Drop sr-Latn?
flodolo opened this issue · comments
That was a language that we used for Firefox OS, and it was generated automatically from Serbian (Cyrillic).
Is there anything generating that now? I'm asking, because my linter caught an error in verificationReminderFirst-description(auth.ftl), and that was fixed a month ago in Serbian.
@clouserw
Putting this on the radar ;-)
Thanks for the ping. I think we generate this at https://github.com/mozilla/fxa-content-server-l10n/blob/master/scripts/extract_strings.sh#L86
Good to know we can drop it, I'll think on if there is anywhere else we reference it.
Ah, the translation was submitted 1 month ago, but only approved 8 days ago
c0173e5
That's why it's not showing up as fixed yet.
Still, I don't think there's value in keeping it around at this point.
Having now more knowledge of the extraction process, I think that we really should remove this locale.
The process only transliterate the po file via msgfilter, there is nothing doing that for the FTL bits.
@clouserw
Do we have any data on how many users end up on the sr-Latn pages, in absolute numbers and compared to standard sr? If the number is particularly high, we should add conversion for FTL files, if not we should drop it completely.
For the past 30 days sr gets about 5-6k hits/day. sr-LATN gets between 0 and 1 per day.