Skip to content

RELEASES.md

Latest commit

 

History

History
86 lines (62 loc) · 3.74 KB

RELEASES.md

File metadata and controls

86 lines (62 loc) · 3.74 KB
Raw

FaceLock Releases & Version History

All official releases, binary assets, and detailed update breakdowns for FaceLock are documented here.

Official pre-compiled application archives are automatically published to GitHub Releases on every version tag.

[v1.0.0] — Initial Production Release (2026-10-20)

Headline: Initial stable release of FaceLock — High-performance native biometric shield and intruder defense system for macOS Apple Silicon.

📦 Binary Assets

  • Download: FaceLock-macOS-arm64.zip
  • Architecture: Apple Silicon (arm64 — M1, M2, M3, M4 series)
  • Minimum OS: macOS 14.0 (Sonoma) or macOS 15.0+ (Sequoia)
  • Dependencies: Pure native AppKit & Vision (Zero external libraries)

🌟 Release Highlights

1. Deep Neural Biometrics (Apple Vision ANFRv4)

  • Leverages Apple's native deep faceprint neural network (VNCreateFaceprintRequest).
  • Extracts 128-dimensional L2-normalized biometric vectors running directly on the Apple Neural Engine (ANE).
  • Dual-Verification Cluster Metric:
    • Requires candidate face to satisfy both $d_{\min} \le 0.22$ and $\text{top3_avg} \le 0.27$.
    • Completely eliminates false acceptances on strangers while maintaining instant verification for the owner.

2. Unbypassable Security Shield (.screenSaver Level)

  • Enforces a fullscreen dark glassmorphic security overlay at CGWindowLevelForKey(.screenSaverWindow).
  • Blocks hotkey escapes (Cmd+Q, Cmd+W, Cmd+M, Cmd+H, Esc).
  • Audio and tactile feedback on authentication attempts (NSSound.beep() on error, Glass click on unlock).
  • Biometric Touch ID or master PIN fallback.

3. Low-Light Intruder Snapshot & Night-Mode Lift

  • Automatic face cropping with generous 80% shoulder headroom.
  • CoreImage hardware pipeline applying real-time exposure lift (+1.3 EV), shadow recovery, and tone curve equalization.
  • Intruder faces remain crisp and identifiable even in dark rooms.

4. Instant Telegram Bot Alerts

  • Multi-part HTTPS multipart/form-data upload of intruder photographs directly to your personal Telegram chat.
  • Includes formatted HTML caption with timestamp and targeted application name.

5. Walk-Away Inactivity Lock

  • Monitors user input inactivity via CGEventSource.
  • After 45 seconds of no keyboard/mouse events, FaceLock silently samples the camera. If you stepped away, the Mac locks immediately.

6. Universal & Selective Vault Modes

  • All Applications: Continuous biometric protection on every window focus switch.
  • Selective Mode: Guards only confidential apps (Telegram, Safari, Chrome, Terminal, Notes, Photos, WhatsApp, etc.).

7. First-Class CLI & Automation

  • Interactive flags: --enroll, --settings, --gallery, --test-lock, --status, --update, --version, --help.
  • Automated sync engine (make commit, make update, scripts/sync-daemon.sh).

How to Update

Method 1: Built-in Self-Updater (Recommended)

facelock --update

Method 2: Via Git & Make

cd FaceLock
make update

Method 3: Download Binary Release

  1. Download FaceLock-macOS-arm64.zip from Releases.
  2. Unzip and drag FaceLock.app into your /Applications/ folder.

Release Process for Maintainers

To publish a new official release:

# Example: cut release v1.1.0
make release VERSION=1.1.0 MSG="Feature release: proximity unlock & stats"

This automatically tags the commit, updates the changelog, and triggers GitHub Actions to build and publish the .zip binary archive to GitHub Releases.