FHIR Chat · identifiers and UUID's

Stream: implementers

Topic: identifiers and UUID's

Jens Villadsen (Feb 26 2020 at 09:22):

Reading through https://www.hl7.org/fhir/datatypes.html#identifier I get the impression that if the value of the identifier is a UUID, the system SHALL be the urn:ietf:rfc:3986.

Naturally globally unique identifiers are those for which no system has been assigned and where the value of the identifier is reasonably expected to not be re-used.

But I easily imagine that some systems would like to 'for (the sake of 'simplicity/unknown/scoping/collision') reasons' to specify a particular system from where the UUID sort of originated. Would it be considered non-conformant to have an identifier as follows?:

{
      "system": "urn:oid:1.2.3.4.5",
      "value": "urn:uuid:0789cfc2-41a9-4999-bbaa-d2ecdf4fceda"
}

Grahame Grieve (Feb 26 2020 at 11:29):

that is not non-conformant but in generally, if you assign a system, I would expect to see:

Grahame Grieve (Feb 26 2020 at 11:30):

{
      "system": "urn:oid:1.2.3.4.5",
      "value": "0789cfc2-41a9-4999-bbaa-d2ecdf4fceda"
}

Jens Villadsen (Feb 26 2020 at 12:18):

right, ok - should the following sentences then be rephrased or refined:

If the identifier value itself is naturally a globally unique URI (e.g. an OID, a UUID, or a URI with no trailing local part), then the system SHALL be "urn:ietf:rfc:3986", and the URI is in the value (OIDs and UUIDs using urn:oid: and urn:uuid: - see note on the V3 mapping and the examples). Naturally globally unique identifiers are those for which no system has been assigned and where the value of the identifier is reasonably expected to not be re-used. Typically, these are absolute URIs of some kind.

UUID is (v4 at least) pretty random and is expected to not be re-used. If UUID's are within the definition of (and I believe they are)'naturally globally unique identifiers' isn't there then some sort of contradiction in the sentence: