ducklake / duckdb / lakehouse

A Hacker News thread highlights a project to create a distributed DuckDB instance that mimics MotherDuck’s differential storage and hybrid query execution on top of DuckDB. The original poster (citguru) says the goal is replicating MotherDuck-style differential storage and adding hybrid execution; commenters debate DuckDB’s expanding ecosystem (DuckLake, MotherDuck) versus its original simple, single-file ethos like SQLite. Other users mention contributing to DuckDB (adding Iceberg view support) and ask for plain explanations of terms like differential storage and hybrid execution. The discussion matters because it touches on trade-offs between simplicity and feature-rich distributed analytics, and signals community interest in scaling DuckDB for distributed/warehouse use.

DuckDB announces DuckLake v1.0, a production-ready lakehouse format specification and reference implementation available as the ducklake extension in DuckDB v1.5.2. DuckLake centralizes lakehouse metadata in a SQL database catalog (SQLite, PostgreSQL, or DuckDB), enabling schema/version control, metadata-driven operations, and features like inlining for streaming small updates, Iceberg compatibility, geometry and variant types, and adding existing Parquet files without copying. The extension has seen rapid adoption—top-10 DuckDB extension downloads—with integrations for DataFusion, Spark, Trino, and Pandas, hosted support from MotherDuck, and production use at dozens of companies. v1.0 focuses on stability, backward compatibility, bug fixes, and a roadmap for future development.

DuckLake v1.0 – The Lightweight Lakehouse Format Reaches Production-Readiness

DuckDB team released DuckLake v1.0, a production-ready lakehouse format specification and reference implementation as the ducklake DuckDB extension included in DuckDB v1.5.2. DuckLake centralizes lakehouse metadata in a SQL catalog (supports SQLite, PostgreSQL, DuckDB) instead of scattered object-storage files, enabling features like inlining small updates, streaming into lakes, Iceberg compatibility, geometry and variant types, and multiplayer DuckDB access via a shared catalog. The project added tools to ingest existing Parquet files, migration guides, and multiple client integrations (DataFusion, Spark, Trino, Pandas). Adoption is growing: top-10 DuckDB extension downloads, hosted offerings (MotherDuck), production use at dozens of companies, community contributions, and an O’Reilly book underway. This release focuses on stability, backward-compatibility, and production robustness.