Skip to content

CHANGELOG.md

Latest commit

 

History

History
307 lines (261 loc) · 12.9 KB

CHANGELOG.md

File metadata and controls

307 lines (261 loc) · 12.9 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[1.4.0] - 2026-06-02

Added

  • SPA health indicator in navbar with real-time backend status polling (green/yellow/red dot)
  • Service availability graph on home page showing 3-day uptime history from Prometheus
  • /api/v1/health-light/ lightweight health endpoint for SPA polling
  • /api/v1/availability/ endpoint returning historical availability data
  • CORS_ALLOW_PRIVATE_NETWORK setting for browser Private Network Access support (needed when SPA is Cloudflare-proxied but API resolves to private IP)
  • Canary cleanup service that removes failed canary images and reports impact in metrics
  • SPA_PORT environment variable for configurable SPA container port

Fixed

  • Health indicator permanently stuck on "Checking..." — HealthProvider now wraps entire body layout including navbar
  • CORS Private Network Access blocked when frontend is public (Cloudflare) but API resolves to private IP
  • APScheduler running under manage.py commands (now gated to only run under Gunicorn)
  • SPA container health check failing in Alpine-based images (127.0.0.1 instead of localhost)
  • CORS_ALLOWED_ORIGINS missing from docker-compose environment

Changed

  • Docker Compose passes CORS_ALLOWED_ORIGINS to both SPA and Django containers
  • Live demo URL updated to openlpr.computedsynergy.com

[1.3.0] - 2026-05-24

Added

  • Runtime backend API configuration via BACKEND_API_URL environment variable (no rebuild needed)
  • /api/config server-side endpoint that returns the backend URL to the SPA client
  • /health route on SPA that redirects to the Django backend health endpoint
  • AppInitializer component with spinner during config loading and error/retry state on failure

Changed

  • SPA now resolves backend API URL at runtime instead of build time
  • spa-dev.py uses BACKEND_API_URL instead of NEXT_PUBLIC_API_BASE_URL
  • Docker Compose SPA service passes BACKEND_API_URL via environment:
  • getImageUrl and getDownloadUrl are now async (runtime-resolved base URL)
  • Recent uploads on home page use shared ImageCard component
  • Health nav link uses /health (SPA proxy) instead of direct backend URL
  • Removed NEXT_PUBLIC_API_BASE_URL build arg from SPA Dockerfile (replaced by runtime BACKEND_API_URL)

Removed

  • Dead NEXT_PUBLIC_MAX_UPLOAD_SIZE build arg from SPA Dockerfile

[1.2.3] - 2026-05-24

Added

  • SPA frontend env vars (NEXT_PUBLIC_API_BASE_URL, NEXT_PUBLIC_UPLOAD_TIMEOUT) to .env.example and .env.llamacpp.example
  • SPA vars documented in AGENTS.md Environment Variables section
  • Conventions section in AGENTS.md: env var documentation requirement, no build: in compose files

Changed

  • Docker image convention documented in AGENTS.md: all images built by CI, compose only references GHCR

Removed

  • Dead NEXT_PUBLIC_MAX_UPLOAD_SIZE build arg from SPA Dockerfile (max upload size fetched at runtime via /api/v1/config/)

[1.2.2] - 2026-05-24

Changed

  • Add path filters to Django backend CI workflow to reduce CI noise on docs-only changes

[1.2.1] - 2026-05-24

Fixed

  • Grant contents:write permission to Docker publish workflows for SBOM release attachment

Removed

  • Deleted .roo/ directory (Roo Code IDE config artifacts)
  • Deleted nginx/ directory (unused — Traefik is the active reverse proxy)

[1.2.0] - 2026-05-24

Added

  • Next.js 16 SPA frontend with React 19 and Tailwind CSS 4
  • Storybook 10 with component stories for all UI components
  • Dark mode support with system preference detection in SPA
  • REST API endpoints: /api/v1/images/, /api/v1/images/<id>/, /api/v1/download/<id>/<type>/, /api/v1/config/
  • CORS support via django-cors-headers for SPA cross-origin requests
  • Configurable fixed-pixel padding for OCR crop regions
  • Disclaimer banner in SPA with no-liability and data usage notice
  • SPA pages: Home (upload), Images (history with search/filter/pagination), Image Detail (detection results, logs, raw API response)

