Feature Request / Improvement
This sub-issue tracks the remaining row-lineage work needed to move v3 support forward under #589.
A lot of v3 foundation is already in place, but one important commit-path piece is still missing:
we need end-to-end wiring so commit output consistently reflects row lineage state.
What’s missing (high level)
- v3 commit flow should use table next-row-id as snapshot first-row-id
- snapshot added-rows should be set from actual committed data-manifest row ranges
- manifest first_row_id values should be consistent with snapshot/table lineage progression
- table next-row-id should advance correctly after commit
Without this, row-lineage support is only partially complete in practice: metadata types/validation exist, but commit-time lineage assignment is not fully closed end-to-end.
Feature Request / Improvement
This sub-issue tracks the remaining row-lineage work needed to move v3 support forward under #589.
A lot of v3 foundation is already in place, but one important commit-path piece is still missing:
we need end-to-end wiring so commit output consistently reflects row lineage state.
What’s missing (high level)
Without this, row-lineage support is only partially complete in practice: metadata types/validation exist, but commit-time lineage assignment is not fully closed end-to-end.