He may have meant accept only UTF-8/16. That is the very nice thing about UTF-8 though is that it plays so nice with routines that can accept ASCII or Latin-1.

He may have meant accept only UTF-8/16. That is the very nice thing about UTF-8 though is that it plays so nice with routines that can accept ASCII or Latin-1. You can't use old routines to character count/change case/etc., but at least they won't corrupt your string by accident.