Totally agree. And when one of the suggestions for porting to Py3 is "Decide if the feature is really that important. Maybe you could drop it?", you know something is wrong.