Stream: united states
Topic: Provenance - what is expectation for _summary=true?
Emma Jones (Aug 11 2021 at 14:12):
1.Is there an expectation for Provenance.agent to be returned when _summary=true?
- Generally, are mandatory elements required in summaries?
John Moehrke (Aug 11 2021 at 14:14):
the _summary for any resource is defined in FHIR core. This has no relationship to any IG profile.
John Moehrke (Aug 11 2021 at 14:15):
thus, yes the agent.who and agent.type would be expected in a summary
Emma Jones (Aug 11 2021 at 14:25):
correct. But they are sub-elements of an element that isn't marked to be included in _summary. In other resources the parent is marked if a child is marked. Is this an error in the spec?
John Moehrke (Aug 11 2021 at 14:27):
I don't know .... @Lloyd McKenzie should the backbone elements be marked with summary flag when any element within them are marked with summary? Should this not be something tooling automatically does?
John Moehrke (Aug 11 2021 at 14:30):
It is impossible to return sub-elements without returning the backbone.. so I never bothered flagging the backbone.
Lloyd McKenzie (Aug 11 2021 at 15:22):
Yes, if descendants are marked as being included in the summary, their ancestors should too.
John Moehrke (Aug 11 2021 at 15:44):
sure would be good if the tooling would just do that. So Provenance and AuditEvent need a CR ... @Emma Jones can you create a CR on this?
Emma Jones (Aug 12 2021 at 20:23):
@John Moehrke Please see FHIR-33170 sub elements are marked _summary but the parent element is not marked _summary
John Moehrke (Aug 12 2021 at 20:24):
thanks
Lee Surprenant (Aug 18 2021 at 14:01):
question: if the top-level element is marked as summary, won't that mean that all its children are also in the summary?
Lee Surprenant (Aug 18 2021 at 14:02):
the point of contention is on this phrase from http://www.hl7.org/fhir/elementdefinition-definitions.html#ElementDefinition.isSummary
Other than Attachment.data, all data type properties are included in the summary form.
Isn't BackboneElement still a datatype?
Lee Surprenant (Aug 18 2021 at 14:05):
actually, let me ask that on #implementers instead...
Ewout Kramer (Aug 18 2021 at 14:18):
That question was raised before, let me check.
Ewout Kramer (Aug 18 2021 at 14:20):
I recently was looking at this too, see https://jira.hl7.org/browse/FHIR-33148
John Moehrke (Aug 18 2021 at 18:23):
oh, I am so glad that this has continued to be discussed. I do NOT like marking a backbone with summary flag.
Lee Surprenant (Aug 18 2021 at 18:35):
sorry, i meant to post back here with the cross-link to the new topic on #implementers : https://chat.fhir.org/#narrow/stream/179166-implementers/topic/Clarification.20on.20Backbone.20summary.20elements
Last updated: Apr 12 2022 at 19:14 UTC
