We had a bug in a library, that was caused by one of the inputs being Unicode.
It was fixed by adding use utf8;
to the script using that library.
However, adding use utf8;
to the library itself (so ALL scripts using that library would be fixed) had no effect.
Why? Can this be addressed?