Tattvas

A tattva is any object that can be placed, animated, and rendered in a scene. The word comes from Sanskrit and means "element" or "essence".

Every tattva is added to the scene with add_tattva(shape, position) which returns a TattvaId for later reference:

let id = scene.add_tattva(Circle::new(1.0, 64, Vec4::new(0.2, 0.6, 1.0, 1.0)), Vec3::ZERO);

This is the preferred authoring API. The lower-level scene.add(...) path is intended for internal or advanced cases where you already have a fully constructed tattva object.

Before diving into a specific family, see Common Tattva Properties for the shared scene-level behavior that most tattvas support.

Tattvas are organized into categories based on their purpose:

• Primitives — basic shapes such as circle, square, rectangle, line, polygon, arrow, path, and cube
• Text — label, latex, typst, and code block
• Tables — structured grids with labels and titles
• Composite — axes, number plane, and 3D axes
• Graphs — function graphs, scatter plots, parametric curves, surfaces, vector fields, and stream lines
• Math — equation layout, matrices, and vector formulas
• Layout — stacking and relative arrangement helpers
• Storytelling — stepwise reveal for guided explanations
• AI Diagrams — attention matrices, neural diagrams, signal flows, and flow-chart style components
• Utility — traced path, screenshot marker, and other scene helpers

What should be covered

As a rule, the docs should cover all shipped tattva families, even when some pages begin as lightweight indexes.

A good content split is:

• family page for orientation
• concrete tattva entries for constructors and notable fields
• one shared page for common tattva properties

That keeps the docs discoverable without repeating the same transform and visibility guidance on every page.