Protocol parity with the JRing app fixes - #77

Summary

PulseLoop's 56ff driver was reverse-engineered from partial knowledge. Decompiling the vendor JRing Android app (SDK com.sxr.sdk.ble.keepfit, app com.jaga.ibraceletplus.jyring) let us diff our driver against ground truth. Three long-standing user reports all trace to verified divergences, and the diff surfaced two outright bugs.

What changed

• JringClock (new): latches the UTC offset when we send 0x01, subtracts it at exactly the four ring-stamped decode sites (0x01 ack, 0x03 activity, 0x11 sleep, 0x16/0xA0 history HR). Deliberately not applied to 0x24/0x14/0x27/0x28/0x3f, which are stamped on arrival: a blanket post-decode shift would corrupt live samples.
• makeTimeSyncCommand now sends local wall-clock seconds. Byte [5] keeps the DST-aware hour (vendor sends the raw one); half-hour zones truncate there exactly as the vendor's do, harmlessly, since bytes [1..4] carry the full offset.
• resyncTime() on RingSyncEngine (default no-op) + a coordinator observer for NSSystemTimeZoneDidChange / significantTimeChange. The vendor has no such listener; this is strictly better.
• runStartup() now emits 0x19 (background monitoring) and 0x20 (capability bitmask): 48 → 02 → [33] → 0c → 01 → 21 → 19 → 20 → 10 → 16.
• Measurement frequency for 56ff: .measurementInterval on JringCoordinator — the only UI change needed, since SettingsView/MeasurementSettingsView already gate on it — plus set/applyMeasurementSettings on JringSyncEngine mapping hrIntervalMinutes → 0x19[6]. stopHeartRate() now restores the user's cadence instead of a hardcoded 30.
• History dedup: persistMeasurement upserts source == .history on (kind, timestamp), two-tier (in-process guard + one indexed fetch for cross-launch). Live samples keep append semantics.

Related issues

This should fix #8 and #14

Type of change

• 🐛 Bug fix (non-breaking change that fixes an issue)
• ✨ New feature (non-breaking change that adds functionality)
• 📟 New / improved wearable support (BLE driver layer)
• 🤖 Coach / LLM change (tools, prompts, orchestration)
• 🎨 UI / DesignSystem change
• 🧹 Refactor / chore (no behavior change)
• 📝 Docs only
• ⚠️ Breaking change (existing data, settings, or APIs change)

How was this tested?

Need testing with 56ff ring.

• Added / updated unit tests (PulseLoopTests)
• Ran the test suite locally (⌘U in Xcode)
• Tested on a physical device with a real ring 56ff.
• Tested with demo data (-seedDemo YES, no hardware)
• N/A (docs / non-code change)

Privacy & data

• This change does not send health data off-device without explicit user action.
• No secrets, API keys, or personal data are committed.
• N/A

Checklist

• My code follows the project's style (SwiftLint passes).
• I ran the tests and they pass.
• I updated docs / README where relevant.
• I read the Contributing guide.