An evidence-governed research programme in resonant geometry: crystal and resonator physics, typed coordinate systems, signal and phase analysis, and the provenance machinery that keeps them honest.
Lore proposes. Mathematics translates. Software attacks. Evidence decides. Provenance remains.
That sentence is the architecture, not a slogan. Every lane below takes a claim, translates it into something computable, attacks it with tests, and records the verdict — including when the verdict is no. Most physical claims in this repository have been refused or left unmeasured, and those results are published beside the positive ones.
| I want to… | Go here |
|---|---|
| See something work in 60 seconds | Map workbench ↓ |
| Validate a crystal, design parts, export build sheets | RGCS Design Studio ↓ |
| Model a crystal specimen | Crystal & resonator lane ↓ |
| Understand the typed maths core | Foundations ↓ |
| Browse everything | Documentation index |
| Know what's proven and what isn't | Claim boundaries |
The typed, deterministic mathematics everything else is built on.
| package | what it is |
|---|---|
rgcs_core |
RGCS v2 computational core — deterministic, typed, tested implementation of the v2 mathematical model |
rscs_core |
RSCS 1.0 — typed coordinate/state-space foundation: periodic phase ambiguity, nested frames, uncertainty, provenance |
rscs2_core |
RSCS 2.0 — capability-aware multiphysics platform: anisotropic FEM, piezoelectric solver stack |
rgcs-v4 --help # RSCS 2.0 multiphysics CLIThe lane the system is named for: resonant geometry in real materials.
| package | what it is |
|---|---|
r1013 |
Custom crystal specimen workflow — schema, Christoffel API, FEM + convergence + piezo boundaries |
resonator_platform |
Closed-loop platform: design → simulate → fabricate → fixture → excite → measure → identify modes |
fkey_instrument |
Frequency-key harmonic excitation and the ESP32-CYD instrument |
rgcs_surface_wave |
Phase-gated dielectric-loaded annular surface-wave research model |
r1015a |
Scale A mechanical crystal lane |
rgcs --help # unified crystal workflow CLIGuides: Calibration · Bench hardware · Canonical 110 mm case study
RGCS is a coordinate, mapping, signal, and provenance research workbench. It converts received or user-supplied numeric vectors into reproducible binary/octal parse receipts, candidate Earth-root positions, great-circle paths, and polygonal map regions.
| package | what it is |
|---|---|
rgcs_coordinate |
Typed structural decoder/encoder for the 30-bit candidate codec |
cwatlas |
CW Atlas and bidirectional geocoder |
r1053 |
V1 Earth-root map workbench — parse, path, polygon, live builder |
| package | what it is |
|---|---|
r13 |
Cross-domain discovery architecture: response functions, Green functions, S-matrices |
r15 |
Experimental phase infrastructure — turns the R13 architecture instrument-ready and calibration-bound |
These exist so that speculative source material can be attacked rather than believed. None of them is load-bearing for any established result.
| package | status |
|---|---|
cspc |
Crystalline Spacetime Coordinate Program — lore-derived, evidence-gated |
pmwr |
Phase memory / worldline multipath recovery / Phryll translation hypothesis |
consciousness_lane |
Quarantined under an explicit contract; isolated from every other lane |
The most immediately demonstrable lane, and the newest.
python -m pip install -e .
python -m r1053 parse 168930443
python -m r1053 map 168930443
python -m r1053 path 167849523 168930443
python -m r1053 polygon 165876523,165892743,165892763,165892783
python -m r1053 servedistance 178.846 km (11.93 depth-9 cell edges)
initial bearing 18.41 deg
midpoint 42.891736, -79.738486
cross-check 3 formulas agree: True (spread 2.67e-11 km)
The polygon builder takes any number of vectors — type one and press Add, paste a comma-separated list, Remove any row, reorder, or hit Order by bearing. Area, perimeter, centroid and the self-intersection check recompute on every edit.
→ Quickstart · User Manual · Map/Path/Polygon Guide
One vector. One octal word 1170616713. One branch. Two positions 5121.7 km apart —
both fitting all three calibration anchors to machine precision.
The projector matrix is scale-invariant: 9 entries, 8 free parameters, fitted against 6 constraints from 3 anchors. A 2-dimensional family survives, and two of its members put the same vector on different continents. A zero anchor residual is arithmetic, not evidence.
Vertex POSITIONS are projector output and remain underdetermined under V1-B01/B02.
Path, polygon, distance, perimeter, centroid, and area geometry are exact for the
selected vertices.
The tool verifies geometry.
It does not verify that a candidate vertex is physically true.
Open blockers B01–B07, stated in full with the numbers that make them real: docs/BLOCKERS_B01_B07.md — pinning irreproducibility (B01), three anchors cannot test an eight-parameter law (B02), the branch-117 conflict (B03), the cell-scale n=1 reading (B04), no coastline data (B05), the Saint-Frédéric proxy (B06), no transport bridge (B07).
