Stream: fhir/infrastructure-wg
Topic: Normative MetadataResource
Grahame Grieve (Feb 17 2022 at 22:21):
MetadataResource is normative. But RelatedArtifact is not. I don't know how this slipped by us in R4.
Do we have a list of Datatypes that are going normative in R5? RelatedArtifact needs to. Though I think it's underdone...
Lloyd McKenzie (Feb 17 2022 at 23:48):
It's definitely underdone. Would it be a technical correction to flag it as non-normative given that the data type is non-normative?
Grahame Grieve (Feb 18 2022 at 01:31):
well, vocab just moved to make CodeSystem, ValueSet and ConceptMap depend on MetadataResource
Grahame Grieve (Feb 18 2022 at 01:32):
so no, we better not make it non-normative. I think we could consider marking the element non-normative.
Lloyd McKenzie (Feb 18 2022 at 15:04):
Making it normative when the type has problems isn't a good plan. If we mark the element non-normative, then having the type non-normative isn't a problem.
Bryn Rhodes (Feb 18 2022 at 15:09):
We're planning to take Library normative, so that's UsageContext, RelatedArtifact, ParameterDefinition, and DataRequirement
Bryn Rhodes (Feb 18 2022 at 15:10):
Agree that RelatedArtifact needs work, but also happy to mark that element as TU in Library.
Lloyd McKenzie (Feb 18 2022 at 15:17):
UsageContext is currently TU. Do you feel ParameterDefinition and DataRequirement are ready to lock down?
Bryn Rhodes (Feb 18 2022 at 15:49):
ParameterDefinition is pretty simple and we haven't needed to expand it so far, so I feel good about that one. DataRequirement has had some changes in R5 but they are backwards compatible and introduce new capability. We could potentially keep those new things TU, but overall I'd say yes.
Last updated: Apr 12 2022 at 19:14 UTC
