Rill is powered by an incredible database that makes your analytics experience faster and more delightful — duckDB. In their most recent release, duckDB 0.7.0 has expanded their data type support to include JSON data ingestion. Though Rill Developer’s goal is to help analysts create One Big (tabular) Table, often data is collected in nested types such as structs, arrays and JSONs. This release introduces more features that unlock the ability to transform nested data into the tabular structure necessary to power your Rill Dashboard. Including stronger profiling for nested types has laid the foundation for a higher quality user experience when we upgrade to duckDB 0.7.0 in our next release!

• STRUCT, LIST, JSON, and UNION support – the future of data is nested; yet Rill works best with big rectangles of data. In this release, we’re finally adding profiling support for DuckDB’s many nested data types. In profiling, we place these nested types above the VARCHAR columns, but before the numeric columns. This is because we suspect that users will need to handle nested columns in a similar way. In a future release, we’ll add automatic profiling of nested values.

• navigation improvements – this release fixes a long-standing issue in navigation, where if you click, we run a bunch of calculations before showing any feedback. While it’s technically correct – seriously, the delay is not because of render jank! –it’s honestly a very grating experience to not see the interface physically respond to your clicks. This release, we’ve fixed this, so that the navigation item that’s clicked instantly highlights. Soon, we’ll also speed up the rendering of the workspace as well. But this already is a big improvement in the felt experience.