Stream: implementers
Topic: Question about Conformance.rest.operation
Yunwei Wang (Apr 05 2017 at 13:18):
I am trying to write a CapabilityStatement to include the "new" CodeSystem/$validate-code. Then I realized it has the same name with ValueSet/$validate-code. So the operation section now has two operation with the same name "validate-code" but different definition reference endpoints.
Yunwei Wang (Apr 05 2017 at 13:18):
Looks like this:
Yunwei Wang (Apr 05 2017 at 13:19):
<operation>
<name value="validate-code"/>
<definition>
<reference value="http://hl7.org/fhir/OperationDefinition/ValueSet-validate-code"/>
</definition>
</operation>
<operation>
<name value="validate-code"/>
<definition>
<reference value="http://example.com/fhir/OperationDefinition/CodeSystem-validate-code"/>
</definition>
</operation>
Yunwei Wang (Apr 05 2017 at 13:20):
Is this valid?
Lloyd McKenzie (Apr 05 2017 at 13:46):
It's valid, but I agree it's confusing. You have to retrieve the definitions in order to verify it's not a collision. (You also have to retrieve the definition to verify where the operations are allowed to be invoked.) However, the alternative is that there's redundancy (and the potential for misalignment) with what's in the CapabilityStatement and what's in the OperationDefinition, so hard to know what's best to do here.
Yunwei Wang (Apr 05 2017 at 14:05):
Maybe we should consider a different operation name for Code System? This is from tracker 12700: http://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=12700&start=0
Lloyd McKenzie (Apr 05 2017 at 14:32):
The operation is essentially the same, so it would be odd for it to have a different name. When specified in context, the meaning is clear.
Elliot Silver (Apr 05 2017 at 16:21):
Is it cleaner to have one operation with one name and multiple definitions?
Lloyd McKenzie (Apr 05 2017 at 16:42):
Well, it would mean you'd need to create a derived operation if you wanted to support one but not the other. And in this case, the technical behavior is sufficiently different that having different operation definitions that each explained what was going on seems appropriate
Sunanda Veeraganti (Apr 05 2017 at 21:01):
What is the cardinality of both the operations? it will become difficult to check the presence of Valueset and Codesystem
Brian Postlethwaite (Apr 11 2017 at 01:17):
CodeSystem validate-code? http://hl7.org/fhir/codesystem-operations.html nothing on here.
(only on valueset http://hl7.org/fhir/valueset-operations.html)
Yunwei Wang (Apr 11 2017 at 17:25):
@Brian Postlethwaite That is new operation for STU4.
Last updated: Apr 12 2022 at 19:14 UTC
