State of the Duck
summary
Release v0.10.0 ships within weeks and the project plans a snow‑leopard style v1.0 focused on stability, backed by 5 million extension installs/month and about 2 TB/day of extension downloads; the site now gets ~500k unique visitors and ranks ~45 on DB‑Engines. v0.10 introduces a beta of a backwards-compatible file format (read-old/write-new partial guarantees), 2–3x faster CSV ingestion, a unified memory manager for out‑of‑core queries, ALP float compression, fixed‑length arrays, multi‑database attach (MySQL), secret manager, and vacuuming/defragmentation.
AI-written summary · about
chapters (12)
- 0:00DuckDB’s growth in 2023
- 3:53New website and milestones
- 7:02DuckDB 0.10 release and 1.0 roadmap
- 13:04Multi-database attachments and storage
- 14:28CSV reader performance gains
- 15:32Fixed-length arrays and functions
- 16:33Unified memory manager
- 17:38Secret manager for blobs
- 19:02Floating point compression
- 20:04Storage defragmentation and CLI
- 21:07Future plans for extensions
- 22:33Q&A
related talks
27mState of the Duck
28mState of the Duck
33mDuckDB – Overview and latest developments
27mDuckDB Extensions: The Past, the Present, and the Future
05mAnnouncing DuckDB Support for Delta Lake and a DuckDB Extension to Unity Catalog
40mDuckDB Meets Data Lakes
18mDuckDB Extensions
16mWhen the sea lion learns to quack: DuckDB as a MariaDB storage engine