Version 6.3.0 - Logging Cleanup & UX Improvements
LATEST
Released October 28, 2024 - Cleaned up verbose logging and improved user experience
User Experience
- Cleaner Logs: Removed verbose trace messages and debug output
- Essential Logging: Kept only critical error messages and important notifications
- Reduced Noise: Eliminated repetitive keyword processing messages
- Better Performance: Less logging overhead during batch processing
Technical Improvements
- Hierarchical Keywords: Maintained full functionality with cleaner implementation
- API Integration: Streamlined prompt generation logging
- Error Handling: Preserved essential error reporting capabilities
- Production Ready: Professional-level logging for end users
What Changed:
This release focuses on improving the user experience by removing excessive debug logging while maintaining all functionality. The plugin now provides cleaner, more professional log output without compromising error reporting or troubleshooting capabilities.
Version 6.2.0 - Stability & Bug Reporting
Critical stability fixes for batch processing and new bug reporting feature
Critical Bug Fixes
- Batch Processing Fixed: Resolved issue where first 2 photos would get stuck when processing 1000+ photos
- Exponential Backoff: Added 2s → 4s → 8s retry delays to prevent API rate limits
- Task Pacing: Added configurable delay between starting concurrent tasks (default 1s)
- Cold Start Protection: Prevents overwhelming API with simultaneous requests at startup
New Features
- Bug Reporting: New "Bug Reporting" section in Plugin Manager
- GitHub Integration: Direct link to report issues on GitHub
- All Providers Fixed: Retry fixes applied to Gemini, OpenAI, and Ollama APIs
- Better Logging: Enhanced logging for retry attempts and delays
Important Fix:
Users processing large batches (1000+ photos) no longer need to run analysis 3 times. The plugin now handles cold connections, rate limits, and concurrent requests gracefully. This release significantly improves reliability for power users.
Version 6.1.0 - Migration Update
Automatic model migration and performance optimization - All users seamlessly upgraded to Gemini 2.5 Flash
Automatic Migration
- Model Upgrade: Gemini 1.5 Flash users automatically migrated to 2.5 Flash
- Zero Configuration: Migration happens transparently on plugin load
- Version Tracking: Plugin now tracks version for future automated updates
- Migration Logging: Clear logs for transparency and debugging
Performance Benefits
- Better Results: Gemini 2.5 Flash provides improved accuracy
- Same Cost: No price increase, just better performance
- Future-Proof: Foundation for seamless future migrations
- Backward Compatible: All existing settings preserved
Migration Info:
Users with Gemini 1.5 Flash are automatically upgraded to 2.5 Flash on next Lightroom restart. Other model configurations (2.0 Flash, 2.5 Pro, OpenAI, Ollama) remain unchanged. No manual action required.
Version 6.0.0 - Major Release
Critical Gemini API compatibility update - Fixes empty results with new Gemini 2.5 models
Critical Fixes
- Gemini API Update: Fixed compatibility with new Gemini 2.5 models
- Empty Results Fixed: Resolved MAX_TOKENS response truncation issue
- Response Parsing: Enhanced parsing for new API response format
- JSON Handling: Improved truncated response recovery
Technical Improvements
- Token Limit: Increased from 1024 to 4096 tokens
- Error Handling: Better finish reason detection (MAX_TOKENS, SAFETY)
- Multi-format Support: Handles old and new Gemini response formats
- Robust Parsing: 5-method comprehensive text extraction
Important Update:
This release fixes critical issues with Gemini 2.5 Flash/Pro models that were causing empty results. Users experiencing "empty results" should update immediately. All three AI providers (Gemini, OpenAI, Ollama) now work reliably.
Version 5.0.0 - Major Release
Revolutionary update with OpenAI GPT-4V integration - Three AI providers in one plugin!
- OpenAI GPT-4V Support: World-class image analysis with OpenAI's vision models
- Three AI Providers: Choose between Gemini, OpenAI, or Ollama
- GPT-4o Integration: Latest OpenAI model with enhanced vision capabilities
- Unified Interface: Seamless switching between all providers
Version 4.0.3
Previous stable release with new prompts and hierarchical keywords
- New Prompts: Enhanced AI analysis with improved prompt templates
- Hierarchical Keywords: Better keyword organization and structure
- TIFF Thumbnail Fix: Improved compatibility with TIFF image formats
- Bug Fixes: Various stability improvements and fixes
Version 4.0.0 - Major Release
MAJOR
Revolutionary update with local AI processing support
New Features
- Ollama Support: Local AI processing for complete privacy
- AI Provider Selection: Choose between Gemini (cloud) or Ollama (local)
- Local Processing: Images never leave your computer with Ollama
- Provider Factory: Clean architecture for multiple AI providers
Improvements
- Plugin Rebranding: Now "AI Image Tagger" (was "Gemini AI")
- Enhanced UI: Provider selection with conditional settings
- Better Settings: Reorganized for provider management
- Full Documentation: Setup guides for both providers
Version 3.5.0
- Collections Simplification: Streamlined auto collections management
- UI Improvements: Removed non-functional smart collections option
- Better User Experience: Cleaner configuration interface
Version 3.4.0 - Global Reach
- Full Internationalization: Support for 16 languages
- Global Accessibility: English, Spanish, French, German, Italian, Portuguese, Russian, Japanese, Korean, Chinese, Dutch, Polish, Swedish, Norwegian, Thai
- Cultural Adaptation: Localized text respecting linguistic conventions
- Unicode Support: Full support for non-Latin scripts