MnM Minutes CC 20091016a
M&M Conference Call 10AM Eastern Time Joint with ITS Work Group (Date above)
Discuss proposals for issues surrounding implementation of Data Types R2 in the absence of a new ITS for Structures.
Mead Walker, Woody Beeler, Grahame Grieve, Paul Knapp, Gunther Schadow, Lloyd McKenzie, Charlie McCay
Proposals for Implementing Data Types R2
There was copious discussion on the M&M list titled Collection Issue implementing Datatypes R2. Late in thaty thread, Grahame Grieve said:
- Anyhow, I would like to propose a joint meeting of MnM and ITS at 10am
Eastern (US) on this Friday where this issue will be resolved. (specifically, MnM hosting, ITS as guests).
- And would like to move the following motion at this meeting:
- As an interim position, while ITS R2 progresses through it's development
cycle, we will only use the simplified form in ITS R1 where the CIM indicates a collection type (BAD, DSET or LIST) (which is scope limited to serialising CIMs). In these cases will not be possible to indicate the type or nullFlavor of a collection, as has always been true. Where the CIM type is COLL<T>, the implementations should assume BAG<T>. Where the CIM type is ANY (i.e. Observation.value), the full form must be used. This decision preserves the intent of ITS R1 with regard to conflation of singletons and collections. ITS R2 should reconsider this issue in a context where wider change is possible.
Meeting and motion
After much discussion, the attendees unanimously approved to the following motion made by Grahame Grieve and seconded by Lloyd McKenzie:
- Delay the formal migration to Data Types R2, for one (or 2) ballot cycles, in order to allow for a Structures ITS R2 whose only changes are to incorporate the full scope of Data Types R2 into ITS Structures. (Beeler)
- Start an immediate ballot on a Normative ITS Structures R2 to start Balloting in the January 2010 Cycle (McKenzie, Grieve)
3. Start an Informative ballot for Data Types ITS Release 1.1 in January 2010 Cycle. (Grieve, Schadow)