LlamaCasty

Files

T

gramps fbc8562cf4 perf(capture): slice 8 — buffer ring + paste-cache Epoch; gen2 visibility (take-11 spikes)

Take 11 (c10ce06c) validated the off-UI architecture: typical frames land
work ~10ms + wait ~6.8ms = 16.7 exactly on the deadline; 212/300 best yet.
The ENTIRE remaining gap is periodic 35-65ms render spikes that WORSENED
across the take (189 -> 147) — the signature of gen2 GC pauses. Biggest
churner is structural: the screen capture minted a fresh ~8.3MB byte[] per
DWM frame (~500MB/s of LOH), a producer OBS never does (it owns fixed
surface pools).

- ScreenCaptureFrameSource: 4-deep buffer ring with size-matched slots (a
  <=17ms consumer cannot be lapped at 60Hz) + reused downscale row scratch.
- VideoFrame.Epoch: monotonic per producer frame. The paste cache keys on
  array IDENTITY, so recycled arrays MUST be distinguished — epoch joins the
  PasteKey. Producers handing fresh arrays leave it 0 (key unchanged effect).
- Stats print 'gen2 +N' per 5s window: next take acquits or convicts GC
  without another guess (rule: prove the stage).
- Test (the ONE): PasteCache_RecycledArrayWithNewEpoch_ReRasterizes_NotStaleHits
  — same array, new content, bumped epoch; fails on the old key by
  construction. 37/37 compositor/pump, clean build.
- Next suspect if gen2 stays hot: the 10Hz WebView2 capture (full-canvas PNG
  decode + fresh arrays on the UI thread) — recorded, untouched.

Creator audio ask queued in the same working session (+40% post-mix master
gain before the -1dBFS limiter) lands as its own commit next.

2026-09-04 12:36:14 -07:00

AboutHubTests.cs

Monetization policy locked + dev pause recorded (docs changeset, 2026-08-14): the one paid line becomes an annual subscription — early access $49.99/yr → $99/yr list at GA, grandfather-while-subscribed with lapse → list on renewal (no escalation matrix); free tier unchanged (branding flash is the free tier's billboard + no Alerts, which is the one paid feature); billing NOT locked to itch.io (hunted fact: itch.io has no native subscription billing — candidates Gumroad, whose native affiliate program is the tiebreaker, and Lemon Squeezy); support = in-app bug-report → git issues with a when-I-get-around-to-it cadence + emergency patches. ai.md Monetization section rewritten + stale paid-unlock seams fixed (L115 subscription-entitlement wording, the About hub's tabled billing PremiumUrl seam); TASKS.md gains TASK 23 (Monetization — billing provider decision, unlock mechanism with lapse→flash-returns enforcement, in-app bug-report loop, Unlock Premium seam lighting, Alerts gating, and line-item pointers for the early-access strategy, affiliate/ads phase 2, and the channel revival); HANDOFF.md records the decisions + the development pause (resume streaming = pair-programming dev streams via OBS + The Division 2 under the gramps/GOAT persona, 67 and gaming since 1981, 3D-printing content retired; dogfood switch to ytLlive once TASK 5 lands; resume point = TASK 5). Stale itch.io assertions corrected across comments and indexes (MainViewModel.cs PremiumUrl/IsPremium seams, MainWindow.xaml tooltip, AboutHubTests comment, ViewModels/index.md, TASK 9 out-of-scope note). Build 0 warnings, 196 tests passing

2026-08-14 18:48:42 -07:00

AudioMixerTests.cs

TASK 8 meter scaling: +10 dB input amplification in AudioLevelMeter.ToDisplay so the meters use the full bar — speech peaks (~0.2 RMS, -14 dBFS) read ~0.93 (red zone) and normal speech (~0.05, -26 dBFS) ~0.73 (yellow) at maxed volume, instead of hovering at the red edge; ≤0.001 linear still reads zero so the meter never idles on background noise; one knob shared by the mic bar and the game bar (× MicVolume gain-on-noise untouched); tests updated + new ToDisplay_Pushes_Speech_Peaks_Into_Red_At_Maxed_Volume — 171 tests passing, 0 warnings

2026-08-14 08:43:06 -07:00

AudioPipelineTests.cs

TASK 25: master limiter on the live mix (2026-08-15, queued from the TRAX discussion) — the review confirmed two of the creator's three instincts were already satisfied (MediaFoundationReader streams from disk so file size needs no guard; the MusicPlayer 0.20 bed rides the same loopback gain as the game, so music is always exactly 20% of the desktop volume by construction) and found one real gap: AudioMixer.FillAndMix summed mic + loopback with no output ceiling, so hot gains could pass 0 dBFS and clip the AAC encode. New pure Services/Audio/MasterLimiter.cs: -1 dBFS ceiling (Ceiling = 0.891), instant attack per frame (a hot frame is scaled exactly to the ceiling — no overshoot), smoothed release toward unity so loud passages don't pump, gain never exceeds 1. Applied at the end of FillAndMix right before the pipe write. 3 unit tests (trim-to-ceiling, never-boost, recovery) + ONE integration test (MasterLimiter_CapsTheLiveMix_OnThePipe: a 0.95 loopback bed capped to exactly 0.891 on the wire while staying audible). Docs in the same commit: ai.md go-live audio section, Services/index.md new row, TASKS.md TASK 25 record, HANDOFF.md shipped state. Build 0 warnings, 193 tests passing (189 + 4 new)

2026-08-15 16:10:16 -07:00

AudioSyncDelayTests.cs

TASK 22: global audio sync offset (positive delay at the mixer out)