feat(simulation): district-window zoom-ladder renderer (T-1123) #181

2026-07-16T22:31:12+02:00

jpmschweitzer commented

2026-07-16 22:31:12 +02:00

Summary

The orbit-to-jet-plane demonstrator Jeroen asked for after the T-1120 planetary-only screenshots. aliveness_probe gains --render

/ --window / --anchor : renders TRUE-2km district maps (derive_district per cell — not the coarse eager grid) as PNG panels windowed on a settlement anchor.

What

5 panels per window: morphology (17-zone palette transliterated from atlas_marker_overlay.gd MORPHOLOGY_COLORS, sync-commented + unit-pinned against the enum), elev_q, temperature (airless sentinel color), moisture, vegetation; anchor crosshair; km-honest filenames + per-image manifest lines.
believability.rs: new cascade_snapshot_for_body() returning (CascadeSnapshot, BodyParams); cascade_for_body is now a thin wrapper — harness untouched. The probe re-derives the D-203-freed TerrainAnalysis via run_layer1 only under --render.
3 unit tests: palette length pinned to the last MorphologyZone discriminant; district inverse-mapping roundtrip vs derive_district forward mapping; radius-less fallback identity.
Text-only probe behavior without --render verified unchanged.

Evidence produced (not in this PR; preserved at .cache/screenshots/district-windows)

35-PNG zoom ladder for GJ380c/GJ820Bc/GJ244Ad at 524km/131km/33km. Determinism: 5/5 panels byte-identical across passes AND cross-process (md5s reproduced in a fresh run). ~8us/district debug.

Findings filed from first-ever look at these maps (under T-1079)

T-1125 (high): D-227 invented detail does not feed district classification — coherent geography at 524km, gradient stripes at 131km, flat color at 33km; ocean_fraction is raw bilinear (no warp), scatter amplitude slope-slaved.
T-1126: derive_vegetation is ocean-blind (Forest across open ocean).
T-1127: frozen bodies read temperate-green (glaciation_grade unrendered; 17-zone vocab deliberately not reopened).

Verification

Gate green: fmt, clippy --all-targets -D warnings (after an items_after_test_module fix), full cargo test, deny. Tool runtime verified live across 7 window renders.

## Summary The orbit-to-jet-plane demonstrator Jeroen asked for after the T-1120 planetary-only screenshots. aliveness_probe gains --render <dir> / --window <N> / --anchor <substr>: renders TRUE-2km district maps (derive_district per cell — not the coarse eager grid) as PNG panels windowed on a settlement anchor. ### What - 5 panels per window: morphology (17-zone palette transliterated from atlas_marker_overlay.gd MORPHOLOGY_COLORS, sync-commented + unit-pinned against the enum), elev_q, temperature (airless sentinel color), moisture, vegetation; anchor crosshair; km-honest filenames + per-image manifest lines. - believability.rs: new cascade_snapshot_for_body() returning (CascadeSnapshot, BodyParams); cascade_for_body is now a thin wrapper — harness untouched. The probe re-derives the D-203-freed TerrainAnalysis via run_layer1 only under --render. - 3 unit tests: palette length pinned to the last MorphologyZone discriminant; district inverse-mapping roundtrip vs derive_district forward mapping; radius-less fallback identity. - Text-only probe behavior without --render verified unchanged. ### Evidence produced (not in this PR; preserved at .cache/screenshots/district-windows) 35-PNG zoom ladder for GJ380c/GJ820Bc/GJ244Ad at 524km/131km/33km. Determinism: 5/5 panels byte-identical across passes AND cross-process (md5s reproduced in a fresh run). ~8us/district debug. ### Findings filed from first-ever look at these maps (under T-1079) - T-1125 (high): D-227 invented detail does not feed district classification — coherent geography at 524km, gradient stripes at 131km, flat color at 33km; ocean_fraction is raw bilinear (no warp), scatter amplitude slope-slaved. - T-1126: derive_vegetation is ocean-blind (Forest across open ocean). - T-1127: frozen bodies read temperate-green (glaciation_grade unrendered; 17-zone vocab deliberately not reopened). ### Verification Gate green: fmt, clippy --all-targets -D warnings (after an items_after_test_module fix), full cargo test, deny. Tool runtime verified live across 7 window renders.

jpmschweitzer added 3 commits 2026-07-16 22:31:13 +02:00

chore(meta): file T-1123 district-window renderer (activated, dudley) + T-1124 Atlas regional-map-view design — Jeroen's orbit-to-jet-plane bridge gap 96737756dd

feat(simulation): district-window zoom-ladder renderer — aliveness_probe --render/--window/--anchor (T-1123) cf6ef2bc31

Renders true-2km district maps (derive_district per cell, NOT the coarse eager grid) as PNGs: morphology (17-zone palette transliterated from atlas_marker_overlay.gd, sync-commented), elev/temp/moisture/vegetation panels, anchor crosshair, km-honest filenames + manifest. believability.rs gains cascade_snapshot_for_body (cascade_for_body now a thin wrapper); probe re-derives the D-203-freed TerrainAnalysis via run_layer1 only under --render. Determinism proven: 5/5 panels byte-identical across passes AND cross-process (md5s reproduced in a fresh run). ~8us/district debug. First-ever look at the district layer produced the T-1125/T-1126/T-1127 findings (D-227 seam ends at ~500km; ocean-blind vegetation; unrendered glaciation).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

fix(simulation): move probe tests module below Args — clippy items_after_test_module 5e48c85a9a

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

jpmschweitzer commented

2026-07-16 22:36:23 +02:00

Review verdict: APPROVED — Hoshe: zero findings after deep verification (refactor byte-equivalent + harness run; inverse mapping algebra hand-checked + non-tautological roundtrip test; north-up orientation proven; seed/heightmap consistency structurally guaranteed — run_layer1 is RNG-free over the identical heightmap; determinism double-pass confirmed genuine re-derivation; live runs incl. crosshair-overflow edge). Tyre: zero findings — clean seam extraction, re-derive-under---render is the architecturally honest cost vs D-203's cache budget, cross-language palette properly insured by the pin test, independent inverse justified. Merging.

## Review verdict: APPROVED — Hoshe: zero findings after deep verification (refactor byte-equivalent + harness run; inverse mapping algebra hand-checked + non-tautological roundtrip test; north-up orientation proven; seed/heightmap consistency structurally guaranteed — run_layer1 is RNG-free over the identical heightmap; determinism double-pass confirmed genuine re-derivation; live runs incl. crosshair-overflow edge). Tyre: zero findings — clean seam extraction, re-derive-under---render is the architecturally honest cost vs D-203's cache budget, cross-language palette properly insured by the pin test, independent inverse justified. Merging.

jpmschweitzer referenced this issue from a commit

2026-07-16 22:38:38 +02:00

chore(meta): T-1123 to review — PR #181 up

jpmschweitzer referenced this issue from a commit

2026-07-16 22:38:38 +02:00

chore(meta): close district-window batch — T-1123 done (PR #181 merged); T-1125 seam story is the recommended next batch

jpmschweitzer closed this pull request

2026-07-16 22:38:49 +02:00