Stream: methodology
Topic: stream events
Notification Bot (Oct 20 2019 at 17:02):
Stream created by Lloyd McKenzie.
Grahame Grieve (Oct 21 2019 at 15:55):
ok. open subjects for me:
- how should we represent type choice
- what changes should we make to our methodology for resources that represent a new green fields space (where we can't fall back on 'look at what systems actually do to handle variance/disagreement')
Grahame Grieve (Oct 21 2019 at 15:56):
- getting some consistency around the subject/focus split from Observation into other resources (using a terminology for that)
- deciding what to do about an object representation of the participations
- formally defining the following types: BaseObject, Type, MetadataResource
Grahame Grieve (Oct 21 2019 at 15:57):
- figuring out what to do about the .identifier elements across all resources
Lloyd McKenzie (Oct 21 2019 at 16:01):
Feel free to start threads for each. I don't understand the first.
Grahame Grieve (Oct 21 2019 at 18:32):
also
- the problem with instantiatesUrl/instantiatesCanonical. We have to do better than that
- The ConceptReference debate (things that can either be a CodeableConcept or a Reference)
Grahame Grieve (Oct 21 2019 at 18:35):
- can we resolve the issue with not defining repeating elements that have type choice
Lloyd McKenzie (Oct 21 2019 at 19:23):
The last seems like something that's rather late to address, given that XML and JSON are now locked down..
Grahame Grieve (Oct 21 2019 at 19:29):
right. they are locked down. but I'm not entirely convinced that there's nothing we can do.
For instance: we could say that instantiates[x] 0..* is ok as long as there is no intention around order, but it manifests as instantiatesUri:[] and instantiatesCanonical : [] in JSON. Since there's no current elements like that, it's not a breaking change
Lloyd McKenzie (Oct 21 2019 at 20:00):
It's a breaking change for the reference implementations and equivalent code at a minimum
Grahame Grieve (Oct 21 2019 at 20:06):
I don't know on what basis that constitutes a procedural issue. it might be input into whether we like the change or not
Lloyd McKenzie (Oct 21 2019 at 20:09):
I don't think it's covered in the set of changes we explicitly said we'd allow. I also don't think it's listed as something that's prohibited. As such, it's at best "suspect". I agree we can pursue the question though.
Grahame Grieve (Oct 24 2019 at 19:19):
additional open subject: ConceptReference
Jose Costa Teixeira (Oct 24 2019 at 20:48):
Some places where this can be used:
MedicationRequest.medication
MedicationDispense.medication
MedicationAdministration.medication
MedicationUsage.medication
Medication.ingredient.item
DeviceDefinition.manufacturer
Careplan.activity.detail.product
ActivityDefinition.subject
ActivityDefinition.product
Jose Costa Teixeira (Oct 24 2019 at 20:51):
since this is a methodology stream - is this something that will be enforced for all cases where it applies, or a volunteer adoption?
Grahame Grieve (Oct 24 2019 at 21:14):
could be either, but probably some smei-coerced adoption
Jose Costa Teixeira (Oct 24 2019 at 21:33):
so we could treat entry[x] with codeableconcept + reference as an anti-pattern.
Last updated: Apr 12 2022 at 19:14 UTC
