Page 1 of 1

Duplicate strings/translations due to language + country combos?

Posted: Tue Jan 30, 2018 10:31 pm
by northgrove

I've been fixing a number of translations to Swedish for Budgie Desktop. Quite a bit of "särskrivningar", a surprisingly common, yet glaring issue with translations made by Swedes (we never write e.g. "cookie cutter", always "cookiecutter")! While doing so, I discovered there were actually two complete sets of strings to translate, both with translation work done to them!

First we have budgie-desktop.pot here: [sv] ... nguage/sv/

Then we have budgie-desktop.pot here: [sv_SE] ... age/sv_SE/

Both are for the same project (budgie-desktop), contain the same strings to translate... and are translated by different translators to different things! :(

I got confused but I think I contributed to the correct one though; "sv_SE". "sv" ("Swedish" language with no "Sweden" country) seemed to be the wrong one and unused? It should perhaps be deleted? A bit funny too that it seemed to sometimes have been more carefully translated, hmm.

PS. There is something similar going on with Finnish (fi) vs Finnish (Finland) (fi_FI) and Norwegian (no) vs Norwegian (Norway) (no_NO). In these cases, fi_FI and no_NO, the languages that logically (?) should have the actually used translations in this case have 0% translated. Here, we instead have the languages with no country specified being the ones translated. Are those also the ones used? A bit confusing... BTW, Norwegian is never just "Norwegian". In Norway you write either Bokmål or Nynorsk, never just "Norwegian". So actually nb_NO and nn_NO should be used here, but nn_NO doesn't even exist, just nn. And I have no idea what "no" is. I assume Bokmål since it's the by far more common one in general throughout Norway. But nb_NO has a different set of contributions to it, like here with Swedish. :( Summarized I wish we only had sv_SE, fi_FI, nb_NO, and nn_NO. No confusion, no risking to translate strings that won't even be used.