The previous commit (8bfce189) does not build: CS0104 ambiguous Random in ThrustAllocatorTests.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Nothing asked, nothing fired is a fact about the intent: when all three clamped intents are exactly zero Allocate returns the zeroed buffer before the solve, so a warm solve that stops at the iteration cap can no longer fire thrusters. The output flush is gone. The warm start is left at the last solved point. Tests: the two 40-thruster hulls Soul found, and a small intent still answered at any column scale. Builds untested locally; verified on Yggdrasil.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Off-centre joint limits (20..50, -10..50, -50..0) so Max-Min is distinguishable from |Min|+|Max| and 2*max/2*Max; Bind walking past tagged
nodes that are not joints; Rigs and the lead mount follow hull order where it differs from name order. Tests only; builds.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Flushing an empty cache writes no file, so the previous fixture still hit the new missing-file refusal.
Builds.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
After a good call the warm start left 1e-34..1e-44 throttles on a zero-intent call. Throttles at or below ThrottleResidue return 0 on the one general path; TA-R3's warm start is kept. The test NothingAskedAfterACallFiresNothing arrived in the previous commit (22cb76e8), which therefore fails that test; this commit makes it pass.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A cost row pulling the errors the other way lowers the pre-onset drive throttle by about 20 percent (1.30e-4 to 1.06e-4 at turn .5); the bounds sit 8 percent under the measured value.
Commit builds: yes (tests/Aetheria.Shared.Tests not yet run on this commit; the full run is the closing job).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The turn leaves a forward error of 2, the top of TA-R2's error box; a bound of 1 or 1.5 gives yaw back.
Commit builds: yes (tests/Aetheria.Shared.Tests not yet run on this commit; the full run is the closing job).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
128 torque-free forward thrusters from a cold start reach the 100-iteration cap through the public surface; the call must keep the iterate it reached, not zero it.
Commit builds: yes (tests/Aetheria.Shared.Tests not yet run on this commit; the full run is the closing job).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Imagination pass (session self-2026-10-10-ag): the row gimbal is sim data,
one shared-throttle actuator per row in the allocator's single solve;
specs vector-rows-record r2 and vector-rows-allocator r1.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The game treats a directory without ship.cc as not a package (TheBootCatalogIsTheShippedOneUntilAModIsInstalled), so da4d8c6c's
Compose exclusion broke that test. Reverts the Compose and ResolveCatalog change and moves the check to the one place that
holds the first-party tree to a stricter rule: TestCatalog.Resolve throws, naming each directory under GameData/Mods that
holds no ship.cc (missing-ship-cc-skipped). The GLB case change and the registry parameters stand.
Builds; not yet run.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The test assembly registers a catalog [CultGlobal], so the global check already threw on a missing
file there and AStoreRefusesAMissingCatalog passed against the old Open (run at fe9da2e: 12 passed).
The refusal is now a FileNotFoundException naming the path, and the test pins that type, which the
global check never throws. Builds.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The test relied on a read-only open of a nonexistent file yielding an empty catalog, which
AetheriaStores.Open now refuses; the fixture writes the empty catalog first. Builds; the suite is green with it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The previous commit did not compile in Unity (CS0117 on PlayerSettings.bundleVersion); this one does,
and a batchmode run of AetheriaBuild.Build succeeded.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
AetheriaBuild (Assets/Editor) sets the version, builds Addressables, builds the
Win64 player, stages GameData through PlayerStaging, publishes AetherDb to
ModTools and removes the Burst debug folder; any failure exits 1 naming the step.
PlayerStaging is Unity-free and linked into Aetheria.Shared.Tests. A read-only
AetheriaStores.Open of a nonexistent catalog throws.
Builds on the headless side; the Unity .meta files and the batchmode run follow.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Deletes the direct GameData/Aetheria.cc opens (FireControlCut124Tests x3, FireControlCut7Tests, MiningCut3Tests x2,
RestoredHullsTests x11) in favour of TestCatalog.Repo, and the eight copies of the shipped-assembly registry in favour of
TestCatalog.Registry() (real-catalog-readers-bypass). Pirates and Roles command tests keep copying the shipped catalog (they
rewrite a copy through AetherDb) and take only the registry.
