293a82c5e8
John looked at R41's proof shots and named three faults. All three are fixed, at zero draw cost, and every lane corrected somebody's numbers on the way — including its own. E-LOOK (§42.1+§42.2a) — THE PREMISE WAS WRONG. "The pub props have no texture" (my brief, quoting E's own R41 note) is false: all 37 carry COLOR_0 and three.js renders it. The colour was BLEACHED — jukebox at saturation 0.02, 15 assets under 0.05; what the source library called colour was baked AO. So the fix was a palette imposed on colour already present, in place: 46 assets, draw delta +0 and tri delta +0 on every one, proved the honest way — the concatenated POSITION/NORMAL/TEXCOORD/index bytes are sha1-identical to R41, so footprints and origins provably did not move. (A first-pass bbox check ignored node rotations and "found" 3 phantom violations; thrown out for the byte comparison.) The banquette was 1990s office teal and went burgundy; the beer umbrella was white and went brewery green. THE DECK, diagnosed before being touched: 7 greys, chassis 0.05 luminance on a 0.033 bench = 1.21:1, and the one high-contrast feature was INSIDE-OUT (white ring around a dark disc; a turntable is a dark record on a light platter). Now Technics silver at 6.74:1, black vinyl, red start/stop. Node contract re-read from the shipped GLB: 4/4 + 1/1 PASS. Filed to R43, not fixed: deck_1200_rigged has NO PLINTH (bottom 43mm of 119mm lost to R41 join damage). E-CAST (§42.3) — 19 of 24 ship. Rejected: comical_luchador_01 (Mixamo Ch43, purple mask reading EL CHUPACABRA, identified from the file's own material name), a 3-heads-tall cartoon child, a robed wuxia elder, a WW2 infantryman — and dj_phrtt_01, AN UNCLOTHED BODY whose only texture is a bare-skin atlas, which was in the DEFAULT crowd and live on the depot. It is the R41 deny-list's failure class through a different door: an innocent name from another pipeline. Only rendering all 24 and LOOKING found it. Ruling 2 (prefer CC0) was REFUSED ON MEASUREMENT: Quaternius is 25 nodes / 23 joints and ZERO of 64 canonical tracks bind — it would have loaded silently and never animated. Replacements came from Mixamo's Ch## pool, filtered to the contract-clean 65-bone family (candidates ran 65..99 bones; named characters carry facial/hair bones). THE LICENCE GAP WAS UNREAD, NOT MISSING: every source FBX carries ApplicationVendor "Mixamo, Inc." internally. Adobe royalty-free, no attribution owed. B's amber rows are now real. B (§42.5) — MY ASSIGNED FIX WAS USELESS AND B PROVED IT. document.fonts.ready shipped, then measured: the atlas canvases are already byte-identical across boots, the game ships NO webfonts (document.fonts.size === 0), and ready fires 50ms BEFORE the first fillText. The real cause of unreproducible bookmarks is the wall-clock CROWD — 983 changed px of 921,600, ALL in the pedestrian band; 0 px with the crowd out. New DBG.freeze() takes stable bookmarks 2/5 -> 5/5. Also found: ?pop=0 boots 140 citizens (0 is falsy against the default). THE BENCH: D's filing correct on both counts. Front-vs-road 90.0deg -> 0.0deg, min=med=max over 60/60 instances; the streetlight arm now reaches 1.05m over the road where it reached 0.00. Both used a base yaw that pointed +Z down the street while their own comments claimed otherwise — the side term was already right, which is why it survived 41 rounds of reading. Also: on any box with CDN access 9 of 60 benches are the depot GLB, 90deg out from the house law — normalised at the geometry, once, at load. benchStops(plan) exported as one truth. C (§42.2b) — 122 vs 123 WAS THE CAMERA ASPECT, and nobody was wrong. Draw counts are post-frustum-cull and the sweep camera inherits the window aspect: 1.25->116, 1.60->122 (C's harness), 1.778->123 (F's), 2.37->124. Pinned MEASURE_ASPECT=16/9 with a 1.25 control. E's "keep the bench dark" advice was measured and declined with reason (deck-vs-bench was already 4.73:1, over threshold). "Reads as a black box" is a claim about pixels, so C measured pixels: near-black in the booth crop 57.5% -> 26.7%, distinct colours 228 -> 389. Bench rebuilt 3->7 boxes with an open record bay; headphones on the mixer; faceDoor aims booths at the player's entry — alignment within 45deg went 49% -> 92%. Two bugs found by looking: a fitting could be planted