MINUTES: 16 March 2003 NDRSC

["Lisa Seaburg" <lseaburg@aeon-llc.com>]

 

1.  Roll Call and Welcome from the chair (Lisa or Mavis?), assignment of  Lisa to take minutes.

 

Attendees:  Mike Grimley, Dan Vint, Lisa Seaburg, Kris Ketels, Eve Maler, Paul Thorpe, Eduardo Gutentag, Anne Hendrey, Gunther Stuhec,   Arofan Gregory, Jon Bosak.

 

Regrets: Mark Crawford, Mavis Cournane,

2.  Acceptance of minutes from the previous meeting.
http://lists.oasis-open.org/archives/ubl-ndrsc/200303/msg00062.html - No Quorum

3.  Adoption of agenda/schedule planning -

 

Noted: Mark sent an email saying the ATG group will be there everyday, they will need 1 hour daily out of our discussions.

 

Schedule for Face to Face, London:

Monday

Opening plenary  (Lisa will present)

Afternoon:

Danish Bankers (Stig Korsgaard) - Lisa needs to talk to Stig to set time and date.

Discussing the weeks schedule

Local vs. Global

 

Tuesday

AM: Common Core Components Document

PM: Code Lists? finalised??

Context Methodology (overview) (Eduardo)

Containership (after 2pm) (Arofan, Gunther)

    - Would like to invite Tim McGrath and Mike Adcock to attend this section of meeting, ask them when they are available.

Namespace and Versioning (overview)

Embedded Documentation (overview)

 

Wednesday

(We should schedule call in, Eve Maler said she will call in, anybody else?) (7-10am ET noon-3pm UK)

AM: Containership

PM: more Local vs Global

 

Thursday

(We should schedule call in, Eve Maler said she will call in, anybody else?) (7-10am ET noon-3pm UK)

AM: edit NDR Doc

Subgroup: Context Methodology

PM: Gunther's paper on Naming Length and Truncation Rules.

PM: edit NDR Doc

 

Friday (Lisa not there)

Mark to present Closing NDR Plenary

4. Review priorities and assignment of work items (Mavis/Lisa/Mark)

A+ NDR document update (Mark not available to give update.)
A Code list rules and schema template - (in progress, discussed and formulated more rules in this call)
A Embedded documentation (may need to review table format and discuss at face to face. Need to decide about Gunther's proposal.)
A Local vs. Global elements - (to be discussed at F2F)

    Discussion about ATG2 joining in this discussion, we need harmonization of opinions across the two groups.

    Involving ATG2 is important to make sure their views are represented within our discussions.

    ATG2 members have been invited to meeting, don't know which ones are able to yet, will look into this.

A Namespace rules (to be discussed at F2F)
A Versioning rules (to be discussed at F2F)
A Context Methodology (to be discussed at F2F)
B Common Core Components documentation - (in progress)

    Working on examples and putting in the different sections editing by others.

    Would like to be finished before the Face to Face.

    Send to the LCSC, asking for review and comments. (before Face to Face)

B Containership (discussion on this call and at F2F)

6. Containership - (I tried to catch most of this discussion, it was quite lively)

    DV: I was wrong (Dan, just had to put that into writing), in looking at redefine for the example it turns out it does work together, which brings Gunthers proposal back into play.

    GS: My colleagues and myself have been using model groups for document assembly.  In design time, you can use templates (model groups).

    DV: The mechanisim of how to extend complex and simple types works for model groups.

    AG: Does this get complex?

    GS: Not really.

    AG: Can Gunther give me a wrap up of how this is used?

    GS: How can we merge the two papers to get a common paper about containership.

    AG: I will be freeer to work on it next week.

    GS: It makes sense to combine semantical containers, at design time it works.  There would be a template for anybody wanting to design business documents.  It is not necessary to see the containers in the XML instances.  You are creating more hierachies, and more elements.  It makes the document much more complicated and not so processable. 

    AG: There are some containers that are not semantical, all they do is facilitate processing and java binding.  Are we talking about the same thing.

    GS:  Yes, we are talking about similar things.

    AG:  I want to know we will have containers for extending elements.  If I wanted to extend the header information, I want to be able to put my element in the right place with the header information.

