Tasks

List of TODOs currently being worked on, prioritized.

Core

• Option to automatically setup A1111 from scratch
  • Clone latest compatible release
  • venv, install process
  • Initial setup for JSON config files
  • Have SettingsView automatically apply changes to JSON config
• Prepare for item generation queue
  • Sidebar item progress should include percentage (loading indicator?)
  • Refactor Txt2Img API interface with modular design
• Transition to Observation framework
  • Research foundational changes (ensure compatibility)
  • Minimum deployment target >= 14.0
• Merge PromptModel, StoredPromptModel into singular model entity
  • Repeat for child models where ideal
  • Remove all unnecessary mapping functionality
  • Research how to best incorporate DirectoryObserver classes

Features

• CoreML
  • Simple prompts foundation
  • Additional configurations
• PyTorch
  • Simple prompts foundation
  • Txt2Img
  • VAEs
  • LoRAs
  • ControlNet (IPAdaptor, ...)
  • Img2Img
  • Plugins (AnimateDiff, ...)
  • Additional APIs
• Prompt Generation Queue
  • Isolate, modulate the prompt items for queuing
  • Queue tracker for individual items
  • Add/remove items from queue
• Refactor DirectoryObserver managing models (checkpoints, LoRAs)
  • Refactor CheckpointsManagerView to generic directory ManagerView
  • Have managing models directly conform for presentation

Improvements

• Rework PromptModel save states
  • Save state triggers on sidebar item change
  • Save state triggers on window/app close
• Rework PromptModel onChange modifiers
• Stored image previews
  • Stored image compression
  • Stored image cache
• Restart script environment
  • Reset app variables on new script load
  • Refactor 'Disable VRAM optimizations' to reload instance of restart app
• Setup proper SwiftData model migration plans, schemas

Future

• Dynamically create views, elements based on Automatic API config
  • Map API hooks for dynamic element input
• SwiftUI view builder for customizing dynamic views
• Mobile client controller
  • Use common UI for presentation
  • Forward API requests to desktop app
  • View, organize generated content