Future of musl locale/strerror(). In LC_MESSAGES=C, strerror(ENOENT) = ?
Obviously things are oversimplified in a tweet. The question is whether to keep English-phrase error msgs in C locale or use the E* codes (more 'language-agnostic').
-
-
In the former case, it costs more size to make the translation framework clean. In the latter, making the translation framework clean also makes it smaller (and you can have English-phrase error msgs still in "en" locale, just not "C").
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
wouldn't that be contrary to the expectation for e.g. LC_ALL=C ?
-
Maybe depending on what your expectations are. There's no std requirement that they be English messages.
-
No, but a locale where strerror(ENOENT)="ENOENT" is kinda useless by default. It is meaningless as localised messages in a (default) language.
-
In the past I'd heard the opposite from a lot of people - that the E* codes were more useful because they could look them up in docs/manpages, whereas the English msgs required figuring out which E* code they came from.
-
I don't disagree with that, been there too.
-
For that reason, I tend to like strerror(EFOO)="EFOO". Also reducing anglocentricism. But I know it's unconventional & likely controversial; that's why I'm trying to get a feel for ppls opinions on it.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.