Delta Lake 3.0.0 Announcement

[Scott Sandre]

Hello everyone,

We are happy to announce the final release of Delta Lake 3.0.0 on Apache Spark 3.5!

Some special highlights:

- Support for Apache Spark 3.5

- Delta Universal Format - Write as Delta, read as Iceberg! See the highlighted section above.

- Delta Kernel - The Delta Kernel project is a set of Java libraries (Rust will be coming soon!) for building Delta connectors that can read (and, soon, write to) Delta tables without the need to understand the Delta protocol details).

- Up to 10x performance improvement of UPDATE using Deletion Vectors

- More than 2x performance improvement of DELETE using Deletion Vectors

- Up to 2x faster MERGE operation

- Support specifying the columns for which Delta will collect file-skipping statistics via the table property `delta.dataSkippingStatsColumns`

And more! See the complete list of new features, improvements, and bug fixes here: https://github.com/delta-io/delta/releases/tag/v3.0.0

Cheers,

--

Scott Sandre

Software Engineer

Delta Ecosystem Team

scott....@databricks.com