Open jkva opened 1 year ago
@sinabahram you suggested that we use canonical_id
as the identifier for a resource, yet this is not a required column.
We could change the data model to either generate a UUID or have the user define their own unique canonical ID, and include a DB migration that would fill in canonical IDs to ensure this holds for existing records.
Apologies, I meant Coyote ID
Via the format
/org/[org-id]/resource/[canonical-id]/simple
, render a view of the image with the expected highest ranked representation, including all of the descriptions that are approved.