Release Version 6.2.1: UI Design Overhaul, PDF Parsing, and Enhanced Vision Requests · TannerMidd/minimal-chat

New Features

PDF File Parsing Support: Added the ability to parse PDF files and upload them as context to conversations. (#138)
Audio Queueing: Implemented audio queueing to prevent overlap TTS tracks and modularized the InteractMode component. (#141)

Settings Dialog Refactor: Refactored the SettingsDialog template code into individual configuration section components for better maintainability. (#143)
Vision Requests Enhancements: Updated vision requests to be saved to the conversation context and render the uploaded image on the screen. Vision requests are now integrated into the conversation flow, rather than being one-off requests. (#142)
Messages List UI Design: Updated the UI design of the messages list with larger left and right paddings to focus interactions in the center. User messages are now uniquely different but distinct from LLM response messages. (#142)
Conversations Panel Design Overhaul: Redesigned the conversations panel for a more streamlined and minimal look. This includes a new context menu, persistent "new conversation" option, and synchronized max tokens for titles across models. (#139)
Enhanced ChatInput UI: Improved the ChatInput UI with a flex-based layout and an InteractMode button. (#140)

Fixed Max Token Param: Corrected the max_token parameter name for GPT requests. (#139)
Fixed Overflow Bug: Addressed an overflow issue and moved the hover delete icon to the beginning of the text in the conversations panel. (#139)
Fixed Click Issue: Resolved a click issue with the conversation panel context menu icon. (#139)
Fixed Issue Importing Settings Files: Corrected an issue related to importing settings files. (#143)
Fixed copyText Logic Issue: Updated copyText to handle the updated messages object structure.

Mobile Layout Adjustments: Made width adjustments for messages in mobile layout for better readability. (#142)
Removed Redundant User Label: Removed the "user" label as it was deemed redundant, and updated message icons to be animated and scale up when hovered. (#142)
Interact Mode Enhancements: Updated interact mode to read responses of vision requests and made various layout and design adjustments. (#142)
Removed User Prompt for Image Upload: Eliminated the requirement for a user prompt when uploading an image to the conversation. (#143)
Tooltip for Token Count: Added tooltips to show token count in the conversation list for a more informative UI. (#139)
- Final update removes tooltip and left data icon and instead display the token count directly below the title in a small font.
Added Copy Icon for User Messages: Added copy icon to user messages.

compressed-demo.mp4

Thanks to @o-stahl for the continued work and improvements!

As always, thank you for your continued support and feedback!