eb317d9933
feat(stats): stat VoxelForge -- make tile skipping and the column memo observable
Fr0zkaandClaude Opus 52026-08-16 16:01:26 +02:00
7909c4f2ca
docs(opstack): handoff -- 001 chains into 002, and PERF is now aimed
Fr0zkaandClaude Opus 52026-08-16 15:46:30 +02:00
7313201e91
docs: spec CODEX-TASK-002 (column-memo thrash) and aim the PERF work
Fr0zkaandClaude Opus 52026-08-16 15:45:44 +02:00
57002b35cf
docs(opstack): log the CODEX-TASK-001 acceptance-bar correction
Fr0zkaandClaude Opus 52026-08-16 15:40:48 +02:00
6bd5589d53
docs(codex-001): the lumped AllSolid counter cannot prove T1.d -- add the op-stack site
Fr0zkaandClaude Opus 52026-08-16 15:40:23 +02:00
251a1288e0
docs: spec CODEX-TASK-001 (tile-skip stats) and refresh the handoff for the Codex tandem
Fr0zkaandClaude Opus 52026-08-16 15:36:29 +02:00
e4d1fdd7ea
docs(opstack): experimental is pushed and tracked -- correct the flat "never push" rule
Fr0zkaandClaude Opus 52026-08-16 14:41:00 +02:00
41ba3c34a9
docs(opstack): rewrite the handoff for the T1.d-delivered state
Fr0zkaandClaude Opus 52026-08-16 14:36:36 +02:00
e002bd4e2e
feat(opstack): VerticalShafts' connector branch never tested a connector -- test the real capsules
Fr0zkaandClaude Opus 52026-07-29 03:36:34 +02:00
861dc8e109
docs(opstack): WARP SHARE measured -- half the blocking is the box; plus two corrections
Fr0zkaandClaude Opus 52026-07-29 03:28:49 +02:00
b2938d38f1
fix(opstack): the warp bound was the dominant term all along -- 2.0 -> 1.5, and MEASURE it
Fr0zkaandClaude Opus 52026-07-29 03:21:37 +02:00
9733179723
feat(opstack): tunnels get a threshold test -- T1.d is measured at 6 of 40 tiles, 0 violations
Fr0zkaandClaude Opus 52026-07-28 18:45:56 +02:00
87a0b996ec
test(opstack): measure the box verdict on BOTH densities -- the fixture cannot prove a tile
Fr0zkaandClaude Opus 52026-07-28 18:37:45 +02:00
f2fefade4c
test(opstack): the tile sampler never left the (0,0) spine -- widen it to 4x RoomSpacing
Fr0zkaandClaude Opus 52026-07-28 18:25:43 +02:00
8043a613c3
perf(opstack): drop the redundant column test, and measure WHICH primitive class blocks the box
Fr0zkaandClaude Opus 52026-07-28 18:16:23 +02:00
f537648867
fix(opstack): the worm inherits the room source's box verdict -- the actual blocker
Fr0zkaandClaude Opus 52026-07-28 18:10:51 +02:00
03ddcde334
feat(opstack): FRoomGraphSource::EffectOverBox answers spatially -- and AUDIT C2 is fixed
Fr0zkaandClaude Opus 52026-07-28 14:48:18 +02:00
e55f12a9de
docs(handoff): ce409e7 is built -- turn the Underwater section from a task into a question
Fr0zkaandClaude Opus 52026-07-28 14:27:19 +02:00
b533294f5d
docs(opstack): record the green build and rewrite the handoff for a fresh context
Fr0zkaandClaude Opus 52026-07-28 14:13:59 +02:00
ce409e7bb1
test(opstack C2): floor-divide the sampled chunk range, and DIAGNOSE the Underwater 0% instead of guessing
Fr0zkaandClaude Opus 52026-07-28 14:09:54 +02:00
46b507a261
docs(opstack): close the unattended run -- the post-8/8 batch and the reviewer pass
Fr0zkaandClaude Opus 52026-07-28 04:08:16 +02:00
6a8390f11f
feat(opstack): ClassifyTile consumes ClassifyBox for cave archetypes -- the T1.d prize, behind the same opt-in
Fr0zkaandClaude Opus 52026-07-28 04:05:42 +02:00
239c037f35
feat(opstack): the twelve detail modifiers inherit the room source's box verdict
Fr0zkaandClaude Opus 52026-07-28 03:55:39 +02:00
353d504169
feat(opstack): the box fold carries NUMBERS -- amplitude bounds alongside EVoxelOpEffect (DECOMPOSITION 0.2)
Fr0zkaandClaude Opus 52026-07-28 03:53:02 +02:00
64d0e11821
docs(audit C2): the within-strate stale-room suspicion is CONFIRMED by reading -- and it is the default config
Fr0zkaandClaude Opus 52026-07-28 03:48:01 +02:00
10dbe64b06
feat(opstack C3): TunnelNetwork + Underwater in the ported list -- 8 of 8
Fr0zkaandClaude Opus 52026-07-28 03:46:17 +02:00
e479fcdf7d
feat(opstack C2): Underwater reaches the tunnel stack from a second case + its own equivalence check
Fr0zkaandClaude Opus 52026-07-28 03:42:07 +02:00
9591088d34
feat(opstack C1): per-room op override -- eleven detail ops read LocalParams() from the room source
Fr0zkaandClaude Opus 52026-07-28 03:39:38 +02:00
8a303cc085
test(opstack B5): prove the bNearCaveSurface gate -- a voxel outside it is bit-identically untouched
Fr0zkaandClaude Opus 52026-07-28 03:33:34 +02:00
b063d437c1
feat(opstack B4): port Columns (4d), Domes (4g), Pinch (4h), FloorBias + Column op in the test pool
