Measured on real hardware: IO1=right IO2=left IO3=top IO4=middle IO5=bottom
(0x01=right 0x02=left 0x04=top 0x08=middle 0x10=bottom). The earlier verbal
'left=IO1/upper=IO2...' spec was wrong; this reverts to the original Phase-1
mapping, which the measurement confirms.
- protocol.js REGION constants + getRegionName + REGION_NAMES
- CRITICAL: buildVendorCommand ioMasks was built from REGION.* constants whose
values just flipped, scrambling IO slot order; pin to literal [0x01..0x10]
- diagnosis.js REGION_DEFS key↔mask; pd_region_map values unchanged (PD is
keyed by orientation, which is stable) — right=PD5 left=PD1 top=PD2
middle=PD3,6,7 bottom=PD4
- common.js region labels back to right/left/top/middle/bottom (上/中/下)
- treatment-setup / manual-treatment / treating / scan-report / ble-debug
face-map bindings and REGION_MAP realigned
- verified: mock pipeline lands left→IO2, middle→IO4 correctly
- Remove parseVendorStatus 17-byte branch (FFE4 is now 19 bytes)
- Remove parseVendorStatus 33-byte branch (FFE1 is now 34 bytes)
- Remove buildVendorCommand33 deprecated alias
- Remove fitting/run_state !== -1 guards in connection.js
- Update commands.js to use buildVendorCommand directly
- parseVendorStatus: add checksum_ok verification for 34-byte and 33-byte params
- command-sync.js: passthrough control, current_gain, hold_time from server payload
- Mark buildVendorCommand33 as deprecated alias
Per 协议简述(3): FFE4 now 19 bytes — adds Byte17 fitting status
(0=not fitted, 1=fitted) and Byte18 run state (0=IDLE, 1=scanning,
2=treating, FF=fault). Maintains backward compat with 17-byte format.
Region name mapping corrected to match actual hardware IO positions:
- 0x01: 左区→右区, 0x02: 右区→左区
- 0x08: 下区→中区, 0x10: 中区→下区
Auto-scan no longer shows timeout error, proceeds to next step instead.
Fixes timeout during binding when targetDeviceId is set but exact
string match fails due to encoding differences. Also skips devices
with no name, and logs match result for every candidate.
- parseDeviceId accepts hex with spaces (e.g. "672B6D5A 4DCD861")
- BLE scan does exact match on "JW_<device_id>" when target is known
- Manual input placeholder updated to show new format
Replace '光子美容仪' with '智美'/'智能面膜'/'我的设备' throughout.
Replace '护理' with '使用' in all UI text.
Replace '皮肤' with '检测'/'智能' where applicable.
Replace 🌸 with 💎 for brand identity.
Internal code (variable names, API paths, file names) unchanged.
- ADC (17-byte) now emits 'adc' + 'battery' events, not 'status'
- Fixes CRITICAL timer jump: remaining_ms:0 was resetting countdown
- Drop ADC packets with bad XOR checksum
- treating.js + index.js subscribe to 'battery' for live battery updates
- onStatus only updates remaining_ms when > 0
Protocol update adds FFE4 Status format: 7×PD sensors (little-endian)
+ VBAT battery voltage (mV) + XOR checksum = 17 bytes.
- parseVendorStatus: handles 1-byte heartbeat, 17-byte ADC, 33-byte params
- handleValueChange: ADC type extracts PD array + battery % from VBAT
- Battery calculated as linear 3000mV(0%) to 4200mV(100%)
- PD data passed through status event for treatment pages
- Updated protocol doc with full byte table and old/new comparison
Miniprogram:
- Add BLE reconnect button on home page when disconnected
- Add loading states to index, profile, subscribe-plans pages
- Add profile editing (avatar + nickname) with COS upload
- Enable pull-to-refresh on history page
- Fix auto-scan self.options → self.data inconsistency
- Add console.error to silent catch blocks
- Gate 'Simple Peripheral' BLE scan behind __DEV__ flag
- Add production config comment to env.js
Admin console:
- Add empty state '暂无数据' to all 5 list views
- Replace plain text plan input with select dropdown
- Add type="date" to record date filters
- Add production config comment
Server:
- CORS origin restricted in production (env CORS_ORIGIN)
- DB pool size configurable via DB_POOL_SIZE env var
- .env.example updated with WeChat Pay + production fields
Heartbeat (1-byte FFE4 notify) now emits 'heartbeat' event only.
Only 33-byte state responses emit 'status' with parsed IO data.
Fixes timer jumping to 0 every second during treatment.
- handleValueChange now calls parseVendorStatus and emits 'status'
with legacy-compatible shape (mode_state, battery, remaining_ms, etc.)
- 1-byte heartbeat mapped to mode_state, 33-byte state parsed for
active IO detection
- Fixes regression where treating, index, wear-check, auto-scan pages
never received status updates in vendor_33 mode
- Also emits 'vendor_raw' with full parsed data for future use
- discoverChars uses properties (write/notify) to distinguish FFE1
command characteristic from FFE1 service UUID
- handleValueChange: vendor_33 mode emits vendor_status (FFE4 heartbeat)
and vendor_data separately, no longer tries parseFrame on raw bytes
- parseVendorStatus handles 1-byte heartbeat and 33-byte state readback
with proper IO1-IO5 field parsing
- Add expire_time > NOW() filter to findActive() preventing stale subscriptions
- Add express-rate-limit on login endpoints (user: 10/15min, admin: 5/15min)
- Add production guard for default admin credentials
- Fix BLE bindDevice userId encoding (uint32 instead of hexToBytes on numeric)
- Wrap adminCreate in transaction to prevent race condition
- Add settings cache invalidation after admin saves
- Read trial_days from settings instead of hardcoding 7
- Fix double JSON.stringify in commandDao.finish call
- Cancel stale pending bindings before creating new ones
- Reduce token refresh grace period from 3 days to 1 day
- Fix subscribe-success to fetch expiry from server (correct for renewals)
- Add keep-alive name property to DashboardView and SettingsView
- Fix BLE disconnect() to preserve listener registrations across reconnects
Guoguo