The shape read in LANE_A_NOTES was written from reading code, so I built it to
find out if it was true. Flyable: web/dev/laneA_world.html?dbg=1&assets=1&lvl=swept
Nothing here is imposed on anyone: every addition is additive and a level that
opts out behaves exactly as today. C can bin the whole proposal and lose nothing.
WHAT THE PROTOTYPE DISPROVED (my own cost estimates, all four):
- "radial resolution must scale with radius" — wrong. 64 segments is a 0.05u
circle error at r=70. The arena needs it because it fbm-displaces; the tube's
wave is smooth in theta. No change made.
- "cap geometry at the span ends" — a thinko. The canal is continuous; a room is
a fat part of it. Only the fundus dome caps, and it stays an icosphere.
- cost — cheaper than feared: 9 draws / 98k tris worst-frame for a radius-70
cathedral vs 8 / 74k for the L2 corridor. No leak over a full sweep.
- "one schema flag" — held. `segments[].mode: "open"` is the entire ask.
THE TWO REAL COSTS, NEITHER PREDICTED:
1. D's `tile[0]` is a COUNT, not a density — only a texel size if radius never
changes. Measured 6.1:1 smear at r=70. The wall now derives the count from
each ring's own arc length (aRadius attr + uThetaSpan); D's authored numbers
keep their exact meaning at their reference radius, no re-authoring. That fix
then laid a seam down the canal (a fractional count can't wrap) -> rounded to
an integer, seam gone. Not a no-op on corridors: L2's hiatus takes 3 repeats
where it took 4, which is the texel size correctly holding still as the pipe
narrows. Eyeballed.
2. MY OWN PINCH GUARD WAS WRONG and would have blocked C. The fixture scored
1.30 and "failed" the >1.5 law; it was never bent. stats() divided the
tightest turn ANYWHERE by the widest radius ANYWHERE — 996 units apart, a
bend in the radius-13 pylorus against the radius of the sea. Locally its
worst point is 6.4. pinchRatio is now min over s of turnRadius(s)/radius(s).
Provably one-directional (global <= local by construction) so nothing that
passed can fail; real levels GAINED headroom: L2 3.32->4.16, L3 1.91->6.99,
L1 2.11->6.06. This mattered — the guard is what backs the round-1 promise
that C never has to think about curvature.
THE ACID SEA (world/acid.js, `level.acid {from,to,height,biome}`):
Flat emissive #c8ff3a, level and NOT tube-following — that's the mechanic. The
best find of the session: `height` is ONE NUMBER that tunes the level. Measured:
-18 -> 0% of the sea floor is dry shallow, -55 -> 14%, -62 -> 48%. So C's "position
IS the resource" falls out of the radius wobble they already author — the mucus
shallows are just "where the wall rises above the waterline", nothing to place —
and rising acid literally drowns them: free escalation, same scalar their event
pump drives. Also disproved my own claim: the centreline stays level (+/-1u over
700u), so shallows come from radius wobble, NOT the canal's bends. C must author
vertical bends deliberately if they want depth by anatomy.
Stated plainly, not papered over: the waterline is prototype quality (the plane
meets faceted rings and the shoreline steps). Acid does not drain the coat —
depthAt(pos) is the hook and the cost is Lane B's call. Nothing drives height but
DBG until C's events do.
qa GREEN incl. the new provenance gate; spline selfcheck green; L2 corridor
re-eyeballed for regression. Rulings requested from F in NOTES §-> Lane F.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Generated on MODELBEAST flux_local, on-device, $0.00. Lane D's pipeline used
unchanged; **D owns these — ratify, retune or bin them.**
Why A did D's job: the six-biome tint contact sheet could only judge 3 of 6
biomes for real (oral/large_intestine/appendix had no pack and fell back to
procedural), and you cannot ratify a tint against a stand-in. All six are now
textured — the set ART_BIBLE §Biome palettes promises.
- wall_oral_a/b, wall_colon_a/b + normals. Seams 6.35/9.30/14.06/6.37 -> 1.01/
1.08/0.74/1.36 (D's "indistinguishable" band; the shipped pack's weakest is
2.39). Means normalised to ~0.50 by D's derive_maps, untouched.