| command | what it opens |
|---|---|
rgcs-lab serve |
Local web hub — nine inspectable modules, each with receipts and claim badges. Loopback, no telemetry. |
rgcs-workbench |
PySide6 desktop research workbench |
rgcs-workbook |
Workbook / reporting CLI |
rgcs-coordinate |
Coordinate codec CLI |
python -m r1053 |
Map workbench CLI |
The desktop workbench (rgcs-workbench) has two modes:
| Mode | Use it for |
|---|---|
| Design Studio | guided crystal validation, certification sheets, printable Phyrll generator templates, coil/pulse design, annular ring prototype design |
| Advanced Scientific Workbench | source library, specimen/model editors, experiment builder, comparison views, reports and bundles |
| I want to… | Open | Output |
|---|---|---|
| Validate a crystal | Crystal Validator | geometry receipt, derived values, certification PDF |
| Design a crystal-first Phryll cone | Phryll Generator v2 | custom cone + coil sleeve: SCAD, STL, 3MF, DXF, SVG, PDFs, receipts — single file or full bundle |
| Design coils and pulses | Coil / Pulse Designer | wire estimates, pulse table, sidebands, build PDF |
| Design an annular ring | Annular Ring Designer | ring diagram, masks, SVG/SCAD, engineering PDF |
| Use frequency keys | Frequency Key Library | sourced key list, sidebands, key relations |
| Make binaural / frequency-key audio | Frequency Key Studio | WAV, recipe JSON, session PDF, YouTube metadata sheet |
| Inspect the research model | Advanced Scientific Workbench | models, sources, experiments, reports, bundles |
Install and launch:
python -m pip install -e ".[desktop]"
rgcs-workbenchGuides: INSTALL.md · Design Studio · Crystal Validator · Phryll Generator v2 · Coil / Pulse Designer · Annular Ring Designer
Every exported sheet carries a claim boundary: designs and estimates are model outputs and reproducibility records — they do not by themselves validate any anomalous physical effect. Measurements decide.
RGCS does not claim that anomalous sources, craft, crop formations, physical propulsion, or non-human communication are proven.
The longer form, emitted by the software itself as r1053.certificate.CLAIM_BOUNDARY:
RGCS V1 is a coordinate/research workbench with a candidate Earth-root projection. It can parse and emit structured vector receipts, reproduce the V1 calibration artifacts, and classify residuals under declared cell-scale and operational-envelope hypotheses. It does not prove physical craft, alien sources, crop-circle authorship, Phryll propulsion, or metric engineering.
Two more, at the root of the repository: SCIENTIFIC_BOUNDARIES.md — what the project has not established, with the measured evidence distribution — and NON_CLAIMS.md, everything not claimed, including claims withdrawn after review.
Refusals are results. A sample of what has been recorded rather than buried:
| finding | class |
|---|---|
| 4096 Hz is not the electromagnetic surface-wave carrier — the annulus is electrically tiny there | NULL |
| Space-time nonreciprocity does not engage at 16 Hz modulation | NULL |
| A constructed harmonic field model did not converge and was withdrawn | NULL |
| An isolated distribution has zero self-force; no configuration produces unbalanced momentum | NULL |
Full ledger: negative_results/
| result | verdict |
|---|---|
| Canonical 110 mm eye-node study | UNCERTAINTY_OVERLAPS_CONVENTIONAL_NODE — the candidate sits 3.906 mm from the nearest conventional comparator and the uncertainty interval overlaps it. Not a distinct node. |
| Quartz mechanism firewall | MECHANISM_NOT_IMPLEMENTED_FOR_MATERIAL — a mechanism absent from the model is not implemented. That is a statement about software, never a claim of physical impossibility. |
Not implemented and does not exist are different claims. Only the first is ever made here. The same rule governs every lore and provenance source — they may motivate a test; they may never stand in for one.
→ Full documentation index — every document, grouped by lane.
Fast paths:
| Quickstart | clone to a working map in five minutes |
| User Manual | the map workbench, with real screenshots |
| Architecture | how the packages fit together |
| Claim Boundaries | claim classes; what is and isn't established |
| Claim Register | machine-lintable register — EST / DER / HYP / SRC / ENG |
| Blockers B01–B07 | the open problems, unsoftened |
| Manuscripts · Examples | long-form treatments; runnable inputs |
Superseded documents live in docs/archive/ with correction banners.
They are never deleted.
pytest -q # everything
pytest tests/test_r1053_v1.py -q # map workbenchThe suite asserts the boundaries, not just the behaviour: that fitted residuals are labelled non-evidential, that nulls travel with the claims they qualify, that no document asserts proof, and that every blocker keeps its detail and clearing condition.
Verification status at v8.4.0 (Design Studio): 8362 passed, 9 skipped, 1 deselected, exit 0.
Verification status at the frozen v8.3.0 release commit:
8129 tests passed, 15 skipped, 1 deselected, exit 0 (expect: 8129 passed);
proof-bundle checksums 115/115 via rgcs-v4 verify-checksums. Counts derive from a
real pytest run recorded in docs/v4/RELEASE_METADATA.json — a document that drifts
from the recorded number fails the build.
The one deselected node is
tests/regression/test_generator_determinism.py::test_generator_deterministic, a
byte-equality check requiring the archived v2.0.0 build environment. Hosted CI
deselects exactly that node id under policy D-V3-04. It is the only known
environment-dependent test in the repository.
current release: v8.5.3 — polish: autosave, undo, render control, library filters, neutral naming
workbench RCs: workbench-v1.0.0-rc1 / -rc2 (pre-releases, shipped)
Current release: v8.5.3.
Frozen history — v2.0.0 (archive/v2.0.0/), v3.0.x, v4.0.0, v4.1.x — tags and records
are never modified. The DOI badge is the latest minted DOI (v3.0.1).
The pending candidate publishes the map workbench as software, not proof.
MIT licensed · Author: Andrew Green · Citing: CITATION.cff