INSIDE A WALL (0.77m of a 0.86m half-depth in plaster), and every booth part sat 25mm sunk into its own worktop. D (§42.4) — five swaps IN PLACE at the same index (pickRig indexes these arrays and pedIndex is in the identity signature and the impostor atlas; removal would renumber the town). Identity signature pre- vs post-swap: 0 differing lines of 150 — who each citizen is did not move, only which body they wear. dj_phrtt_01 verified gone AT THE WIRE, not from source: a fresh default boot fetches 32 ped GLBs, none retired. THE CONTROL ARM FOUND A 41-ROUND-OLD BUG: the ped being RETIRED, comical_boy_01, binds only 3,132 of 3,460 tracks — six thumb bones lack the mixamorig prefix. It had been silently dropping six tracks per clip since it shipped. Testing the thing being thrown away found it. Imported B's benchStops mid-round: sitters now 0.0deg to the road, signatures byte-identical across the rewire. Corrected F's filing: the browse smoke was flaky-by-construction (a 16-25s race against a live sim), not deterministically red. F (§42.6) — qa.sh --strict 24 passed / 0 failed / 0 warn / 0 skipped. New Gate 11: period law (0/5 banned; CONTROL: R41's roster trips all five) and roster licence (24/24, 0 unverified; CONTROL: an amber row turns 20 red). r41_shots.py promoted warn -> strict. R41'S BYTE-HASH BAN IS REVERSED AND ITS STATED REASON WITHDRAWN AS WRONG. A gate may pin a hash under five in-run assertions (classic+flags boot, freeze().stillLive all-false, a second freeze removing 0 draws, settle until DBG.info() stops changing, hash with the dbg panel hidden): 5/5 bookmarks now pin to one sha256 over 3 fresh boots. Two findings got it there — the chunk streamer (a fixed settle is a bet on load) and the fps readout, which with everything else frozen was the last thing moving at 53-62 px. ?pop=0 fixed (Number.isFinite) — a no-op for 35 rounds. Three red flags_check gates fixed. Street: noon 282 · NIGHT 291/300 — the nine-draw margin INTACT, every R41 row reproducing to the digit. Interior at pinned 16:9: 123 @ dept/auto, margin 227. selfcheck 157,647/157,647, 0x5f76e76 unmoved. No tag — John's playtest rules the epoch. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
17 KiB
PROCITY pipeline (Lane E) — how to re-run everything
The asset pipeline: audit ultra's library → normalize the good bits to house GLB law → publish
to the 3GOD depot → ship web/assets/manifest.json (the contract Lanes B/C/D read) → fill facade
- prop gaps. Everything is optional at runtime — the game runs asset-free; this lane makes it gorgeous, it doesn't make it work.
Where things live
| what | where |
|---|---|
| ultra (library + Blender) | ssh johnking@100.91.239.7 — key auth, no coreutils/timeout, py3.9, Blender at /Applications/Blender.app/Contents/MacOS/Blender (not on PATH) |
| model library | ultra ~/Documents/3D=models/{shop-fittings,street-furniture,furniture,props-scenes} |
| 3GOD depot | https://digalot.fyi/3god — GET /a/<f>, GET /api/list, POST /api/upload?name=<f> |
| OpenRouter key (skins) | env OPENROUTER_KEY or ~/Documents/dealgod/lore/keys.env (never committed) |
| depot password (publish) | env GOD3_PW (never committed; ask John) — cookie = hmac(GOD3_PW,"3god-ok") |
| Ollama (free LLM) | http://100.89.131.57:11434 (qwen3) — for names/lore/cataloguing, unused so far |
| MODELBEAST (free local gen) | this box, ~/Documents/MODELBEAST + its own queue on :8777. flux_local = FLUX.2-klein-4B text→image on MPS (the facade skins — ungated, weights cached). sf3d/trellis_mac = image→GLB, local + PROVEN, weights already cached — trellis_mac (TRELLIS.2 1024_cascade) went 15/15 on thriftgod's Tier-1 props (Lane G/ultra, R1); sf3d = seconds-fast draft. Also live on the queue: bg_remove_local, mflux_image_edit, seedvr2_upscale. (The old "HF-gated / setup needed" note here was stale — flagged by ultra, corrected R22.) |
The steps (each is idempotent / resumable)
1. Audit — AUDIT.md (free)
scp pipeline/glb_stat.py johnking@100.91.239.7:/tmp/
ssh johnking@100.91.239.7 'cd ~/Documents/3D=models && python3 /tmp/glb_stat.py --json shop-fittings street-furniture furniture props-scenes' > pipeline/_audit_stats.json
glb_stat.py is a dependency-free GLB parser (tris, bbox, texture sizes, Draco flag). Findings +
licensing triage are written up in AUDIT.md.
