On 2023-11-01 19:38, Andrei Borzenkov wrote:
On 01.11.2023 21:17, Carlos E. R. wrote:
On 2023-11-01 18:58, Andrei Borzenkov wrote:
On 01.11.2023 17:27, Carlos E. R. wrote:
On 2023-11-01 15:16, Andrei Borzenkov wrote:
On Wed, Nov 1, 2023 at 5:11 PM Carlos E. R. <...> wrote:
...
Why, what is going on?
You need to export them. Show
env | grep -E 'LANG|LC_'
cer@Laicolasse:~> env | grep -E 'LANG|LC_' LC_MEASUREMENT=es_ES.utf8 LC_MONETARY=es_ES.Utf8 LANG=en_US.UTF-8 LC_NUMERIC=es_ES.UTF-8 LC_TELEPHONE=es_ES.utf8 LC_TIME=en_DK.UTF-8 cer@Laicolasse:~>
Hum. Some are missing.
What is not clear in "you need to export them"?
Everything.
How do I do that?
Is it a joke?
No.
export VARIABLE_NAME
Ok, so doing . ~/.i18n in .bashrc is not enough, now I have to export individually all the LC_variables one by one, something I wanted to avoid.
assuming you are using bourne-compatible shell.
Why is, for instance, LC_MONETARY exported, but LC_NAME is not?
Because something exported LC_MONETARY before your script was sourced. If you want to know, you can examine environment of parent processes upwards to find out where these variables appeared.
I handle both the same way.
Which does not mean both have the same initial state when you start handling them.
Ok, thanks for the explanation. -- Cheers / Saludos, Carlos E. R. (from openSUSE 15.5 (Laicolasse))