Changed

  • Breaking: Django backend is now API-only — all server-rendered templates removed
  • Django upgraded from 4.2.30 to 5.2 LTS (supported through April 2028)
  • Python minimum version raised from 3.8+ to 3.10+
  • Frontend migrated from Bootstrap 5 / Django templates to Next.js SPA
  • Removed all Django HTML templates, forms, template tags, and web views
  • Removed legacy monolithic views.py and views_refactored.py
  • Removed nginx/ directory (Traefik is the active reverse proxy)
  • File upload accepts WEBP in addition to JPEG and PNG
  • README updated: tech stack badges, API endpoints, file structure, acknowledgments
  • AGENTS.md updated to reflect API-only Django + Next.js SPA architecture

Fixed

  • Hydration mismatch in Next.js SPA disclaimer banner (extracted to "use client" component)
  • .gitignore lib/ pattern fixed to /lib/ (was matching single-page-ui/src/lib/)

Removed

  • templates/ directory (all HTML templates)
  • lpr_app/forms.py (Django forms — only used by removed web views)
  • lpr_app/templatetags/ (template tag libraries)
  • lpr_app/views/web_views.py (template-rendering views)
  • lpr_app/views.py (legacy monolithic views)
  • lpr_app/views_refactored.py (unused leftover)
  • lpr_app/utils/response_helpers.py WebResponseHelper class
  • nginx/ directory (unused — Traefik is the active reverse proxy)

[1.1.0] - 2026-05-15

Added

  • Two-phase LPR pipeline with dynamic detection scaling for improved accuracy
  • Comprehensive canary monitoring system with automated health checks
  • Profile-based Docker Compose with merge design pattern (replaces individual compose files)
  • Custom monitoring Docker images (Prometheus, Grafana, Blackbox Exporter) with embedded configs
  • GitHub Actions workflow for building and publishing monitoring images
  • NVIDIA CUDA GPU support for local LlamaCpp inference
  • AMD Vulkan GPU support for local LlamaCpp inference
  • Dark mode support with system theme detection and automatic switching
  • Dynamic font scaling for OCR text visualization
  • Favicon for the LPR application
  • Star History chart in README
  • Coolify deployment support with custom Docker images
  • Configurable Traefik router rules via environment variables
  • Configurable service ports to avoid conflicts
  • CSRF trusted origins configuration for cross-origin requests
  • Prometheus monitoring with LPR app dashboard
  • Grafana dashboards for both LPR app and canary service
  • Blackbox exporter for HTTP probing
  • Simplified mobile pagination
  • Liability disclaimer in application header
  • Live demo link in README

Changed

  • Split monolithic views.py into modular view components (views/ subpackage)
  • Django upgraded from 4.2.7 to 4.2.30
  • Pillow upgraded from 10.1.0 to 12.2.0
  • Gunicorn upgraded from 21.2.0 to 22.0.0
  • Maximum image upload size reduced from 10MB to 250KB for optimized processing
  • Homepage updated to show 9 processed images with annotations
  • UI condensed for better screen utilization
  • Image list page now shows annotated images with OCR text and plate count
  • README reorganized with collapsible sections and compact navigation

Fixed

  • Fixed django.logstatic typo in docker-entrypoint.sh causing chown error on startup
  • Fixed metrics directory permissions by adding /app/metrics to Dockerfile
  • Fixed metrics state file (metrics_state.json) creation and permissions in entrypoint
  • Fixed canary service HTTP 502 error and connectivity issues
  • Fixed canary image cleanup functionality
  • Fixed canary dashboard configuration and connectivity
  • Fixed Prometheus and Blackbox Exporter duplicate binary names in CMD
  • Fixed Grafana provisioning duplicate UID warnings
  • Fixed metrics endpoint errors
  • Fixed multi-platform Docker build issue
  • Fixed Traefik Docker API compatibility issue
  • Fixed database file ownership to prevent readonly database errors
  • Fixed hardcoded API response data in LPR processing
  • Fixed KeyError in get_first_ocr_text() method
  • Fixed OCR results display in dark mode
  • Fixed pagination URL encoding and error handling
  • Fixed reverse pagination on image list page
  • Fixed duplicate timestamps (removed redundant ones)
  • Fixed display of correct max file size on upload page
  • Eliminated duplicate Traefik configuration
  • Fixed navigation links in README.md

