Stream: conformance
Topic: CapabilityStatement.fhirVersion
Steve Munini (Jan 14 2017 at 01:01):
This is probably a nit, but why is CapabilityStatement.fhirVersion of type id? CapabilityStatement.software.version is of type string. Seems like they should be of the same type. In our implementation, we are internally mapping the FHIR id type directly to UUIDs. I understand from the spec that ids can be any combination of characters with a length limit of 64 characters, but for those implementations who are mapping id to something other than a string internally (like some custom struct for example), fhirVersion will be a special case to code around if it is of type id. fhirVersion feels more like a string than an id to me.
Grahame Grieve (Jan 14 2017 at 02:33):
I don't really know. it's always going to be x.y.z - nothing else.
Grahame Grieve (Jan 14 2017 at 02:34):
where as the software version - that's other people's versions- that could be anything.
Brian Postlethwaite (Jan 14 2017 at 02:34):
DataType ID is not just UUIDs, that might get you in trouble elsewhere with that mapping.
Ids have a validation regex of [A-Za-z0-9\-\.]{1,64}. So a Z in your Id which is valid, is not valid for a UUID.
Last updated: Apr 12 2022 at 19:14 UTC
