Caret movement api These should give the user a logical result based on what they see on screen... thus they locate predominately by *pixels* not char index. (These will generally coincide with monospace fonts tho!)