FHIR Chat · FHIR-32689 TestScript.scope.conformance codes · TestScript Resource

Stream: TestScript Resource

Topic: FHIR-32689 TestScript.scope.conformance codes

view this post on Zulip Richard Ettema (Sep 14 2021 at 20:16):

During the May 2021 Connectathon 28 event, a new element was approved and implemented for the r5 TestScript resource - the TestScript.scope element. An HL7 FHIR JIRA ticket FHIR-32689 was introduced since then to expand the set of codes used by the scope.conformance element. This original expansion added the 'recommended' code as follows:

Code Display Definition
required Required All tests are expected to pass. This is the default value.
optional Optional All tests are expected to pass but non-pass statuses may be allowed.
recommended Recommended All tests are expected to pass but warning statuses may be allowed.

During a recent FHIR-I WG Monday call, this JIRA was reviewed and the proposed change was updated to the following:

Code Display Definition
required Required All tests are expected to pass. Warning statuses are permitted. This is the default value.
optional Optional All tests are expected to pass but non-pass statuses may be allowed.
strict Strict All tests are expected to pass. Warnings are treated as a failure.

FHIR-I has requested feedback from the FHIR community on this proposed updated change before approving this JIRA issue. Please let us know if these updated codes are well defined or if they may require additional refinement.

view this post on Zulip Jose Costa Teixeira (Sep 14 2021 at 20:45):

I was somewhat uncomfortable with the notion of "required" and "really really required" which i how I read required vs strict.

view this post on Zulip Jose Costa Teixeira (Sep 14 2021 at 20:46):

I was also struggling with the absence of recommended, but perhaps this is because "recommended" may have different valid flavours.

view this post on Zulip John Moehrke (Sep 14 2021 at 21:14):

you need to support the normative word as part of IETF RFC 6919 

   The key words "MUST (BUT WE KNOW YOU WON'T)", "SHOULD CONSIDER",
   "REALLY SHOULD NOT", "OUGHT TO", "WOULD PROBABLY", "MAY WISH TO",
   "COULD", "POSSIBLE", and "MIGHT"

view this post on Zulip John Moehrke (Sep 14 2021 at 21:15):

:-) ---- some will know that this is an April 1st publication -- April Fools

view this post on Zulip John Moehrke (Sep 14 2021 at 21:16):

but... they are useful --- especially when we consider some of the IGs definitions of "Must Support"

view this post on Zulip Richard Ettema (Sep 20 2021 at 14:23):

While considering "required" vs "strict" as Jose suggests is "required" vs. "really, really required", I believe what we are trying to convey is more like the following:

Code Display Definition
required Required All tests are expected to pass. Warnings are treated as a failure. This is the default value.
supported Supported All tests are expected to pass. Warning statuses are permitted.
optional Optional All tests are expected to pass but non-pass statuses may be allowed.

Thoughts?

view this post on Zulip Ron Shapiro (Sep 22 2021 at 16:33):

How about something like the following?

Code Display Definition
ignore-warnings Warnings Allowed All tests are expected to pass. Warning statuses are permitted. This is the default value.
fail-with-warnings Warnings Not Allowed All tests are expected to pass. Warnings are treated as a failure.
ignore-errors Errors Allowed All tests are expected to pass but non-pass statuses are allowed.

view this post on Zulip Richard Ettema (Sep 22 2021 at 16:55):

Ron, thank you. I like the new code and display values. I've updated the definitions to (hopefully) better represent their meanings:

Code Display Definition
ignore-warnings Warnings Allowed All test outcomes will not have errors. Only Pass and Warning statuses are permitted. This is the default value.
fail-with-warnings Warnings Not Allowed All test outcomes will not have errors or warnings. Only Pass statuses are permitted. Warnings are treated as a failure.
ignore-errors Errors Allowed All test outcomes are allowed. Pass and non-pass statuses are permitted.

view this post on Zulip Jose Costa Teixeira (Sep 23 2021 at 15:30):

fail-with-warnings or fail-on-warnings ?

view this post on Zulip Richard Ettema (Sep 23 2021 at 20:49):

I would say "fail-on-warnings" is more appropriate.

Last updated: Apr 12 2022 at 19:14 UTC


An HL7 Website -