mikaela.github.io/n/localectl.md

3.4 KiB
Raw Blame History

I have started using my devices in en_GB, but on PCs that can have issues with different types of different locales unless some things are borrowed from other locales.

% sudo localectl set-locale LANG=en_GB.utf8 LC_TIME=en_DK.utf8 LC_MONETARY=fi_FI.utf8 LC_NAME=fi_FI.utf8 LC_TELEPHONE=fi_FI.utf8

  • LANG is the language and defaults for other variables if they are unset and as I said I prefer en_GB which is better in Europe for matching the standards better than en_US.
  • LC_TIME sets the time format and en_DK gives ISO 8601 date format for everything respecting LC_TIME (at least date and M̀ATE panel clock dont respect it).
  • LC_MONETARY sets the currency and how sums of it are managed and the only issue with en_GB is it using £ instead of € which affects at least spreadsheets.
  • LC_NAME sets name format and I prefer Finnish (just the name) to British which has ms/mrs/mr and I hate them being gendered and I doubt it understands mx.
  • LC_TELEPHONE sets telephone number format and I set it to fi_FI as it defaults international numbers to +358 and I am more likely to be typing Finnish numbers than British. I imagine it can affect office tools, and its here more just in case.

And naturally to use these locales, they must be compiled.

So /etc/locale.gen must include the lines

en_DK.UTF-8 UTF-8
en_GB.UTF-8 UTF-8
# If you don't have en_US, things will break!
en_US.UTF-8 UTF-8
fi_FI.UTF-8 UTF-8

and as always, after editing that file, you must run sudo locale-gen.

(Debian & deriative users, you have dpkg-reconfigure locales that merges the editing and locale-gen and that is probably what you are supposed to use.)

It seems like I didnt even say anything about that UTF-8 part, but its 2016 and everything is UTF-8 (or your things are horribly broken and the rest of the world hates you).

Sources/thanks:

  • Arch Wiki: Locale, LC_TIME
  • Locale Helper: en_GB
  • Locale Helper_ fi_FI
  • Wikipedia: UTF-8
    • Random misplaced advice on disabling charset fallback in your IRC client and being UTF-8 only so you see when someone is not using UTF-8 and dont submit useless bug reports to bots/bridges/whatever thare are UTF-8 only as supporting every charset is not possible as IRC has nothing to declare character set and there is zero reason why you woulnt be using UTF-8!
  • ISO 8601
    • You should read it or at least be aware of it especially if you are in contact with people from other countries and even more if you are in international communities!

2019-12-27: I dont see LANGUAGE mentioned here, but it was blocking me from changing language of GNOME and sudo dpkg-reconfigure locales in the end gave me *** update-locale: Warning: LANGUAGE (en_US:en) is not compatible with LANG (fi_FI.UTF-8). Disabling it..

Either I was wrong on it being list of fallback languages I wish to use, or GNOME has different view on it, but as I think I have seen errors related to it before, I will drop LANGUAGE from the variables I set.