Questing · 2026-06-11 · Simulation · Zero dependencies
MYCEL
Type any word. Each letter places a food source on a dark canvas. 1,500 particle agents — inspired by Physarum polycephalum, the slime mould that rediscovered Tokyo’s rail network — grow glowing mycelium networks between them. Same word, same network, always.
Open Mycel →What is Mycel?
Mycel is a browser-based Physarum polycephalum-inspired particle simulation: type any word and 1,500 agents grow a glowing mycelium network between the letters within seconds, entirely in the browser at 60 fps. Each letter of the word places an amber food source on the canvas; the agents use local chemical sensing to form near-minimal spanning paths between them — the same optimisation strategy used by the real slime mould to design transportation networks.
The word determines where food sources appear: each letter’s position in the 26-letter alphabet modulates its radius from centre (±30% variation), so every distinct word produces a unique source constellation. The network that grows to connect them is fully deterministic — clear the canvas, type the same word, and the identical topology re-emerges.
The Biology Behind It
In a landmark 2010 study published in Science (Tero et al., vol. 327, pp. 439–442), researchers fed oat flakes to Physarum polycephalum at positions matching Tokyo’s 36 major satellite cities. Within 26 hours, the mould had grown a network whose structure closely mirrored the actual Tokaidō Shinkansen rail system — achieving similar efficiency, fault-tolerance, and construction cost to the human-engineered railway. The organism reached this solution with no brain and no global coordination: only local chemical signalling between adjacent cells.
Mycel distils that behaviour into a browser toy. The colour palette shifts from deep forest-green at low trail density to bright lime-amber at peak density — reflecting the qualitative glow transition seen in bioluminescent fungi under ultraviolet light.
How to Use It
- Open slayerblade.site/q/mycel/ — a demo word loads automatically.
- Type any word of up to 10 letters and press Enter (or Return). The canvas clears and the network begins growing.
- Watch the amber food-source markers pulse at each letter’s position. Within a few seconds, green trails connect them.
- Type a different word to grow a new network. The topology is unique to each word.
- The URL fragment updates to your word — copy and share the link to show anyone the same network.
Technical Notes
Mycel runs as a single 10 KB static HTML file with no external JavaScript dependencies. The simulation targets 60 fps on any modern browser (Chrome, Firefox, Safari). Agent state is stored in three parallel Float32Array buffers (x, y, angle) for cache-friendly iteration; 1,500 agents complete a full sensing + move + deposit cycle in under 2 ms per frame on a mid-range device.
| Parameter | Value | Effect |
|---|---|---|
Agents | 1,500 | Particle count; higher = denser networks but slower |
Sensor angle | ±42.9° | Left/right sensing arc; wider = more meandering paths |
Sensor distance | 13 px | How far ahead each agent looks for trail |
Trail decay | 0.962×/frame | Trails fade in ≈27 frames; keeps paths clean |
Trail resolution | 36% of canvas | Downsampled for performance; upscaled via bilinear |
Steps/frame | 4 (early) → 2 | Faster early growth, stabilises to 2× real-time |
The trail map runs at 36% of canvas resolution (≈692×388 on a 1920×1080 screen) and is rendered via an offscreen canvas scaled up with browser bilinear interpolation, giving smooth glows at a fraction of the per-pixel cost of full-resolution rendering. The URL fragment encodes the word — share the link to share the exact network.