- appendix wears the colon pack at its own gold tint (TEXTURE_SHARE in
world/index.js) — anatomically it IS colon tissue, and it's ART_BIBLE's
"one texture, two biomes" law for zero bytes. Distinct from the temporary
slug map; keep it when that dies.
- oral_a took three subjects. v1 named an ORGAN ("tongue") and FLUX drew the
organ's silhouette — D's "one urchin of villi" in a field's clothing. v2 named
the tissue but papillae are 3D projections, so perspective gave it a vanishing
point that tiles into starbursts. v3 asks for squamous cell pavement: flat by
construction. Rule for the kit: prompt the SURFACE, not the FEATURE.
Evidence: docs/shots/laneA/round2_oral_prompt_evolution.png
PROVENANCE DEFECT FOUND AND FIXED (-> Lane D):
batch_textures.json recorded, for 4 of 6 walls, the prompt from the framing
experiment D tried and rejected — not the prompt that made the shipped pack.
record() only runs on generation, so reverting a prompt while keeping its image
(exactly what D correctly did: "four winners kept untouched rather than churned")
drifts the record silently. Proved by regenerating esophagus_a at seed 1101 under
both prompts: the code's prompt reproduces the shipped wall, the recorded one
gives the bland wood-grain D described rejecting. Evidence:
docs/shots/laneA/round2_provenance_probe.png. Records re-pointed at the code's
prompt (images untouched — they were never wrong).
New: gen_textures.py --verify-provenance + qa.sh gate 5b, so it cannot rot again.
qa GREEN incl. the new gate; 11/11 provenance verified; all six biomes eyeballed
in engine (docs/shots/laneA/round2_tint_*.png).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Finishes the lane A round-2 list that f0982e7 (the rescued first half) left open:
- Biome-tint re-eyeball under the colorspace law (#4 second half). Verdict: all
six ART_BIBLE tints stand, biomes.js untouched — the wash was wall_material's
round-1 constants, tuned on the raw framebuffer. Retuned: rim pow 2.2->3.0,
rim gain 0.9->0.35, textured curve (0.35+d*.95)*.9 -> 0.12+d*0.60, procedural
d*.8 -> d^2*.65, sheen .10->.04, matcap .22->.14. Rationale documented
in-shader. Verified on a new six-biome contact sheet (?lvl=biomes, SIX_BIOMES
fixture — kept as permanent kit), the real L2 hiatus, and an arena interior.
Evidence: round2_tint_*.png x6 + two *_retuned frames. qa GREEN.
- Stomach-arena shape read for C (#7): recommend the spline-swept room (arena =
"open" mode over an s-span; fundus dome stays a welded icosphere; chained
spheres rejected — crease rings + union collision for a worse look). Full
argument NOTES §-> Lane C + round2_stomach_arena_sketch.svg. Triplanar
deliberately deferred: the swept room obsoletes it where it mattered.
- Ruling #5: harness -> web/dev/laneA_world.html (+DBG.post to the house shot
sink); _fixture.js stays — the spline selfcheck asserts against it, now
documented in-file. launch.json: guts-a on 8145.
- NOTES + progress written for both halves; ROUND2 mid-round box updated: lane
A's round-2 list is closed, only round-3 items remain.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
All round-2 lane sessions were cut off 15:08-15:16 before NOTES/commits; the three prior
commits preserve their work verbatim. This commit is the wiring B was cut off inside of,
plus the run-state boot owes per ruling #4:
- combat/index.js: fiction-id -> archetype resolve via C's getEnemy() (the round's declared
ONE blocker), theta-array + spread fan-out for group spawns, hazards + pickups constructed
and updated (order: enemies -> weapons -> hazards -> pickups, so a torpedo can neutralize
a surge the frame it detonates), pickup score/samples onto the combat:state scoreboard,
combat.reset(fromS) for respawn.
- flight/player.js: kill/shove/refill — the API B's new modules call; kill ignores iframes
by design (boost-dodging through C's lethal acid wall is not a read). Deleted the
crestSpeed fallback guard per its own comment — A's law landed.
