v1.1.1 - Adding Languages

📜 Changelog

🎉 v1.1.1 - 2025-08-23

✨ New Features

• Chat Memory: AI now remembers the last 8 user turns.
• Profile System: Add about, job, and user facts to enrich AI responses.
• Model Selection: Users can select from installed Ollama models (default: gemma3:4b).
• Multi-language System:
  • Added support for 9 languages: English (US, UK), Bahasa Indonesia, 日本語, 한국어, 中文（简体）, Português, Español, العربية.
  • Fallback chain: en_us → id → target.
  • UI displays native names (e.g. Bahasa Indonesia, 日本語).
  • Language dropdown moved from header → Settings dialog.
  • Backend endpoints: GET /i18n/<lang>, GET /settings, POST /settings.

🐛 Fixes

• Fixed bug: Chat history could not be accessed in v1.0.0.
• Fixed bug: GPT-OSS "thinking" text leaking into UI.
• Fixed bug: Language loader not detecting Indonesian.
• Fixed bug: POST /settings writing config to wrong path.
• Added _DEFAULT_I18N in UI to prevent raw keys (e.g. btn.send) from showing.

---

🎉 Release v1.1.0 - 2025-08-23

This update introduces several important new features and bugfixes, improving the overall functionality and flexibility of the Local AI Assistant.

✨ New Features

• Chat Memory: the AI can now remember up to the last 8 user messages in a conversation.
• Profile System: add custom information about yourself (e.g. "What do you do", "Anything else the AI should know about you"), which is used to enrich responses.
• Model Selection: you can now select which Ollama model to use from the ones installed on your system (default still gemma3:4b).

🐛 Bug Fixes

• Fixed issue where chat history could not be accessed in v1.0.0.
• Cleaned up UI "garbage" text that leaked from GPT-OSS thinking process.
• Minor stability improvements.

📌 Notes

• Default persona and identity system remain minimal, but more customization will come in future updates.
• Ollama integration is expanding: while gemma3:4b is still the default, users can swap to other models installed locally.

---

v1.0.0 - 2025-08-10

🚀 Initial Release

• Flask backend with /config, /chat, /chats.
• PySide6 UI with chat window, settings, history, and identity prompt.
• Ollama integration (default gemma3:4b).
• Local storage in /data/.
• Known Issues:
  • Chat history not accessible.
  • Occasional leaked meta/thinking text.

v1.1.0 - Chat Memory, Profile & Model Selection

🎉 Release v1.1.0

This update introduces several important new features and bugfixes, improving the overall functionality and flexibility of the Local AI Assistant.

✨ New Features

• Chat Memory: the AI can now remember up to the last 8 user messages in a conversation.
• Profile System: add custom information about yourself (e.g. "What do you do", "Anything else the AI should know about you"), which is used to enrich responses.
• Model Selection: you can now select which Ollama model to use from the ones installed on your system (default still gemma3:4b).

🐛 Bug Fixes

• Fixed issue where chat history could not be accessed in v1.0.0.
• Cleaned up UI "garbage" text that leaked from GPT-OSS thinking process.
• Minor stability improvements.

📌 Notes

• Default persona and identity system remain minimal, but more customization will come in future updates.
• Ollama integration is expanding: while gemma3:4b is still the default, users can swap to other models installed locally.

v1.0.0 - Initial Release

🚀 Initial Release (v1.0.0)

This is the very first public release of Local AI Assistant powered by Ollama.
It provides the foundation of the project, with Flask backend, PySide6 UI, and local data storage.

✨ Features

• Flask backend with endpoints: /config, /chats, /chat
• PySide6 chat UI with modern layout
• Chat history (saved locally)
• Custom persona & identity support
• Ollama integration (default: gemma3:4b)
• Local storage in data/ folder

⚠️ Known Issues / Bugs

• Chat history could not be accessed properly (error when loading sessions).
• Occasional UI glitches with some responses.
• Early integration with Ollama (limited to gemma3:4b).
• Basic persona system, still minimal.

---

This release sets the baseline. Upcoming versions will include bug fixes (history access), AI memory improvements, profile feature, and UI cleanup.