30 Hz out-of-band energy 15.45% -> 0.00%; the 29 Hz rate-2.0 detune -133 -> +0 cents.
An unknown period keeps the fixed-window geometry bit for bit. The detector cannot
reach process(): sampler_core does not link it.
Retire the effectiveGain blend so the meter's minimum tracks smoothGain's own
reduction against real input, unscaled by the transition mute — a toggle over
quiet material now reads no reduction instead of pinning to 0.
Fold setLimiterEnabled's restart request into setInstrumentParams so every writer
keeps the host's latency report in sync. Pin the window-sizing identity, drop the
per-sample modulo, tighten the ceiling tolerance, publish the blended gain.
Zero-crossing counting was anti-correlated with the real defect (splice debris
fools it). Now asserts energy outside the fundamental, with an alignable control,
matching test_preserve_low_frequency.cpp's approach.
New pure curve_tessellate joins the overlay's node vertices through curveMap,
one sample per pixel column; the knot no longer floats off its own trace.
A splice can only relocate by [0.75, 1.25]*window, so periods with no multiple
in that interval never phase-align — at 50 ms, f < 16 Hz and 26.7-32 Hz.
Harness runs by hand; too slow to gate.
Generalize curveMidLevel/curveFromMidLevel to curveLevelAt/curveFromLevelAt at
arbitrary phi; knotVtx and its drag inverse now read the phi a knot's truncated
x actually implies, not always 0.5.
Re-derives the splice-cadence inequality and adds a corner probe that FAILS at
P=500 by design, pending a ruling. Names the baseline commit and harness edit,
fixes measurement methodology, corrects the Trigger-AHD/rate coupling framing.
Swap nearbyint for std::round (MXCSR-independent); derive the finest-drag test
from the editor floor, not the knob; verify resets against fields, not norms;
record the spline-point modifier exclusion.
Generalizes the correlation-aligned SOLA delay line so the feed and the shift are
independent rates over one ring. Unity is bit-identical to the shipped read, asserted
against a hash baseline captured pre-change.
MSVC's C4062 is off by default and GCC/Clang's -Wswitch only warns without
-Werror; this repo sets no -Wall/-Werror anywhere. /we4062 and -Werror=switch
now cover both, scoped to pure libraries only.
REAPER's GetUserFileName serves both pickers — the no-save-picker finding was a
regex miss. A new enum value is structural, not additive: BankModel rejects an
out-of-range SourceMode/Tier rather than degrading.
The dedup path claimed it directly, so a key could be answered Ok pointing at an entry a failed persist never wrote. Write-back is now guarded and three-valued, and landing one request split out of the scan pass.
SetProjExtState's return covers the whole extname, so it never saw one key. The persist verdict now reaches the report, and a throw mid-write no longer claims the landing left nothing behind.
Splits Ignore into unreadable vs not-a-request and counts every verdict; the report prints only when the pass answered nobody, so its absence proves the action never ran.
Batch captures now name the retained-render folder once instead of nothing; Auto/Manual tail modes refuse a 0-byte render like None does; VERIFICATION.md steps 1-3 no longer invite a false conclusion; docs/comments no longer overclaim.
Adds the Phase Λ entry: six settled rulings, four open forks, and thirteen tracks across seven waves, from docs/product/linux-readiness.md. Adds the l slug transliteration, and scopes the plan-wide "nothing unanswered" claim, which Λ's open forks break.
Refuse multi-track; result track always Arrange; follow panel tail. Rho-F2 overrides the mode-following framing: explicit Arrange tags on track and item, plus an explicit-tag-wins filter in detectNewContent.
Product doc for the extension + ReaSampler 9000 on Linux, downstream of the two Λ-W1 audits: Daniel's six rulings, seven waves, the non-REAPER-host safety contract, and four open forks.
The one-frame bound is not provable for a per-edge renderer; the test now shows where it breaks. Refused renders move out of the bank instead of being deleted, so the DAW experiment has something to read.
Ε-F1 proprietary RSBK, Ε-F2 new bank with auto-suffix, Ε-F3 refuse on a
degraded ledger. No Ε track is gated now. Also carries Phase Ρ's plan
sections, authored concurrently in this shared checkout.
Product notes for rendering a track's output to a new sibling track at the exact render position, source parked in Design mode, bank never touched. Framed as a third verb (arrange->arrange), not an exception to capture/placement separation. Three open forks.
Rewords Cleared/Unanswered/ForeignRequest to name only observed facts and every live hypothesis instead of picking a winner; adds a guarded answeredOutcome accessor against a future unhandled BakeAnswerKind; folds an unreachable BakeScanKey/Context test state; fixes generation-stamp timing.
Corrects a false unguarded-include claim, re-scopes hot-path evidence to what's actually linked, replaces a nonexistent Fork-A collision scenario, fixes several miscounted TUs/libraries, re-grades two Fork-B items from Blocker to Major/non-blocker, and adds missing owners/done-conditions.
A no-answer stays a failure; it now says whether the extension never ran the landing, answered a stale generation, spoke a wire this build cannot read, cleared the request, or refused it.
The tolerance is unchanged and now derived, not assumed: frameCountFor lands in {floor(L), ceil(L)}, so a non-frame-aligned window can never miss by more than a frame. Naming the source is what tells a self-bounding render from a short one.
Two one-line compile blockers, a locale hazard in the persisted number codec, and a cost split showing the extension is nearly there while the instrument editor is a from-scratch X11 job.
daniel