New datasets to use the current default encoding.