- boot.js: run state. checkpoint crossed -> remembered; death -> 2s -> respawn at it +
combat.reset() re-arms hazard timelines (pump deliberately does NOT rewind: no
double-spawns, collected stays collected); gate -> level:complete {stats, par} for E's
medal card. Exported step(dt) so the stepped-sim harness drives the REAL loop.
- ROUND2_INSTRUCTIONS: MID-ROUND STATUS box for the resuming lanes; f-progress entry.
Verified (deterministic stepped sim via boot.step, :8140): full L2 run fires all 47 events
with ZERO console warns (was: 11 spawn no-ops), all 5 hazard telegraphs, 2400 proximity
ticks, 14 pickups collected, finish 157s vs par 180 while surfing 94s of it. Death path:
throttle 0.7 into the finale -> killed by acid s=3386 -> respawn at Cardia Approach 2980 ->
surge re-arms -> second attempt escapes -> level:complete {deaths: 1}. Survivable AND
losable. Evidence: docs/shots/laneF/round2_L2_integrated_full_run.png. qa.sh GREEN.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Landed before the cut: schema v2 (segments[].wave.amp override, validated — breathe stays
biome-owned); hiatus re-authored tight AND calm (radius back to 6.5, amp 0.6, 3.3u steady
clearance); par 210 -> 180 with full rationale (below throttle-only par-pace, above a surf
run — the medal teaches the mechanic); finale surge 21 -> 24 so surfing is the stylish
escape (required 1.20x); surf profile reads A's live CREST_FACTOR; chaseAnalysis() +
escapability laws in validate; NEW --probe builds A's real canal and measures clearance
against the model (all L2 segments clear the 2.5 floor, hiatus 3.75); isMain guard +
importSpline shim — fixes A's selfcheck process.exit killing C's mid-run (qa.sh had been
green on one level out of three).
Cut off before: fly-through of the integrated build (blocked on B), pickup economy handover
(B proposed numbers in balance.js meanwhile — reconcile), L1 encounter design doc, GDD boss
folding, NOTES. Committed by F to protect the shared tree; levels selfcheck 3/3 GREEN.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Landed before the cut: surf as a speed-lock onto world.crestSpeed(s) with surf.authority
(ruling #1) — riding beats throttle and holding a crest is throttle discipline; tuning.js
header corrected to the real wallRho frame (C's flag); enemies.js reads rho as a FRACTION
of safe radius (C's placement law); balance.js hazard + pickup constants; NEW hazards.js
(reflux_surge / aortic_squeeze / ring_gate, self-scheduling telegraphs, reset(fromS) built
for respawn) and NEW pickups.js (all five kinds, shape-coded).
Cut off before: wiring — nothing imports hazards/pickups yet; the fiction-id -> archetype
resolve (round-2 task #1) does NOT exist despite the enemies.js comment saying index.js
does it; player.kill/shove/refill are called but were never added; checkpoint/respawn and
gate -> level:complete not started; NOTES unwritten. F lands the glue next commit.
Committed by F to protect the shared tree.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Landed before the cut: crestSpeed(s) + CREST_FACTOR 1.6 (ruling #1) with selfcheck
asserts; radius blend widened +/-12 -> +/-25 (C's #4 dependency) + no-cliff selfcheck;
per-segment wave.amp override as per-vertex aWaveA (ruling #8); colorspace law in the
wall shader (ruling #2); TBN normal maps + matcap + dual detail layers (D's perturb(),
trap documented in-shader); sample(s, out) v1.2; slug map shrunk to the two real
mismatches. Evidence: docs/shots/laneA/round2_L2_*.png.
Cut off before: NOTES/progress, stomach-arena shape read for C (task #7).
Committed by F to protect the shared tree; spline + qa selfchecks GREEN at commit time.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Wired B's player/combat + event pump, D's assets, C's level pick into boot.
Fixed qa ESM gate (node --check no-op, found by B) + added spline selfcheck.
TECH: world contract FROZEN v1.1, bus events ratified, colorspace + crest-speed
laws (CREST_FACTOR 1.6 — B proved surfing lost to throttle at 1.0). Stub complies.
