Abstract:

This video provides a comprehensive tutorial on Rerun Blueprints, a feature for programmatically and manually controlling the layout and state of the Rerun data viewer. The core concept presented is that "data + a blueprint equals the viewer," allowing users to create customized and persistent visualization templates. The tutorial explains that blueprints control panel visibility, layout structure, and view-specific properties like background color and zoom levels. A key mechanism highlighted is the application_id, which links a specific blueprint to a data logging session, ensuring consistent layouts across runs. The video demonstrates how to define and send blueprints from Python code, build complex multi-panel layouts using containers (Grid, Horizontal, Vertical, Tab), and manually rearrange, save, and import blueprint files directly within the viewer UI.

Mastering Rerun Blueprints: A Guide to Customizing Your Viewer

• 0:00:08 What are Blueprints? Blueprints are used to control the layout of the Rerun viewer. They define which panels are visible, how viewers (e.g., charts, 3D spaces) are arranged, and the state of those viewers. The fundamental concept is data + blueprint = viewer.
• 0:02:00 The Importance of application_id: The application_id is a crucial identifier that links a data session to a specific blueprint. All applications that share an application_id will use the same blueprint, providing a consistent layout for recurring tasks or datasets.
• 0:03:27 Programmatic Control: You can define and send blueprints directly from your code. This allows you to create default or dynamic layouts tailored to your application's state. Examples include creating a simple TextLogView or a Spatial2DView with a specified background color.
• 0:09:00 Building Complex Layouts: Blueprints can combine multiple different views, such as a 2D scene, a map view, and a text log, into a single, cohesive interface. You can also programmatically control which side panels (like the timeline or selection view) are expanded or collapsed by default.
• 0:11:23 Manual Customization: The viewer allows for direct, manual manipulation of the layout. Users can drag and drop panels, split views horizontally or vertically, and rearrange content to suit their immediate needs.
• 0:11:51 Layout Containers: Rerun uses different container types to structure the viewer layout. These include Grid, Horizontal, Vertical, and Tab containers, which can be nested to create highly customized and information-dense dashboards.
• 0:12:52 Saving and Reusing Blueprints: Any manual changes to the layout are automatically saved and associated with the application_id. You can also explicitly save a desired layout to a blueprint file (.rbl) for sharing or reuse with other data.
• 0:13:43 Importing Blueprints: A saved blueprint file can be imported directly into the viewer. This allows you to apply a pre-configured layout to any compatible dataset, regardless of the application_id.