bug#57531: 28.1; Character encoding missing for "eo"

[Jonathan Reeve]

> The characters in that post are supported by Latin-3, and I had no
> problem saving them.

It’s not about saving them, it’s about how they’re displayed. If the rest of my 
system correctly uses unicode for the `eo' locale, /because it’s a unicode 
locale,/ but emacs is the only one that guesses it should be Latin-3 instead 
(for no reason that I can find), then it’s emacs which incorrectly handling 
this locale.

> That’s okay, but then all you need to say in Emacs is
>
>   (prefer-coding-system ’utf-8)

If emacs were really following system settings, it would set the encoding to 
utf-8 without needing extra customization from the user, since `eo' is a UTF-8 
locale. And incidentally, there’s no such locale as `eo.utf-8', from what I can 
tell.

> I don’t know why glibc did that, and glibc-supported systems are not
> the only ones where we want to support Esperanto.  So if the above
> simple customization fixes your problem, I’d prefer not changing the
> default for everyone.

It’s not about changing the default for everyone, it’s about fixing an 
incorrect character encoding, making it so that emacs correctly respects the 
locale’s character set. There’s no reason to have the latin-3 character set, 
except for backwards compatibility, but that’s irrelevant in the case of 
Esperanto, since I know of no program, document, or anything else that would 
use latin-3 for Esperanto. Since the locale for Esperanto is a relatively new 
invention, it doesn’t have the same history of legacy character encodings as a 
language like English.

In fact, not even emacs seems to define it as such, judging from this line in 
`locale-language-names': `("eo" . "Esperanto")'. It just seems as if the author 
of that variable didn’t know what character set to assign, and left it blank.