And next time you cite the docs, please try to link to it like [doc://tr]

> I didn't think that was needed anymore.

Perl still needs to be able to operate with byte strings and this backwards compatibly.

So you need to indicate that the literals in your code contain multi byte characters in utf-8 ( provided your editor is set to produce utf8)

Otherwise you'll need to use Encode to convert between the two string states.

This is also necessary when converting from another encoding than utf8.

Unless this has changed again, (I normally work in ASCII) use utf8 specifically declares that the Perl program itself is in UTF-8. As a side effect, (because the DATA handle is the handle the parser used to read the code) it causes any __DATA__ block to also be read as UTF-8.