Verified integrated build via 60s stepped sim: 9 draws, 0 errors, 0 asset misses;
one known gap (fiction-id spawn resolve) confirmed and assigned as B's top item.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Textures (6 walls + normals + wet matcap, $0, ~64s on the m3ultra MPS):
- FLUX ignores "seamless tiling" (seam_error 3.7-17.9); a 4-way cosine
partition-of-unity blend of the four half-rolls fixes it exactly -> 0.87-1.32
- levels-normalised the pack to one exposure (means were 0.18-0.32); these are
detail maps multiplied into a biome tint, so per-prompt exposure read as a
broken tint rather than as dark tissue
- prompt-kit experiment rejected + recorded: PROCITY's "uniform coverage" clause
in the STEM cured centred subjects but flattened the pack; framing words now
live only in the two subject prompts that needed them
Audio (1 bed + 4 sfx, 4.2s render, 0.61/10MB, ogg+m4a dual-ship):
- bed-esophagus 24.000s, loop seam 0.32 (wrap step 3x smaller than inner step)
- cascaded lowpass poles: one-pole at 900Hz left 16kHz only ~25dB down = hiss
- rewrote the spectrogram sheet (per-clip peak, log freq) — the first one was a
saturated rectangle, and evidence you can't read is not evidence
assets.js: miss ledger + misses() — Lane A found that a drifted slug falls back
procedurally forever with no error. Each distinct miss now announces itself.
Also: shot_sink.py (canvas -> docs/shots/laneD, correctly named) and a dev
texture viewer that imports the stub world read-only for the eyeball law.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Implements THE WORLD CONTRACT (TECH.md) in web/js/world/; drop-in for the stub.
boot.js now runs this on C's L2_esophagus (3600u, hash cefc4f83), zero console errors.
- spline.js: centreline + arclength LUTs + parallel-transport frames + radius law +
peristalsis phase + project() + hash(). No three.js import, so qa can run its
14-assertion selfcheck headless.
- curviness is solved from a curvature budget, not authored as an amplitude: the pinch
guard caught the tube folding through itself on a wide+curvy join (turn radius 9.9 vs
tube radius 17.2). C can no longer author a pinch.
- peristalsis: global OMEGA, k(s) = OMEGA/flow(s), phase K(s) = integral k ds. A crest
therefore travels at exactly the local flow speed, so "surf the crest" == "ride the
current". Esophagus wave amp 0.9 -> 1.4 (measured): moves wallRho for Lane B.
- tube.js: chunked streaming, seams aligned to biome joins, arenas own their s-span.
- arena.js: displaced icosphere shells v0 (three's polyhedron detail is 20*(detail+1)^2,
not 20*4^detail — 720 tris was far too coarse); fog sized to the room, since biome fog
tuned for 100u corridors renders C's 360u acid sea as a black rectangle.
- index.js: prefers assets.texture() over get() (get returns raw JSON a shader can't eat),
plus an explicit slug map — D's wall_smallint_a vs biome id small_intestine would miss
silently forever under the assets-optional law.
Measured (C's real L2 + D's real pack, 1920x1080, renderer.info): 8 draws worst frame
(budget 150), 74k tris (500k), no geometry leak across a full sweep, <120ms load,
pinch ratio 3.32. fps not claimed — the screenshot pane runs tabs hidden, pausing rAF.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Flight (js/flight/): tube-mode controller in spline space (s,x,y) — flow-locked
forward motion, throttle as a spring-back lean, boost + i-frames, banking chase
cam on the parallel-transport frame, arcade wall response (shove + graze damage,
never a hard stop). Twin-stick input: WASD/left-stick moves the ship in the disc,
mouse/right-stick aims, Shift/Ctrl is throttle. Gamepad supported.
Combat (js/combat/): lysozyme cannon (pooled, instanced, heat-limited with
hysteresis lockout) + antacid torpedo emitting level:neutralize. Enemy framework
with floater/seeker/turret and turn-rate-limited homing darts, spawned from C's
level:event. Coat/hull model and the full bus surface E builds against.