2. Normalize — normalize.py → 16 GLBs + thumbs (free, runs on ultra)
scp pipeline/normalize.py pipeline/batch.json johnking@100.91.239.7:/tmp/
BL=/Applications/Blender.app/Contents/MacOS/Blender
ssh johnking@100.91.239.7 "\$BL --background --python /tmp/normalize.py -- /tmp/batch.json /tmp/procity_norm"
scp johnking@100.91.239.7:/tmp/procity_norm/_results.json pipeline/_normalize_results.json
scp 'johnking@100.91.239.7:/tmp/procity_norm/thumbs/*.png' web/assets/thumbs/
scp 'johnking@100.91.239.7:/tmp/procity_norm/*.glb' pipeline/_normalized/
Enforces the house GLB law (metres, +Y up, base origin, ≤5k tris, WebP ≤1024, no Draco), renders a
256px thumbnail each. Extends MESHGOD/scripts/finish_glb.py (solidify/decimate/bright/base-origin)
- adds texture-webp, height-normalize, outlier cleanup, per-asset up/yaw. Edit
batch.jsonto add assets; always eyeball the thumbnails — node-transform-distorted sources (see AUDIT rejects) look fine in-frame but have wrong footprints.
3. Manifest — build_manifest.py + validate_manifest.py (free)
python3 pipeline/build_manifest.py # → web/assets/manifest.json
python3 pipeline/validate_manifest.py # Lane F gate: 0 errors = green
python3 pipeline/validate_manifest.py --depot # stricter: require GLBs live on the depot
build_manifest.py derives skins from web/assets/gen/ + fittings from _normalize_results.json;
the curated facade→shop-type maps live at the top of that file (edit there when adding skins).
4. Publish — publish.py → 3GOD depot (the one outward-facing step)
python3 pipeline/publish.py --dry-run # free: list what would upload
export GOD3_PW=... # ask John
python3 pipeline/publish.py # POST each GLB + thumb to the depot (idempotent)
5. Facade skin gaps — gen_skins.py --local (DONE, free, on-device)
The 23-skin gap pack is already generated + harvested + mapped. To re-run or extend:
python3 pipeline/gen_skins.py --local --dry-run # free: list the pack, no GPU
python3 pipeline/gen_skins.py --local # MODELBEAST flux2-klein-4b on MPS → .genraw/ (~5.7s/img)
python3 pipeline/gen_skins.py --harvest # → web/assets/gen/facade-*.jpg, then re-run build_manifest.py
--local drives ~/Documents/MODELBEAST/server/operators/flux_local/run.py directly (no server,
no key, weights cached). It runs one GPU job at a time. Add skins by editing PROMPTS in
gen_skins.py and their shop-type map in build_manifest.py (FACADE_TYPES). Always eyeball
.genraw/ and delete rejects before --harvest. Cloud fallback (drop --local): OpenRouter
google/gemini-3.1-flash-image, ~$0.004/img, needs OPENROUTER_KEY. Manual flow: SKIN_PROMPTS.md.
