Datatypes R2 Issue 31

Introduction

Null flavor co-occurrence constraints need to be more clear

Backwards compability: No change, just better documentation

Discussion

Thurs Q3 May 2006: Motion in favor of this.

Disposition

Well, the underlying confusion here is the contention that setting a datatype to null actually prohibits any properties from being populated. But generally this is not the case. There are some specific cases, and these are clearly documented. In fact, all the required co-occurence constraints are documented.

But that's what we don't document. So the disposition to this is to add a general paragraph saying that.

What are we voting on here? Gschadow 07:35, 24 September 2007 (CDT)

Agree that this seems somewhat devoid of content - but happy to vote in favor of adding a short paragraph as described above Charliemccay 12:00, 24 September 2007 (CDT)

Changed vote to in favor, based on documentation being added that says any components can be populated that don't break an existing explicit "can't be populated if null" prohibition. --Lmckenzi 00:25, 3 October 2007 (CDT)

Despite the fact that I don't like what they end up as, I'll vote in favor of this as it is simply to make what we currently do more clear. LeeColler 13:42, 03 October 2007 (PDT)

Vote:
 For: Grahame, Charlie, Lloyd, Lee
 Against:
 Abstrain:

Links

Back to Data Types R2 issues