Measured (deterministic stepper): flow-lock exact (3s at flow 14 -> s=44.0);
boost peak 30.8 u/s; heat 4s fire -> 2s lockout; wall slam 13.3 coat at 11.3 u/s
with forward speed untouched; 10 combat draws at 55 live enemies (budget <=80);
0.10 ms/frame CPU; 0 leaks over 20 create/dispose cycles. fps NOT claimed — rAF
does not run in the automated pane. Evidence: docs/shots/laneB/.
Three findings escalated in LANE_B_NOTES:
- qa.sh gate #1 is a no-op: node --check silently exits 0 on ESM, so the syntax
gate has never checked anything. I shipped a real SyntaxError past a GREEN qa.
Fix + verification handed to F.
- Surf is structurally broken: the peristalsis wave (13.64 u/s) is slower than
the player (19.6 u/s), so level 2's signature mechanic loses to mashing
throttle. Not tunable from this lane; escalated to A/F with options.
- Custom ShaderMaterials need #include <colorspace_fragment>, else ART_BIBLE's
hostile amber reaches the display as rgb(255,26,6). Fixed here; stub/walls
inherit it.
js/flight/dev.html is a dev harness (boot.js has no player wiring yet); it
retires once F pastes the snippet in LANE_B_NOTES.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Data spine:
- Schema v1 ratified (TECH.md, C-owned). Additive over v0: schema/par/next, segment
name, cross-section placement (theta/rho as a FRACTION of safe radius), spread, span,
pickup+gate event types, hazard params. STUB_LEVEL synced (that const only, per
ROUND1 §Lane C.1) — world.hash() unchanged (1bdf001), geometry provably untouched.
- levels/index.js: registry + loader (node+browser, no THREE), real schema selfcheck
(qa's --selfcheck hook is now live) and a pacing simulator (--sim). Consumes the other
lanes' contracts live rather than restating them: A's biomes.js for ids and
wave.amp+breathe, A's SKIN, B's tuning.js hull radius, B's balance.js for scores.
- levels/enemies.js (fiction id -> archetype join table), levels/difficulty.js.
L2_esophagus — the vertical slice, complete: 3600 units, 7 beats, 47 events, 65 enemies,
10 checkpoints, 3 samples, reflux-surge finale. L1/L3 skeletons (shape only).
Anatomy is the level designer: the three real esophageal constrictions are the three
setpieces, and all three enemies are real esophageal pathology (bolus chunk, eosinophilic
esophagitis, candida esophagitis).
Measured (node web/js/levels/index.js --selfcheck|--sim; qa GREEN):
- selfcheck 3/3 valid, exit 0.
- pressure curve 0.27 -> 2.14 -> 4.35 -> 0.69 -> 4.53 -> 6.35 -> 4.40 (breather/pressure/
spike/breather/pressure/peak/finale).
- par 3:30 vs measured 2:47 speedrun / 3:19 par-pace / 4:22 timid; par.score 9000 of a
9400 kill budget (95%), computed from B's scores.
- all 10 checkpoint gaps within the <=30s death-cost law (max 28.8s); min clearance 2.73.
Two contract collisions found and resolved (LANE_C_NOTES):
- Biome ids: authored mouth/pharynx/cardia/pylorus; A's registry has six canonical ids and
rejected all four. A is right — biome = look, segments[].name = anatomy. Conformed.
- Balance: my first enemies.js restated hp/size/speed/score that B owns in balance.js.
Cut to a join table; index.js reads B's scores via scoreFor().
Caught live: A raised esophagus wave.amp 0.9 -> 1.4 mid-session, making the diaphragmatic
hiatus unflyable (2.33 vs the 2.5 floor). Widened 6.5 -> 6.8. The invariant also failed my
own aortic constriction's first draft (radius 7, symmetric squeeze) -> radius 9, directional.
No in-engine flythrough: ?stub=1&lvl= does not compose (snippet offered to F) and the stub
is single-segment, so it could only render beat 1. The pacing evidence is a kinematic model,
not a playtest — it proves L2 is flyable, paced and survivable, not that it feels good.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
