Rio Webgpu Tiles

Rio WebGPU Tiles

GPU-accelerated tile rendering in Rio Terminal using wgpu and the sugarloaf brush architecture.

Architecture Overview

OSC 1337 Sequence → rio-backend (parse) → RioEvent::InsertTile → rioterm (frontend) → sugarloaf TileBrush → wgpu render

Core Files

Protocol: OSC 1337 Tile Extension

# Format: ESC ] 1337 ; Tile = key:value,key:value,... BEL
printf '\033]1337;Tile=shader:plasma,x:50,y:50,w:200,h:150\007'

Parameters

Tile Lifecycle

Persistent Tiles

Remain until explicitly removed by ID:

let id = sugarloaf.create_persistent_tile(scene);
// ... later
sugarloaf.remove_persistent_tile(id);

Transient Tiles

Cleared at begin_frame(), must be re-pushed each frame:

sugarloaf.push_transient_tile(scene);

TileWorldState

Manages CPU state with high-precision time (f64), converted to modular f32 for GPU:

pub struct TileWorldState {
    persistent: HashMap<TileId, TileScene>,
    transient: Vec<TileScene>,
    world_time_seconds: f64,
    next_id: TileId,
}

impl TileWorldState {
    pub fn begin_frame(&mut self, dt_seconds: f64) {
        self.world_time_seconds += dt_seconds;
        self.transient.clear();
    }
}

Writing WGSL Shaders

Uniform structure available in shaders:

struct Uniforms {
    position: vec2<f32>,  // NDC position
    size: vec2<f32>,      // NDC size
    time: f32,            // Animation time (modular)
    custom: vec4<f32>,    // r,g,b,a from protocol
}
@group(0) @binding(0) var<uniform> u: Uniforms;

Vertex Shader Pattern

@vertex
fn vs_main(@location(0) pos: vec2<f32>, @location(1) uv: vec2<f32>) -> VertexOutput {
    var out: VertexOutput;
    let screen_pos = u.position + pos * u.size;
    out.position = vec4<f32>(screen_pos, 0.0, 1.0);
    out.uv = uv;
    return out;
}

Fragment Shader Pattern

@fragment
fn fs_main(in: VertexOutput) -> @location(0) vec4<f32> {
    let t = u.time;
    // Your shader logic using in.uv, t, u.custom
    return vec4<f32>(color, 1.0);
}

Demo Scripts

# Plasma effect
printf '\033]1337;Tile=shader:plasma,x:50,y:50,w:200,h:150,r:0.5,g:0.2,b:0.8\007'

# Clock display
printf '\033]1337;Tile=shader:clock,x:300,y:50,w:250,h:80\007'

# Remove by ID
printf '\033]1337;Tile=remove:42\007'

Integration Checklist

1. Add TileBrush to sugarloaf - Register in brush list, call in render loop
2. Parse OSC 1337 in rio-backend - Extend performer handler
3. Create RioEvent variant - InsertTile(TileSpec)
4. Handle in frontend - Convert TileSpec → TileScene, call sugarloaf API
5. Manage state - Call begin_frame(dt) before each render

Shader Examples

See reference/shaders.md for complete plasma, clock, and noise shader implementations.

Scientific Skill Interleaving

This skill connects to the K-Dense-AI/claude-scientific-skills ecosystem:

Graph Theory

• networkx [○] via bicomodule
  • Universal graph hub

Bibliography References

• general: 734 citations in bib.duckdb

SDF Interleaving

This skill connects to Software Design for Flexibility (Hanson & Sussman, 2021):

Primary Chapter: 9. Generic Procedures

Concepts: dispatch, multimethod, predicate dispatch, generic

GF(3) Balanced Triad

rio-webgpu-tiles (+) + SDF.Ch9 (○) + [balancer] (−) = 0

Skill Trit: 1 (PLUS - generation)

Secondary Chapters

• Ch4: Pattern Matching
• Ch10: Adventure Game Example

Connection Pattern

Generic procedures dispatch on predicates. This skill selects implementations dynamically.

Cat# Integration

This skill maps to Cat# = Comod(P) as a bicomodule in the equipment structure:

Trit: 0 (ERGODIC)
Home: Prof
Poly Op: ⊗
Kan Role: Adj
Color: #26D826

GF(3) Naturality

The skill participates in triads satisfying:

(-1) + (0) + (+1) ≡ 0 (mod 3)

This ensures compositional coherence in the Cat# equipment structure.