Topics discussed:
- The work that started as "add tags to the test definitions" grew into a structure for the whole spec. Reserved tags now line up with profiles — shareable, tabular, experimental — so a tag can link to the profile it means. The FHIRPath subset moved out of the ViewDefinition page and into the shareable profile, and the test report groups results under those headings, with the boundary functions and
joinsitting under experimental: intended for the required subset eventually, not part of the normative release, still subject to change. - The rationale, when someone asked why the FHIRPath subset isn't in the ViewDefinition part of the spec: it follows how FHIR itself works. Base Observation and base Condition are deliberately loose, and in practice you constrain them with a profile to make them portable between systems — which keeps the base stable and reusable. Same here. You can put any FHIRPath in a ViewDefinition; the subset only starts to matter the moment you want that view to run on someone else's implementation, so it belongs in the profile that makes that claim.
- A pull request removing the metadata fields the profile duplicates from ancestor resources was closed rather than merged. The duplication is a side effect of ViewDefinition being a logical model rather than a real FHIR resource; if it is ever promoted to a resource, those fields should come through inheritance the way they do for other canonical resources. Closed for now, to be revisited then.
- The functional model page is orphaned — it exists, but nothing in the spec links to it, and nobody had a good answer for where the link should come from. It was noticed independently by more than one reviewer.
- Roughly ten ViewDefinition examples for common resources need translating into FHIR Shorthand before they can go in. That turned out to be less thankless than expected: the IG build validates against the ViewDefinition logical model, so it catches things like a string where a code should be, and does a better job than people assumed.