Builds; not yet run.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
AetheriaStores.Open, ShipModCatalog.Compose and ResolveCatalog take an optional registry (the game passes none and keeps the
process-wide one). TestCatalog.Registry() is the shipped assembly's registry, so the test assembly's TestCatalogGlobal is no
longer a catalog global the real catalog must hold (fixture-fails-on-real-package). TestCatalog.ForRepo(root) is the one
binding Repo uses, pinned by a test over a real catalog plus a package (repo-wiring-unpinned). A package directory without
ship.cc is excluded with a reason, in Compose and so in ResolveCatalog's pre-check (missing-ship-cc-skipped). The loader takes
only a lowercase .glb, matching the case-sensitive LFS rule (lfs-glb-case).
All commits build; not yet run (verification job follows).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Serves follow-ups vectored-thruster-rows and gimballed-thrusters: virtual
actuators, nonlinear and cascaded allocation, rate and gimbal limits as
per-tick bounds, the zero-thrust singularity, games, shared-angle rows.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Behaviors.cs union list: keep Union(41, ThrottleLockData) from master and
Union(42, VapourDumpData) from this cut; no key renumbered. All commits build
as a merge result only after the verify job.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
The first commit's gear modifiers never executed (test gear has no heat response, so it read offline). Builds; the first commit's new tests fail without this.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Pins ruling boost-stacking-multiply through the production path (gear and consumable StatModifierData on a ship, read via EquippedItem.Evaluate): TwoScalesOnOneStatMultiply, AConsumableBoostMultipliesWithGear, TwoConstantsOnOneStatAdd. Tests only; no production change. Not yet run; the commit is unverified.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Replaces TheLockHoldsThroughMostOfItsDurationAndEndsAfter with a theory over Duration 1 and 2 at 0.25 s ticks. Builds.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Deletes the articulation anchor role. Adds ShipJoint (key 5 Joints), ShipAnchor.Joint (key 5),
the judge's joint-shape, mount-on-joint and yaw-reach checks over RigChains, Bind's GLB checks
(ReadNodeIds returns the nearest tagged ancestor), and ShipModPlan.Rigs. Presentation only:
arc stays HardpointData.FiringArc and traverse WeaponData.Tracking (ruling mount-arc-gun-traverse).
Not yet built or run when committed; the verification commit follows.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Imagination r2 re-map (session self-2026-10-10-ag): the sim already carries
per-hardpoint arc and weapon traverse; links are rig only. Target doc moves
articulated weapon mounts into scope under ruling mount-arc-gun-traverse.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Every entry of A is now assigned each call or never written, so the clear was a second
owner of the same state (hand mutants a-never-cleared and a-cleared-by-division survived
it as equivalent).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Stryker survivors 12390 (initial _columns +1 skips the first resize at n = 1) and
12399/12400 (A not cleared between same-count calls) get tests. Array.Clear on b was an
equivalent mutant: rows 0-6 are assigned every call and the ridge rows are never written.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Eyes pass before the hull shader fork: DECALmachine 2.17 export paths,
forward-renderer decal practice, parallax for panel lines, layered wear
on the built-in pipeline, and this project's rendering constraints.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Removes the scratch probe. The spec pinned 1e-4; the optimum leaks 1.2e-4 (Duel) and
1.6e-4 (halved Talaria) from the first touch of Turn. Builds; ran 18 tests.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Rules TA-R1 to TA-R4 of cut-thrust-allocator-core.r2, tests on synthetic columns
(Duel, damaged Duel, symmetric hull). First draft; not yet run.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ambientCG 2K sets under Assets/Materials/PBR (GlowFade .mat per material,
packed metallic/smoothness), and Asset Sources/PBR Materials.blend as a
UV-mapped Blender asset library linking the same textures. Sources and
licence in Assets/Materials/PBR/SOURCES.md. Unity import unverified until
the project is opened.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
12 asset-marked materials, UV mapping with a scalar Tiling node, images
linked by relative path to Assets/Materials/PBR, previews generated.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2K-JPG Color/Roughness/NormalGL (+Metalness, AO where the set has them),
generated MetallicSmoothness.png (R metallic, A 1-roughness), texture and
material .meta files, SOURCES.md. Unity side not opened yet.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
CultMathRevision 252e226c (cultmath-unity-v0.4.0), manifest tag and lock entry.
Generated content that samples noise may move; fixtures follow in later commits.
Builds as a pin; the suite may fail on re-rolled content until then.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>