TODOs:
- Limit selection area to actual displayed image area
- Attention: Window resize
- Exchange corner markers (X, Y, Width, Height) when dragging one edge past another
- Convert selected area data into image coords (applying scale-to-fit diff to them) for data extraction
- Save selected data to ~/.face
- Move pages into separate QML files, for code org
IDEAS:
- Support other profile image locations (somewhere in AccountsService?)
- Load current ~/.face (and other profile images)
- Store previously saved profile pictures, allow reverting back to them