6. Bespoke hero props — gen_props.py (MODELBEAST) (round-3: DONE, free, on-device)
The meshgod_batch.json gap list now goes through MODELBEAST instead of fal.ai — free, local:
python3 pipeline/gen_props.py --dry-run # list props + what's done
python3 pipeline/gen_props.py --concepts # flux_local concept images → .genprops/*.png (eyeball!)
python3 pipeline/gen_props.py --mesh --only glass-case # trellis_mac 1024_cascade image→GLB (~5-8 min)
python3 pipeline/gen_props.py --mesh # all props with a concept, no glb
# then normalize (uses local Blender), stage, catalogue:
BL=/Applications/Blender.app/Contents/MacOS/Blender
"$BL" --background --python pipeline/normalize.py -- pipeline/batch_props.json pipeline/_normalized web/assets/thumbs
cp pipeline/_normalized/_results.json pipeline/_props_results.json && python3 pipeline/build_manifest.py
5 shipped (arcade_cabinet, listening_booth, drinks_fridge, milkshake_mixer, novelty_record) + cash-register
reused from depot. magazine-rack → primitive (thin-wire gotcha); glass-case/bus-shelter → fal.ai
fallback (TRELLIS can't do transparent glass; sf3d --draft gives a featureless blob). TRELLIS meshes run
heavy (25–77k tris) — see AUDIT provenance. Eyeball every thumbnail; ≤2 local attempts then fal (John's go).
⚠️ Finishing a trellis_mac mesh — ultra's post-mortem (Lane G, R1) — folded in verbatim
⚠️ Do NOT run trellis_mac output through the MESHGOD finish farm — its fill_holes+decimate shreds 200k-tri marching-cubes meshes (proven on ultra, post-mortem in dealgod memory). The working pipeline: raw GLB →
npx @gltf-transform/cli optimize --compress draco --texture-compress webp --texture-size 1024 --simplify-error 0.01→ 0.4–1.8 MB at raw-quality look.
Credit: Lane G / ultra (G-progress.md), proven post-mortem — it cost them a shredded batch so it
shouldn't cost us one. One amendment, per Fable's Ruling 1 (docs/LANES/FABLE_TO_LANE_G.md):
NO-DRACO WINS — the depot speaks uncompressed GLB and PROCITY's loader stays decoder-free. So drop
--compress draco; everything else in ultra's command stands (WebP ≤1024 is already our law):
npx @gltf-transform/cli optimize IN.glb OUT.glb \
--texture-compress webp --texture-size 1024 --simplify-error 0.01 # NO --compress draco
Files land ~2–3× bigger than the Draco'd version and that's fine — the depot holds ONE law.
Note our own in-house paths remain valid for their own cases and are NOT the finish farm:
bake_lowpoly.py (voxel-remesh + bake, for shell-soup that must hit a low budget) and
decimate_props.py (collapse, detail-preserving — see the R13 instrument diet).
7. Local-depot mode — loaders.js ?localdepot=1 + stage_local_depot.py (round-3: DONE)
Lets Lanes B/C/F validate GLB loading with no network / before publish:
python3 pipeline/stage_local_depot.py # mirror _normalized/*.glb → web/assets/models/ (git-ignored)
# then load the game with ?localdepot=1 — depot:<file> resolves to assets/models/<file> instead of the CDN
8. Audio pack — gen_audio.py (round-11: DONE, free, on-device)
Tool choice (Fable asked E to pick + document): pure procedural synthesis in numpy, not a neural model. MODELBEAST has no audio operator, and for this pack synthesis wins outright — $0, no model download, deterministic/seeded (house law), perfectly seamless loops (neural gens don't loop), parody-safe by construction (every sample is an original oscillator/noise render — nothing to imitate), and precise ≤25 MB budget control. Pipeline: numpy synth → WAV → ffmpeg OGG/Opus (+ M4A/AAC).
PY=~/Documents/MODELBEAST/venvs/mflux/bin/python # numpy lives in the mflux venv
$PY pipeline/gen_audio.py --dry-run # list the 23-asset pack + budget
$PY pipeline/gen_audio.py # render all → web/assets/audio/*.ogg + *.m4a
$PY pipeline/gen_audio.py --only music # just the music beds
python3 pipeline/build_manifest.py # catalogue → manifest.json audio:{ambience,sfx,music}
23 assets, 8.1 MB shipped (opus+aac): 6 ambience beds (street day/night, rain, 3 roomtones), 13
sfx (doorbell, door, till, riffle, toast, tram bell/rumble, 6 footsteps), 4 instrumental 90s music
loops. Loudness: beds/music loudnorm -16 LUFS, sfx peak-normalised. Ear-ball = spectrogram sheet
docs/shots/laneE/audio_spectrograms.png + numeric loop-seam/peak check (I can't subjectively listen —
final ear-check is John's/F's). Add assets by editing PACK in gen_audio.py + the AUDIO_* maps in
build_manifest.py. WAV scratch pipeline/.genaudio/ is git-ignored; OGG/M4A ship committed.
