Possible bug in time schema definition

[photonbit]

I am working on pygeometa output validation geopython/pygeometa#264 and I've found something that I am not sure if it a problem with my validation implementation, or it is a bug in the standard specification.

In https://schemas.opengis.net/ogcapi/records/part1/1.0/openapi/schemas/recordGeoJSON.yaml, time is defined as

  time:
    oneOf:
      - type: object
        nullable: true
      - $ref: 'time.yaml'

and time.yaml indicates:

---
nullable: true
type: object
properties:
# ...

That makes me see two potential problems:

• time.yaml is already defining a nullable object, so the case added in recordGeoJSON.yaml is duplicating this case.
• Any object, empty or not, will validate -type: object and time.yaml. If the duplicated case needs to exist in recordGeoJSON.yaml, then anyOf should be used instead of oneOf.

[pvretano]

18-AUG-2025: Assiging to @pvretano to review and comment or fix.