Fr0zkaandClaude Opus 52026-07-28 03:31:55 +02:00
6ec60099d6
feat(opstack B3): port cave Overhang, Cliff, Scallop, Arch (STEP 4c)
Fr0zkaandClaude Opus 52026-07-28 03:26:40 +02:00
ab1a996cc5
feat(opstack B2): port Terrace, LayerLines, Ribbing (STEP 4c) + hoist the room source's per-worker state
Fr0zkaandClaude Opus 52026-07-28 03:22:21 +02:00
6e29cbfe4c
feat(opstack B1): port TunnelNetwork surface roughness (STEP 4b) + share the cellular noise
Fr0zkaandClaude Opus 52026-07-28 03:18:21 +02:00
7a87cdda14
docs: handoff — take the unattended queue all the way to 8 of 8
Fr0zkaandClaude Opus 52026-07-28 03:05:33 +02:00
473bccb43f
docs: handoff — add an unattended-run work queue for stage B
Fr0zkaandClaude Opus 52026-07-28 03:02:43 +02:00
b888b86729
docs: stage A verified with real coverage; rewrite the handoff
Fr0zkaandClaude Opus 52026-07-28 02:58:33 +02:00
68e43238bd
test: pits and chimneys never ran — ask the bake, not the density
Fr0zkaandClaude Opus 52026-07-27 19:22:07 +02:00
5bf61c1815
test: stage A is bit-identical — strengthen the test its own counters indicted
Fr0zkaandClaude Opus 52026-07-27 18:41:56 +02:00
ef5bda3d8a
feat: TunnelNetwork stage A — the SDF spine, wrapping BuildChunkCache
Fr0zkaandClaude Opus 52026-07-27 18:28:06 +02:00
96e75abe57
feat: port FloatingIslands — the stack that runs backwards
Fr0zkaandClaude Opus 52026-07-27 17:50:36 +02:00
f5d5a03ad1
docs: end-of-day handoff — 5 of 8 ported, 11 tests green
Fr0zkaandClaude Opus 52026-07-27 17:17:35 +02:00
d0a9ce3018
fix: the column key omitted the params — my perf fix broke the overhang
Fr0zkaandClaude Opus 52026-07-27 17:13:50 +02:00
3acb3fbc6b
feat: port VerticalShafts — three ops reused from Maze unchanged
Fr0zkaandClaude Opus 52026-07-27 17:11:07 +02:00
6a60732c98
build: FPSemantics = Precise + clear the IWYU debt that blocked it
Fr0zkaandClaude Opus 52026-07-27 15:53:23 +02:00
f1fd1e0b05
docs: new acceptance bar (peer agreement, not fidelity) — and C1's documented fix is wrong
Fr0zkaandClaude Opus 52026-07-27 15:43:23 +02:00
51d8db842b
test: fix the ULP yardstick — it measured output density, not where the error is born
Fr0zkaandClaude Opus 52026-07-27 15:39:03 +02:00
85993199fb
test: the slab test proved less than it claimed — add the pass that varies CeilingRoughness
Fr0zkaandClaude Opus 52026-07-27 15:34:16 +02:00
644339def5
feat: Phase 2 first port — FlatPlain + CrystalChamber collapse into one op
Fr0zkaandClaude Opus 52026-07-27 15:30:26 +02:00
974e795b66
fix: call PrepareChunk and guard the degenerate strate on the wired path
Fr0zkaandClaude Opus 52026-07-27 15:15:59 +02:00
f7cccb044b
docs: add OPSTACK-HANDOFF.md; install the 39 Unreal skills
Fr0zkaandClaude Opus 52026-07-27 15:11:26 +02:00
256a262140
feat: Phase 1 step 3 — wire the operator stack into GetDensityAt behind an opt-in
Fr0zkaandClaude Opus 52026-07-27 15:03:00 +02:00
34f06dfea7
docs: park the ULP residue as AUDIT C10; strip the diagnostic scaffolding
Fr0zkaandClaude Opus 52026-07-27 15:00:46 +02:00
b6ccb1c9f1
test: constness isn't it either — add the one unambiguous check, then stop chasing
Fr0zkaandClaude Opus 52026-07-27 14:57:21 +02:00
7189d51b7b
test: the variable is compile-time-constant vs runtime Blend — one line to confirm
Fr0zkaandClaude Opus 52026-07-27 14:53:02 +02:00
34f8ca7953
test: localised to FSdfCarveOp — now testing inlining context, the right variable
Fr0zkaandClaude Opus 52026-07-27 14:48:56 +02:00
210602586e
test: three-way says the fault is mine — compare the SDF channels directly
Fr0zkaandClaude Opus 52026-07-27 14:44:11 +02:00
d5c71d6b68
build: revert FPSemantics (drops the shared PCH); answer the FP question in the test instead
Fr0zkaandClaude Opus 52026-07-27 14:39:09 +02:00
a49d440bf6
build: put FPSemantics=Precise on the VoxelForge module — the experiment never ran
Fr0zkaandClaude Opus 52026-07-27 14:34:50 +02:00
cca9e83182
test: /fp:precise reproduced the residue byte for byte — instrument instead of guessing
Fr0zkaandClaude Opus 52026-07-27 14:32:22 +02:00
0379d59c1c
docs: AUDIT C9 — the FP default differs BY TOOLCHAIN, not just "fast is allowed to drift"
Fr0zkaandClaude Opus 52026-07-27 14:19:42 +02:00
af5f2103b3
test: the Maze residue is /fp:fast, not port drift — encode the real bar
Fr0zkaandClaude Opus 52026-07-27 14:15:43 +02:00
23605d5350
test: bisect the residual Maze difference instead of guessing at it again
Fr0zkaandClaude Opus 52026-07-27 14:07:14 +02:00
Fr0zkaandClaude Opus 5