#Engine
This section describes how the website works internally. Use these pages when you want implementation details about Angular orchestration, workers, WebGPU, WGSL, storage, exports, formats, limits, and performance behavior.
#Pages
- Technical architecture: high-level Angular, worker, WebGPU, OPFS, and export architecture.
- WebGPU webengine: worker runtime, rebuild lifecycle, message routing, and run loop.
- WGSL shaders: generated compute WGSL, render WGSL, brush WGSL, and metrics WGSL.
- Rule expressions: canonical selector, clause, outcome, tribe, and rule JSON reference.
- Rule cost model: rough per-cell cost formulas, shared count reuse, and worked rule examples.
- Rule engine internals: how rule clauses, selectors, outcomes, ordering, and muted rules become shader code.
- VRAM and packing: packed grid formats, frame byte formulas, WebGPU limits, and estimated VRAM budgeting.
- Recording and OPFS: recording buffers, chunks, staging, compression, backpressure, and browser storage.
- Downloads and exports: ZIP output,
.goltsaves, metrics, PNG frames, MP4, and chunk export. - Compressed chunk export: chunk ZIP layout, manifest fields, payload decoding, and Python reader example.
- Snapshot format:
.goltpreamble, JSON header, compressed payload, streaming path, and Python tools. - Metrics internals: live GPU metrics and offline export metrics.
- Limits and performance: hard-coded limits, practical constraints, and performance tradeoffs.