acdream/tests/AcDream.App.Tests
Erik 4ad6fb9184 close #127 (user-gated + desk pin): distant-building flood flap died with the W=0 clip port
User re-gate 2026-06-12: ran past distant buildings, 'Seems to have
been fixed' - no flicker/vanish. The per-building flood-admission
bistability (#127, the building-flap mechanism behind the tower roof
flap and #123 'buildings vanish when running past') is gone.

Root: the bistable knife-edge admission died with the W=0
polyClipFinish clip port (987313a - the #119/#120 work that 'kills the
knife-edge class everywhere') plus the #120 containment-rejection
growth fix. The captured-pair evidence (tower-viewer-capture.log,
2026-06-11) PRE-dates all of those - it was that same near-eye knife
edge, not a separate distant mechanism.

Desk confirmation (both green at HEAD):
- CapturedFlipPair_AdmissionIsStable: the original 4 cm flip pair is
  now |A|=|B| with zero diff across all FOVs and both pre-gate states.
- DistantBuildingStrafe_NoAdmissionChurn (new regression pin): 0
  admission churn across all 21 building groups x {10,30,60,120,190} m
  x 100 mm-step run-past strafes, both pre-gate states. A stable flood
  toggles each cell at most once over a monotone eye path; this asserts
  no cell toggles >=2x.

ISSUES.md #127 -> CLOSED with the DO-NOT-RETRY note (no re-opening the
BuildFromExterior seed gates for a flap symptom without a fresh HEAD
repro - the captured-pair lead is dead). Render digest banner updated.

Suites: App 264+1skip / Core 1443+2skip / UI 420 / Net 294.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-06-13 10:03:33 +02:00
..
Rendering close #127 (user-gated + desk pin): distant-building flood flap died with the W=0 clip port 2026-06-13 10:03:33 +02:00
AcDream.App.Tests.csproj refactor(app): extract typed RuntimeOptions for startup env vars (Step 1) 2026-05-17 09:16:55 +02:00
RuntimeOptionsTests.cs chore(render): Phase A8.F — strip ACDREAM_A8_DIAG_* step-disable flags (keep PROBE_VIS) 2026-05-29 11:25:00 +02:00