# iOS V0 acceptance — provisional, hardware gates pending

Date: 2026-10-01. Original brief: `docs/requirements/ios-v0-brief.txt`; contract: `docs/validation/verification-matrix.md`. **The V0 definition of done has not yet been demonstrated.** Synthetic fixtures and SDK compilation cannot answer the research question about torso-mounted iPhone measurements.

The app has been prepared for physical preview installation, but launch was awaiting device developer trust at the time of this preliminary report. The coordinator inspected its signing profile: created 2026-10-01, expires 2026-10-08 15:46:09 UTC. Rebuild/reinstall through Xcode to refresh signing after expiry. No actual motion session, mounting/sign check, approximately 50-transition dataset, independently labeled movement count, observed missed/false-positive count, onset error, or hardware range/variance/smoothness conclusion is available here. Device OS/build and exported artifact identifiers must be recorded from the actual trial, not inferred from simulator or SDK versions.

## Requirement evidence register

| Requirements | Software evidence sources | Acceptance status / outstanding evidence |
|---|---|---|
| R1–R5 acquisition/statistics/clocks | U1/U2 reports and independent reviews; capture/statistics/schema tests; new U8 exact artifact test | Device delivered rates, clock mapping, raw/fused impact and three matched stationary/movement trials per mode pending |
| R6–R7 neutral/rider coordinates | U3 reports/review/retry evidence; synthetic known-axis math and calibration integration | Physical oblique mount, all ±axis signs, repeat neutral and mounting rigidity pending |
| R8–R11 recording/events/cues/durability | U4/U5 reports/reviews; persisted session, lifecycle, cue and UI tests; U8 fault tests added | Authoritative new tests pending; hardware controls/markers/audibility/interruption/relaunch/share pending |
| R12–R13 schema/export | U1/U4 evidence, schema/CSV/archive tests; U8 independent ZIP/JSON/CSV artifact assertions added | Device metadata agreement and external Files/AirDrop reopen pending |
| R14–R18 analysis/results | U6 evidence/review; numeric/phase/smoothness fixtures; U7 integration underway; repeated analysis/input identity assertions | Final integration/review pending; independently recomputed real metrics and labeled detector assessment pending |
| R19 diagnostics | Portable diagnostics tests; U7 mode/review in progress | Approximately 30 s stationary and controlled slow/rapid movement capture pending |
| R20 Watch portability | Saved synthetic Watch pipeline test passed; full core watchOS 10 SDK typecheck passed (`u8-evidence.md`) | No Watch hardware/runtime evidence; adding a Watch target is the next experiment |
| R21 comparison evidence | Separate sensor/host times, clock mappings, receipt/cue dispatch times, rates/jitter, raw measurements, calibration and analysis configurations preserved | Real noise/drift/sensitivity/onset/peak/range/smoothness/repeatability comparisons pending; physical sensor/acoustic latency and cross-device synchronization unmeasured |
| R22 iPhone 16 Pro acceptance | Automatable persistence/analysis/export workflow test added | Developer trust/launch, torso-mounted garage protocol, roughly 50 movements, independent labels, observed research report and external export pending |

## Research artifact accounting to complete after capture

Record device/OS/app/build, protocol/mount, master UUID/path and SHA256, accepted/written/lost rows by stream, start/end clock anchors, achieved rate/interval/gap statistics, calibration identity/configuration, analysis version/configuration/hash and exported filenames. Preserve stationary and movement captures even if the conclusion is inconclusive.

For every deliberate movement, retain an independent label/time interval and pair it to detected transitions or mark it missed/uncertain; retain unmatched detections as false-positive candidates. Report observed versus detected counts, onset error with clock uncertainty, range/duration/variance by side and repeat block, and candidate metrics for simple and intentional-correction subsets. Marker/cue timestamps alone are not movement ground truth. No composite rider-quality score is an acceptance output.

Physical receipt age describes callback delivery after an explicitly validated mapping; it does not measure sensor/fusion physical latency. Audio dispatch is not acoustic onset. iPhone and Watch boot clocks are independent; simultaneous recording alone does not synchronize them. Matched torso locations, mount differences and future Watch screen-off/off-wrist/runtime behavior require explicit follow-up.

Final authoritative automated results and actual hardware artifacts should be appended to this report by the coordinator. This preliminary register intentionally makes no pass claim for unexecuted checks.

## Desktop integration update

The authoritative core (41) and iOS integration (29) tests passed, including exact persistence/export and lifecycle failure recovery. Source review completed without outstanding findings; optional cross-model review was unavailable. Review and diagnostics software is integrated. Build 0.1 (2) installed successfully and was confirmed in the iPhone inventory; launch was blocked by the locked phone. The initial pending software status in the register above is superseded by `calibration-desktop-update.md`. All stated physical acceptance gates remain pending.