9. Gig layer — round-12 (v3.0-alpha, ?gigs=1) (DONE, free, on-device)
Three asset streams for one-pub-end-to-end, all $0:
# instruments (5 GLBs): flux concept -> trellis -> normalize, via the --batch override
python3 pipeline/gen_props.py --batch pipeline/gig_instruments.json --concepts # eyeball first
python3 pipeline/gen_props.py --batch pipeline/gig_instruments.json --mesh # ~3-4 min each
BL=/Applications/Blender.app/Contents/MacOS/Blender
"$BL" --background --python pipeline/normalize.py -- pipeline/gig_norm_batch.json pipeline/_normalized web/assets/thumbs
# posters (4 skins): portrait flyers with a blank name-band (B/C overprint like signboards)
python3 pipeline/gen_skins.py --posters && python3 pipeline/gen_skins.py --harvest
# gig audio (3): pub-rock live bed + crowd-walla + applause — already in gen_audio.py PACK
~/Documents/MODELBEAST/venvs/mflux/bin/python pipeline/gen_audio.py --only music # etc.
python3 pipeline/build_manifest.py # catalogues skins.poster + audio gig assets + fittings
Instruments: amp/guitar/bass reconstruct well; thin objects (mic-stand pole, drum hardware) may fall
to D's primitive (round-3 thin-wire lesson — spec allows "primitives when absent"). Posters carry
nameZone [x,y,w,h] in skins.poster. Gig audio is flagged gig:true in manifest.audio.
Cost summary
| step | cost |
|---|---|
| audit, normalize, manifest, validate, publish | free (compute + bandwidth) |
| facade skin gap-fill (23 imgs) | free — done on-device (MODELBEAST flux_local); cloud fallback ~9¢ |
| hero props (round-3) | free — 5 via MODELBEAST trellis_mac; glass-case/bus-shelter on fal.ai list (≤$2.70, John's go) |
| audio pack (round-11, 23 assets) | free — 100% procedural synthesis (numpy → ffmpeg), $0, no model |
Artifacts (not committed — regenerated from the steps)
pipeline/_normalized/ (staging GLBs, live on depot), pipeline/.genraw/ (raw skin gen).
Committed: the scripts, *.md, batch.json, meshgod_batch.json, _audit_stats.json,
_normalize_results.json, web/assets/manifest.json, web/assets/thumbs/, docs/shots/laneE/.
10. Motion library — clips_export.py → clips_pack.py → clips_verify.py (round-41 §41.1: DONE, free, on-device)
46 Mixamo clips → 6 grouped skeleton-only GLBs in web/models/clips/ (3 498 124 B, 0 tris) + a
new web/assets/motion_manifest.json. Source is ultra's ~/Documents/MOTIONLIB/green/mixamo_full/
only (Mixamo = GREEN: usable in games, no attribution, never redistribute as clip packs).
The selection is checked in at pipeline/clips_r41.json — every asset by name, no wildcard.
