Concepts¶

The core mental model of Visualli. The Specification page covers the exact JSONL format; this page explains what the pieces are and why they exist.

The Infinite Canvas¶

Visualli treats information as living on an infinite 2D canvas. Content is not forced into a rigid tree — positions are free-form, but hierarchical relationships between layers and nodes are always preserved.

Layers and Levels¶

Layers¶

The fundamental unit of organization is the Layer. A layer is a self-contained collection of nodes, connections, and containers:

• Self-contained. Everything a layer needs to render its content lives inside it.
• Independent. Layers can be loaded, shown, or hidden individually — enabling "detail on demand": show the summary first, fetch deeper layers only as the user zooms or expands.
• Hierarchical. A layer can be a child of another layer, attached to a specific node in the parent layer.

Levels¶

Multiple layers can sit at the same Level. Think of a level as a "floor" in a building:

• Depth. Level 0 is the root summary, Level 1 is the next drill-down, and so on.
• Grouping. Several independent layers can share one level; the user switches between them (like apartments on one floor), but only one is visible at a time.
• Navigation. Users progress from lower levels (overview) to higher levels (detail).

How Layers and Levels relate¶

graph TD
    %% LEVEL 0: THE FOUNDATION
    subgraph L0 [**LEVEL 0**]
        direction LR
        subgraph LA [Layer A]
            direction LR
            A1((N)) 
            A2((N))
        end
    end

    %% LEVEL 1: LAYER B ON LEFT, LAYER C ON RIGHT
    subgraph L1 [**LEVEL 1**]
        direction LR
        subgraph LB [Layer B]
            direction LR
            B1((N)) --> B2((N)) --> B3((N))
        end
        subgraph LC [Layer C]
            direction LR
            C1((N)) --> C2((N))
        end
    end

    %% LEVEL 2: LAYER D -> LAYER E -> LAYER F
    subgraph L2 [**LEVEL 2**]
        direction LR
        subgraph LD [Layer D]
            direction LR
            D1((N)) --> D2((N)) --> D3((N)) --> D4((N))
        end
        subgraph LE [Layer E]
            direction LR
            E1((N))
        end
        subgraph LF [Layer F]
            direction LR
            F1((N)) --> F2((N))
        end
    end

    %% CONNECTIONS
    A1 -.-> LB
    A2 -.-> LC

    B1 -.-> LD
    B2 -.-> LE
    B3 -.-> LF

    %% STYLING
    style L0 stroke-dasharray: 5 5
    style L1 stroke-dasharray: 5 5
    style L2 stroke-dasharray: 5 5

Entities (mental model)¶

These are the building blocks. See the Specification page for exact JSONL fields and validation rules.

Extension example¶

Extensions are identified by ID, and their payload shape is up to the extension's contract. For instance, a semantic-anchors extension links terms to descriptions:

{
  "type": "extension",
  "id": "semantic-anchors",
  "data": [
    {
      "word": "atmosphere",
      "description": "The envelope of gases surrounding the earth or another planet.",
      "knowMoreUrl": null
    },
    {
      "word": "water circulation",
      "description": "The continuous movement of water on, above and below the surface of the Earth.",
      "knowMoreUrl": null
    }
  ]
}