Storage service documentation - kind and id
Please check this page -
See how kind and id are defined. These need revision.
id: (optional) Unique identifier in the Data Ecosystem. When not provided, the service will create and assign an id to the record. Must follow the naming convention: {Data-Partition-Id}:{object-type}:{uuid}.
kind: (mandatory) Kind of data being ingested. Must follow the naming convention: {Data-Partition-Id}:{dataset-name}:{record-type}:{version}.
During a recent thread with @gehrmann and @doniger , here is some clarification from Thomas.
• id: At some point in time the term data-partition was considered alien and confusing – I started using namespace in Data Definitions. Over time the community adopted (only to some extent?) the term data-partition and the generic term namespace became alien and confusing.
• kind: The first part is schema-authority. The (old) Storage Schema endpoints require the schema-authority to be the data-partition-id. This service is deprecated. The Schema Service has taken over and the schema-authority is now independent from the data-partition.