This task is linked to #153 and #127, as it describes the technical implementation.
It is, however, also related to #95 and #22.
To enable better curation and debugging, we need to ship metadata about the metadata we collate. What are sources of the data? Which one was selected (and why)?
The internal data model ADRs do not yet describe how we implement this. A new ADR is needed.
This task is linked to #153 and #127, as it describes the technical implementation. It is, however, also related to #95 and #22.
To enable better curation and debugging, we need to ship metadata about the metadata we collate. What are sources of the data? Which one was selected (and why)?
The internal data model ADRs do not yet describe how we implement this. A new ADR is needed.