    GS:  Model groups you can not do this.  We need to find out if we can extend the model group itself?

    AG:  This could take a long time.

    DV:  This is the same question as when I started out. 

    AG:  How does this work?  If we choose to use model groups for things, can we do this?

    DV:  A couple of weeks ago we reviewed the groups usage.  At that point I thought you could not do the extensions there.  What is the actual steps for getting to the extension capabilities that we need?  So far it looks like using redefine would work.

    EG: We had this discussion in Context Methodology, Dan is going to present how it could work.  We have the methodology for Context, but what we need is what features in XSD Schema will allow us to have what we need.

    DV: If you have other ideas of how this pulls together and what it looks like, I would love to see that.  Redefine is the only way I have found so far.

    EG:  Redefine works in a separate namespace.

    GS:  I did not find a way either.

    AG:  I think there are non-semantic containers that we will want to have in our documents.  For purposes of extension and for processing.  The other thing you need is to encapsulate within the body, any repeatative lists that we are using.

    GS:  you do not define additional element groups?

    AG:  If I have a lineitem I have a list of ScheduleLine which I could have 1 to n, put this ScheduleLines element around these.  This allows for better processing.

    GS: You would see this element in the instance?

    AG: Yes, our positions are not that different.  I am thinking more of processing.

    GS:  It is very complicated to process the kind of hierarchies you are talking about.

    GS: You will have trouble with the database processing with this kind of structure.

    AG:  You don't store the containers in the database, you would know when they are there and not need to store them.

    GS:  Normally we are very clear about how you define your tables in your database.

    AG:  These are rules that are about the document in tranmission, not in the database.

    GS:  If we have too many hierarchies it slows down the seirilization.

    DV:  Hierarchies are good, they are our friend.

    AG:  If we can agree on fairly simple mapping rules on these, then the cost could be marginal.

    GS:  If we are processing many documents in a short time (say 20 million per day), this could be costly.

    AG:  I was saying in my document that we only use these containers in key spots to help with the processing.  This could actually be beneficial to the processing.  How many levels of hierarchies is too many?

    GS:  Probably 10 levels could be too many.

    AG:  We don't want to take it to that extent.

    AG:  I think what we are talking about may not be 1 to 1, but is not far off.

    GS:  I will do some testing next week,

    AG:  We should be able to use the results to come up with some rules for the levels of containers.

    AG:  We will hold off until we see this, lets postpone discussion until then.

    LS:  I am happy with seeing the two papers merged and coming up with a set of rules that all can live by.

    AG:  We may get some push back on this from LCSC.

    AG:  NDR should work it out first, then use the second day to work with Mike and Tim. 

    LS:  I will shift around on schedule the part where we have.

 

AI:  Gunther has to put together two things, semantic grouping, then use Arofan's paper to come up with a reasonable number of levels.

 

Discussion to continue.

 

7. Issues from last week's call - Issues of the ATG2 group joining

Lisa to ask Mark about status of this item.

 

Did anyone read Gunther's paper about element names.  This is added to Thursday at Face to Face.

 
8. Common Core Components - see above.

 

9.  New Business:

CCTS Implementation Verification:  Tim is drafting the UBL response to the CCTS Implementation Verification and need some input on the Core Components Context Constraints Language (section 6.2.4 of the CCTS).  Is it appropriate to say we intend to implement this? 

 

Lisa will email Tim with reply:  In the next phase we have always said we would implement this.  Basically our current extension methdology is predicated on the use of it in the future.   Eduardo agrees with this wholeheartedly.

 

Next Weeks Agenda Items:

 

Discussion of Redefine

Status of CCTS Implementation Verification

Element Naming and Truncation Rules (Gunther)

Common Core Components Document, should be finished and sent out to LCSC for review.

9. Adjourned at 11:50am CDT

 

++++++++++++++++++++++++++++++++++++++++++++++++++++
Lisa Seaburg
AEON LLC
Website: http://www.aeon-llc.com/
Email:  lseaburg@aeon-llc.com
Alternative Email: xcblgeek@yahoo.com
Phone: 662-562-7676
Cellphone: 662-501-7676

 

"Remember that great love and great achievements involve great risk."
++++++++++++++++++++++++++++++++++++++++++++++++++++