonging to PSC handled by
      PSC, etc., and preliminary dispositions recorded in a
      separate resolution list.

   Work/meeting schedule

      We note that the week of 8 May is a trade show week in

      ACTION: JB to ask CEN for advice on the hotel situation.

      AGREED that Mavis Cournane will chair the meeting in

      AGREED to add a TC meeting 14-18 August 2006 in Montreal and
      move the Singapore meeting to November, say the week of 6
      November.  The August meeting will focus on the following:

       - Where we are with the UN/CEFACT transition

          - What practical steps are needed to effect a transition

          - If we can't make the transition, formation of a Plan B

          - If we can make the transition, how to maintain UBL 2

       - Possible actions required by UBL 2 OASIS Standardization

      ACTION: JB to find a host and facilities in Montreal,
      beginning with McGill University.

   TC conference calls during the public review period

      AGREED to devote the second half of Atlantic TC calls to NDR
      work, using this slot for code list work as needed.

      AGREED to hold regular PSC, TSC, HISC, and SBSC calls during
      the public review period.

   NDR progress report

      The NDR editors have reviewed the issues raised by SW on the
      issues list, finished updating the rules, and added a
      section on QDT.  They are now half done generating the
      checklist, after which they will begin a review of the PRD
      schemas against the NDRs.  They believe that the NDR
      document is stable and complete with the exception of issues
      that arise in schema/NDR Q/A.

      Issue: We decided long ago that acronyms had to be approved
      by the TC as a whole, but we don't have a process for such
      approval.  Example: GUID.

      ACTION: PSC and TSC to produce a list of new acronyms (i.e.,
      acronyms used in PSC and TSC content that are not on the
      list at the end of the 1.0 NDR) so that the TC can
      approve/disapprove them.  Deadline: end of February.

      ACTION: JB to find out which of GUID and UUID is an open

   Content team progress report

      Completed the Tax XML spreadsheet.  About 20 percent done
      with the Danish scenarios and have uncovered a number of
      issues for the issues list in the process.  Have formed a
      proposal for documenting business rules (i.e., preparing
      implementation guidelines); BH will make a schema for this.


      ACTION: DK to appoint someone to coordinate with Carol Geyer
      at OASIS regarding marketing/publicity around

       - Releases

       - Approval as an OASIS Standard

       - Marketing after release as an OASIS Standard

      AGREED that the dormant MSC will meet in Montreal to restart

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]