python3 - <<'EOF' > /tmp/fbx_list.txt # 1. the 46 names
import json;print('\n'.join(c['file'] for g in json.load(open('pipeline/clips_r41.json'))['groups'].values() for c in g['clips']))
EOF
mkdir -p /tmp/fbx && ssh johnking@100.91.239.7 \
"cd ~/Documents/MOTIONLIB/green/mixamo_full && tar czf - $(tr '\n' ' ' </tmp/fbx_list.txt)" | tar xzf - -C /tmp/fbx
BL=/Applications/Blender.app/Contents/MacOS/Blender # 2. FBX -> per-clip skeleton-only GLB
"$BL" --background --python pipeline/clips_export.py -- --list pipeline/clips_r41.json --fbx /tmp/fbx --out /tmp/clip_glb
python3 pipeline/clips_pack.py --stage /tmp/clip_glb \ # 3. merge to <=6 groups + manifest
--out web/models/clips --manifest web/assets/motion_manifest.json
python3 pipeline/clips_verify.py --stage /tmp/clip_glb # 4. gate: 0 errors = green
No retarget is run and none is wanted: the Mixamo bank and PROCITY's peds are the same
mixamorig skeleton (rigs.js _canon folds mixamorig1/4/…). A mixamorig→mixamorig pass through
MIRPAMO would add error and bake the ped mesh into the clip file, destroying the zero-draw
property. MIRPAMO (~/Documents/MIRPAMO, make smoke = its ground truth, green on Blender 5.1.2)
is for foreign skeletons (CMU/BVH) and for rigging raw meshes. Blender-5 gotcha avoided by
construction: FBX goes in, glTF comes out — nothing is ever exported through FBX.
Packer flags (each measured in AUDIT.md R41): --tracks all restores translation+scale (the
engine's _rotOnly discards them, so they ship off by default) · --no-quantize keeps float32
rotations (int16-normalized is ~2× smaller at a worst measured error of 0.0034°) ·
--drop-fingers removes the 40 finger channels = 49.7% of the payload, taking the set to
1.71 MB. Accessor dedup is lossless and always on (saved 1 341 724 B).
⚠️ Licence zoning, ruling 1 (closed R41): MIRPAMO's library/downloads/ also held 3,077
CC BY-NC (never-shippable) Bandai-Namco BVH, unzoned and unmarked. They now live at
library/downloads/red/bandai_namco_1_NC-research/ with the _NC-research suffix on every file
(so mirpamo batch's <rig>@<clip>.glb naming propagates the marker), and MIRPAMO's
library/manifest.json licence label is reconciled to ultra's MOTIONLIB/SOURCES.md.
R41 ships one licence regime: Mixamo. CMU/100STYLE/ACCAD are CC-BY — they wait on 41.5's credits
surface.
11. Palette — palette.py + isolate.py + montage.py (round-42 §42.1/§42.2a: DONE, free, on-device)
The R41 fittings shipped geometrically correct and visually unfinished. They were not
untextured — they carry COLOR_0 and three.js renders it — they were bleached (15 of 37 pub
props below 0.05 saturation). The fix is a palette imposed on colour that is already there, which
costs zero draws: the COLOR_0 accessor is rewritten in place, so primitive count, triangle
count and every POSITION/NORMAL/index byte are untouched. Adding baseColour maps instead would
have re-imported the cost merge_prims.py bought back in R41 (14,140 → 117 interior draws).
python3 pipeline/palette.py --report web/assets/models/procity_fit_jukebox_01.glb # what colour IS it?
python3 pipeline/palette.py --spec pipeline/r42_palette.json \
--src pipeline/.props_orig/r42 --out pipeline/_normalized # repaint the named set
python3 pipeline/strip_transmission.py --dry-run pipeline/_normalized/procity_fit_*.glb
python3 pipeline/check_contract.py --list pipeline/_normalized/procity_fit_deck_1200_rigged_01.glb
BL=/Applications/Blender.app/Contents/MacOS/Blender
"$BL" --background --python pipeline/thumbs.py -- web/assets/thumbs pipeline/_normalized/<set>.glb
GOD3_DEPOT=... GOD3_PW=... python3 pipeline/publish.py --files pipeline/_r42_publish_list.txt
⚠️ palette.py is NOT idempotent. ramp and tint normalise against the source's own
luminance spread, so a second pass re-stretches an already-stretched file. Every output is stamped
asset.extras.procity_palette; a stamped input is skipped unless --force. Point --src at the
pristine copies in pipeline/.props_orig/r42/ (git-ignored) when re-running. Routes, per-asset
stops/rules and the reason for each are in pipeline/r42_palette.json — every asset by name.
Proof tools: isolate.py renders ONE staged scene from a perspective camera at eye height on a
bench of a stated colour (the only rig that can answer "does this read from where the player
stands" — it is what diagnosed the DJ booth); montage.py builds labelled before/after grids from
finished PNGs; contact_sheet.py --pitch M spaces tiles when the set mixes a 2.68 m umbrella with
a 0.2 m ashtray.