Skip to main content

Rill 0.19 – Embed data sources in models

· 2 min read
note

⚡ Rill Developer is a tool that makes it effortless to transform your datasets with SQL and create powerful, opinionated dashboards.

To try out Rill Developer, check out these instructions and let us know over on Discord if you encounter any problems or have ideas about how to improve Rill Developer!

This release of Rill Developer introduces a new way to start your analysis by directly embedding sources into model FROM statements. This means you can jump directly into your data project without fiddling with the inputs and tabs needed to manually create a source reference through our modal. With embedded sources, we check for valid URIs to ingest and cache the source in Rill Developer to keep things incredibly fast as you elaborate upon your query. The result is a magical, fast, frictionless querying experience from the moment you identify an interesting data set.

  • Embed sources directly into models — Fast and simple are core value here at Rill. To this end, we are introducing an experimental feature where you can define your source directly in the model. The ability to embed sources in models eliminates the navigation complexity of sources and makes it easy to jumpstart modeling with a simple FROM statement and a uri:
     FROM "s3://bucket/partition.parquet"
    — check it out and let us know what you think on Discord! Note that embedded sources do not have total feature parity with parameters explicitly defined sources in the modal.
  • Create dashboards without timeseries data — Data comes in all shapes and sizes, however historically our dashboard had hard requirements around the types needed to create an interactive dashboard - primarily a timeseries column to generate temporal line charts. Though the dashboard was designed with time series line charts in mind, digging into catalogs and lookup table metrics can be incredibly useful for for understanding what is available and how to refine dimension value labels for consumption by people. To support this workflow, we have relaxed timeseries constraints and it is now possible to present leaderboards and big-numbers without charts.