Skip to content

Roadmap

Manuel Luján edited this page Nov 6, 2026 · 2 revisions

Roadmap

LineageIQ is at version 0.2.0. This page lists what may come next and what is deliberately out of scope. It has no dates: the project has a single maintainer, and the order depends on feedback.

Candidates for 0.3.0

The next release, 0.3.0, will ship one or two of these, chosen from user feedback. Both are about interoperability: getting LineageIQ's object-level lineage into the tools people already use.

  • Export lineage to Microsoft Purview through its lineage REST API, so the object-level lineage that Purview does not scan for Fabric items appears in Purview itself. This is the "complement Purview" direction.
  • Export lineage in the OpenLineage format, so other catalogues and lineage tools can read LineageIQ's graph.

Both are tracked in the v0.3.0 milestone.

Also open

These come from the known limitations and the maintenance notes in the code. They are not tied to a release, and contributions are welcome. See the open issues.

  • Stable identity by Fabric item GUID, so a renamed item is tracked as a rename instead of a removal plus an addition.
  • Automate the Fabric-native discovery pipeline setup. Today the notebooks are imported and the Data Pipeline is wired by hand.
  • Keep the route registry in sync with handler.py automatically, starting with a test that fails when they drift. Marked as a good first issue.
  • A parity test for the PowerShell and Python build pairs (11, 17 and 21), which today must be kept identical by hand.

How to influence the choice

  • Add a thumbs-up reaction to the issues you care about, and comment with your use case: what you would do with it and what you use today. Reactions and comments are the main input for choosing what ships next.
  • Suggest something new in Discussions, Ideas or with a feature request.

Not planned

These are deliberate choices, not oversights. They can be revisited if there is a strong case, but nobody is working on them.

  • Column-to-column lineage. Doing it reliably needs deep analysis of SQL and PySpark, and the project prefers not to promise results it cannot back up. Lineage stays at object level; columns stay catalogued and profiled.
  • Writing to the scanned tenant. Discovery stays read-only. Any operational change is proposed as text for a person to review and apply.
  • Replacing Microsoft Purview. LineageIQ complements Purview, and interoperability is the direction.
  • Item-level incremental sync. Microsoft's APIs offer only workspace-level change polling, so there is nothing to build true item-level sync on today.

Roadmap board

The open work is tracked as issues on the public LineageIQ roadmap board, grouped by status: Backlog, Planned, In progress and Done. No dates are promised.

Clone this wiki locally