minutes of Wednesday 3 September joint LC/NDR call

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

The UBL Library Content and Naming and Design Rules subcommittees
met jointly by telephone on Wednesday 3 September at 16:00 GMT.

U.S. domestic toll-free number: (866)839-8145
Int. access/caller paid number: (865)524-6352
Access code: 5705229

Attendees:
Jon Bosak,
Mavis Cournane,
Stephen Green,
Mike Grimley,
Eduardo Gutentag,
Anne Hendry,
Ken Holman,
Tim McGrath,
Sue Probert,
Marion Royal,
Lisa Seaburg,
Paul Thorpe

Regrets:
Bill Burcham,
Tony Coates

Agenda:

1. Welcome from Co-Chairs (Tim and Mark) and appointment of Secretary to 
take minutes.
   Anne will take minutes.

2. List Containership
   Presentation of LCSC position paper
   (Please read paper attached to agenda email and be prepared to discuss)

   Tim summarized the LCSC position:
   LCSC needs resolution on this question of (list) containers for next 
week,
   so don't have much time to debate this matter.  Yesterday we spent quite
   a bit of time going over the LC position paper.  The emphasis of today's
   meeting is to redefine the problem for which Rule 116 was created.
   It has never been clear why this issue has been so important to people
   with a document processing background.  We need to have a solution that
   works more practically with the current library.  The overall position
   within the LCSC regarding NDR Rule 116 is that it doesn't work for 
modeling.
   [Jon: Do we get enough benefit from definiing non-semantic containers
   to deal with the issues that arise from that - those problems seem quite
   considerable.]  In terms of the final schemas that are produced, there
   is difficulty in several areas as identified in Tim's position paper
   (this has been distributed to the NDR and LC list - see archives).
   LC is also convinced that there are no performance gains inherent in
   the use of non-semantic containers.  Therefore, if we are to implement
   list containers in UBL 1.0 we need a different, more workable, set of
   rules related to how to implement them from NDR.  This issue is not 
closed.

   Jon stated that it appears that if there is a problem that requires
   containers we have demonstrated that we don't understand the problem
   well enoguh at this point to come up with an appropriate solution
   at this time.  Currently, if someone is quite queasy about having
   naked line items running around inside a box, then the way to deal
   with it is to create a semantic container around these items as
   appropriate, which becomes a semantic list.  If we can show common
   properties, etc, these can be grouped within the kind of analysis
   LC has been doing.
   
   The question of how much containerization is needed is to some degree
   a matter of art.  At this point we should go ahead and do what's 
workable.
   Jon has been convinced that the result of applying the rules is worse 
than
   the problem we're trying to solve.  So his inclination is to drop 
Rule 116
   and go ahead with what we have.

   Tim added that Rule 116 says 'and lacking a qualifying structure',
   but we do have qualifying structures if you consider semantic containers.

   Lisa voice concern that ndr had a quorum vote on this.
   Eduardo suggested putting it out for an email re-vote.
   Jon agreed, reminding us that prior notification has to be sent.
   Sue thought that notification was sent with the email announcing these
      past two joint meetings.
   Jon suggested putting out a notice that in next week's NDR meeting it be
      brought up for reconsideration with the recommendation that Rule 116
      be dropped.
   Eduardo asked, on the expectation that the rule be overturened,
     whether LC has an alternative.
   Tim responded that on expectation that Rule 116 will be overturned,
      LC will continue to work without containers, as it has been to date.
   Mavis agreed to publish this information with next week's NDR agenda and
      will ask for votes in advance of the call for those that can't attend.

3. Scheduling for 1.0 release (see LC schedule attached to agenda email)

   Meeting schedules:
   A suggestion was made that we have a joint meeting every two weeks.
   Several options were tabled on the logistics, since each team also needs
   their individual meeting time. Jon will propose a plan to the lists.

   Release schedule:
 
   - LC will be producing first draft of alpha schemas next week (8 
September)
     This is dependent on the conclusion of the codelist and containership
     discussion.
   - Early in the week following (15 September) we will need a review 
against
     the rules checklist, and any other issues about usage, from NDR.
   - Need sign-off by NDR week of 22 September.
     This will give us the early part of October for instances, 
documentation,
     packaging, etc.  We want to go into the Plenary with a committee draft
     (new OASIS term), so to go into the 4 November meeting with a complete
     packagae and draft allowing the first item of business at Plenary to be
     acceptance of the specification.  Therefore, somewhere in that 
timeframe
     we'd informally want to (before 22 Sepember week) get closure/signoff
     on the appropriate NDR rules to use for the 1.0 release.  The NDR 
rules,
     in whatever format, will need to be part of the package on 20 October.
   - Coming out of the Plenary we'd have Beta.

   QA schedule:
   The only open item is how to document the NDR rules within the package.
   The next QA meeting will be next week, 12 September.

   Coordination of NDR/LC schedules:
   Tim will look at the NDR schedule and respond with any issues.

4. Other Business?

   Codlelists:
   Ken related that he has made a posting to LCSC regarding issues with 
proper
   prototyping of codelists.  He would like to talbe an approach for 
discussion,
   so that by a week from tomorrow the codelist group will sign off on 
something
   (11 September).  This work is stalled until we get xsd files and 
sample instances.
   The pdfs are up to date (from Bill Meadows).  The codelist team is 
pulling together
   all necessary documents and are in a holding pattern until the xds's 
come along.

   Tim will distribute draft 9 of the models by Friday.  This may 
becomes 1.0.

   Ken asked if we are attempting to include validation of codelist 
enumerations?
   The only barrier may be deciding the future use of the adjunct file.  
Ken will
   incorporate this into the codelist report.

   Ken asked that for tomorrow's codelist meeting everyone who has concerns
   please post them and note them in the meeting.  The codelist team needs a
   someone to write portions of the findings that should go to LCSC.  
Ken has
   been focusing on the technical mechanics and needs someone to do the 
writing
   of conclusions of the codelist group for LCSC.

5. Next meeting(s)
   LCSC: Tuesday 9 September 2003
   NDR: Wednesday 10 September 2003

6. Adjourn: 17:35 GMT

Action Items
------------
20030903-01: Mavis to publish list container vote request along with 
reasoning
             from LCSC with next week's NDR agenda and ask for votes in 
advance
             of the call for those that can't attend.
20030903-02: Jon propose a plan for joint meetings.
20030903-03: Tim review NDR schedule and respond if any issues.
20030903-04: Tim send draft .90 models by Friday.