FHIR Chat · Can / should an Extension.url contain an optional version? · implementers

Stream: implementers

Topic: Can / should an Extension.url contain an optional version?


view this post on Zulip John Timm (Jul 21 2020 at 15:42):

The spec mentions resolving the StructureDefinition for an extension using a "canonical" URL though Extension.url is typed to uri. Can / should Extension.url be used in a similar fashion to Resource.meta.profile where an optional version can be appended?

"The url SHALL be a URL, not a URN (e.g. not an OID or a UUID), and it SHALL be the canonical URL of a StructureDefinition that defines the extension. Except for child extensions defined within complex extensions, the URL SHALL be an absolute URL."

When I think of a canonical URL, I think about the potential for an optional version (especially if we are looking up a StructureDefinition from a registry).

view this post on Zulip Lloyd McKenzie (Jul 21 2020 at 16:00):

Extension.url is not allowed to contain "|version" because it does not have a type of canonical. This is intentional.

view this post on Zulip John Timm (Jul 21 2020 at 16:02):

So perhaps the wording should be changed to not include "canonical URL"?

view this post on Zulip Lloyd McKenzie (Jul 21 2020 at 16:04):

It's a reference to the canonical URL, but it's not of type 'canonical'.

view this post on Zulip John Timm (Jul 21 2020 at 16:13):

Perhaps the language should then be changed from SHALL be the canonical URL of a StructureDefinition that defines the extension to SHALL be a reference to the canonical URL of a StructureDefinition that defines the extension though the nuance between an element of type canonical and a reference to a canonical URL still seems a bit ambiguous to me.

view this post on Zulip Lee Surprenant (Jul 21 2020 at 16:18):

Extension.url is not allowed to contain "|version" because it does not have a type of canonical. This is intentional.

Mind expanding on the intention behind this? Elsewhere, when we have a reference to a definitional knowledge artifact that needs to get looked up, it tends to be a canonical (so that it can include the version), right? Why not here?

view this post on Zulip Lloyd McKenzie (Jul 21 2020 at 17:13):

@Grahame Grieve

view this post on Zulip Grahame Grieve (Jul 21 2020 at 21:29):

it's deliberate because you should not change the definition of an extension so much that the version becomes relevant to the interpretation

view this post on Zulip Grahame Grieve (Jul 21 2020 at 21:29):

but I do think that it's reasonable to clarify the language

view this post on Zulip Lloyd McKenzie (Jul 22 2020 at 05:04):

@John Timm Can you submit a change request asking for the clarification?

view this post on Zulip John Timm (Jul 24 2020 at 20:34):

Created: https://jira.hl7.org/browse/FHIR-28143


Last updated: Apr 12 2022 at 19:14 UTC