Repository navigation
Home
HYWE is an open-source computational spatial design environment and open research project where structured programmatic intent resolves into spatial configurations through deterministic design computation.
Rather than beginning with fixed geometry or stochastic image generation, HYWE represents architectural intent as hierarchical and relational information, resolves that information onto a discrete integer lattice, generates comparable deterministic configurations, and provides analytical tools to evaluate spatial properties.
-
HYWE Web App (
hywe.in) — Interactive spatial design environment. -
HYWE Engine (
repo.hywe.in) — Computational engine and open-source implementation. -
HYWE Wiki (
wiki.hywe.in) — Technical documentation and research reference. -
HYWE Spatial Dataset (
data.hywe.in) — Structured spatial configuration dataset. -
Software Archive & DOI (
10.5281/zenodo.22967097) — Permanent Zenodo software archive and citation record. - AEC Open Source Ecosystem — Listed in the AEC Open Source Directory for architecture, engineering, and construction tools.
- F# Community Highlights — Featured as a highlighted project in F# Weekly #18 (2026) and F# Weekly #35 (2026).
Foundational philosophy, conceptual positioning, ontology, and system guarantees.
- Computational-Spatial-Design — The theoretical foundation: relationships preceding geometry.
- Design-Principles — Core principles guiding discrete spatial synthesis and UI design.
- Platform-Ontology — The two-layer vocabulary: architectural domain concepts vs. computational engine modules.
- Architectural-Spatial-Configuration — Case studies comparing topological intent to resolved spatial forms.
- Guarantees-and-Non-Guarantees — Formal epistemic rubric distinguishing implemented, verified, observed, and horizon goals.
- Terminology — Concise dictionary of system terms and abbreviations.
- System-Status — Operational domain, maturity levels, and current technical scope.
User-facing workflows, direct manipulation tools, and analytical outputs.
- Getting-Started — Quick-start tutorial: declare hierarchy, pick sequences, and compare configurations.
- Input — The Program panel: Interactive Node Tree, Lexel Code View, and Waggle spatial exchange.
- Boundary — Site boundary capture, dynamic bound/unbound states, and internal keep-out islands.
- Layout — Planar spatial canvas, embedded sequence scrubber, and 2D packing.
-
Analysis — Quantitative compactness metrics (
$I_{\text{comp}}$ ), topological graph divergence ($\Delta J$ ), room proportion, and solar analysis. - 3D — Volumetric 3D spatial massing, camera cover-lock, and vector/raster snapshots.
- Batch — Parallel computation across the 24 canonical sequence operators.
- Teach — Structured intent documentation, pre-flight cache validation, and dataset commit.
- Report — Automated multi-page architectural spatial dossiers and PDF export.
Discrete mathematics, formal models, parser architecture, and authoritative specifications.
-
Canonical-Conformance-Specification — Authoritative Specification: Mathematical invariants, sequence operator derivations (
$\mathcal{Q}_{24}$ ), topological semantics, and dataset serialization encoding. -
Grammar — Canonical
.hywabstract syntax tree, tokenization rules, and container semantics. - Hygrid — Custom integer hybrid orthogonal-hexagonal lattice, parity rules, and anisotropic neighbour tables.
- Formal-Model — Mathematical formulation of spatial transformation functions and cluster expansion.
- Determinism — Bit-exact reproducibility invariants and zero-divergence validation across engine versions.
- Parsing — Lexel token parsing, structural validation, and tree deserialization.
- Technical-Architecture-Expanded — High-performance WebAssembly compilation, WebGPU rendering pipeline, and memory layout.
Empirical benchmarks, research hypotheses, and spatial datasets.
- Research-Questions — Core computational questions, including the emergent circulation hypothesis.
- Benchmarks — Empirical WebAssembly performance, scaling experiments up to 1,000 nodes, and determinism benchmarks.
- Experiments — Canonical topological comparisons and structural variance studies.
- Data-Collection — The HYWE Spatial Dataset, Base34 coordinate compression, and the Hynteract ingestion gate.
- Distributed-Analysis — Future multi-agent architectural analysis and evaluation networks.