Deprecated

  • Individual Docker Compose files (docker-compose-llamacpp-cpu.yml, docker-compose-llamacpp-amd-vulcan.yml)
  • Use profile-based approach with main docker-compose.yml instead

Security

  • Django security patches included in upgrade to 4.2.30
  • Pillow security patches included in upgrade to 12.2.0
  • Non-root Docker user maintained
  • Input validation and sanitization maintained
  • CSRF protection with trusted origins support

[1.0.0] - 2025-11-23

Added

  • Initial release of Open LPR - License Plate Recognition System
  • Django-based web application with license plate recognition capabilities
  • RESTful API with comprehensive endpoints for image processing
  • User-friendly web interface with drag-and-drop upload functionality
  • Image processing pipeline with bounding box visualization
  • Search and filter functionality for processed images
  • Responsive design that works across all devices
  • Docker deployment support with optimized multi-stage Dockerfile
  • Automated CI/CD pipeline with GitHub Actions
  • Multi-architecture support (linux/amd64, linux/arm64)
  • Automatic publishing to GitHub Container Registry
  • Software Bill of Materials (SBOM) generation for security scanning
  • Local inference support with LlamaCpp server
  • CPU and GPU deployment options for LlamaCpp
  • Comprehensive documentation for local deployment
  • Environment configuration templates
  • Model download automation
  • Visual showcase with screenshots in documentation
  • API documentation with detailed endpoint descriptions
  • Docker deployment guide with multiple deployment options
  • LlamaCpp and ROCm resources collection
  • Troubleshooting guides and best practices

Features

  • AI-powered detection using qwen3-vl-4b-instruct vision-language model
  • Advanced OCR integration with confidence scores
  • Bounding box visualization for detected plates and OCR text
  • Permanent storage of uploaded and processed images
  • Side-by-side comparison of original and processed images
  • Health check endpoint for monitoring
  • Download functionality for original and processed images
  • Processing status tracking with AJAX endpoints
  • Comprehensive error handling and logging
  • Security best practices including non-root Docker user
  • Environment-based configuration management
  • Input validation and sanitization

Documentation

  • Comprehensive README with visual showcase
  • Detailed API documentation
  • Docker deployment guide
  • LlamaCpp deployment guide
  • LlamaCpp and ROCm resources collection
  • Contributing guidelines
  • Code style guidelines

Technical Details

  • Django 4.x web framework
  • Bootstrap 5 for responsive UI
  • Gunicorn WSGI server for production
  • SQLite database with migration support
  • Docker containerization with multi-stage builds
  • GitHub Actions for CI/CD
  • Semantic versioning with automated tagging
  • SBOM generation for security scanning

Security

  • Non-root Docker user for security
  • Environment-based configuration
  • Secure file upload handling
  • Input validation and sanitization
  • CSRF protection
  • Security headers configuration

Performance

  • Optimized Docker image for production use
  • Gunicorn WSGI server with 3 workers
  • Efficient image processing pipeline
  • Configurable batch processing
  • Multi-architecture Docker images

Developer Experience

  • Well-structured codebase following Django best practices
  • Comprehensive documentation
  • Environment configuration templates
  • Management commands for project setup
  • Clear code organization with separation of concerns

Deployment

  • Standard Docker deployment with cloud-based AI services
  • LlamaCpp CPU deployment for local inference
  • LlamaCpp GPU deployment with AMD Vulkan support
  • Environment configuration templates
  • Automated model download
  • Health checks and monitoring

API Endpoints

  • Web endpoints for user interface
  • REST API endpoints for programmatic access
  • Health check endpoint
  • File upload and processing endpoints
  • Image download endpoints
  • Search and filtering endpoints

Known Limitations

  • Large images (>10MB) may require increased memory allocation
  • GPU acceleration is currently limited to AMD GPUs with Vulkan support
  • Concurrent processing is limited by Django's development server (use Gunicorn in production)

Dependencies

  • Django 4.x
  • Bootstrap 5
  • Gunicorn
  • Docker
  • LlamaCpp (for local inference)
  • Qwen3-VL model

License

  • MIT License

[Unreleased]

Planned

  • Batch processing capabilities
  • Real-time video stream processing
  • Additional language support for license plates
  • Performance optimizations
  • Mobile application