[pacman-dev] Localization issues

[Dan McGee]

On 3/6/07, Roman Kyrylych <roman.kyrylych at gmail.com> wrote:
> 2007/3/6, Dan McGee <dpmcgee at gmail.com>:
> > On 3/5/07, Nagy Gabor <ngaba at petra.hos.u-szeged.hu> wrote:
> > > Hi!
> > >
> > > It would be nice, if pacman accepted Y and N as yesno answer
> > > too. Because if a message hasn't been localized yet (for
> > > example 'Proceed with installation? [Y/n]' to hungarian now), it's
> > > annoying that if answer Y, pacman exits (because Y is localized to I).
> > >
> > > Bye, Nagy Gabor
> > >
> > > P.S.: Yes, you may say: There may exists a language where
> > > localized(I)==N or similar; so the localized answer should be checked
> > > first. In this pathological case (again .-) the untranslated
> > > messages are really annoying :-)
> >
> > The best solution is to either update the translation or drop it (put
> > it in storage until someone wants to update it). This would be very
> > hacky to implement what you suggested, and not that useful to
> > languages where Y and N have flipped usage from their English
> > counterparts.
>
> Translating Y/N to use native language chars (i.e. J/N in German) is
> bad for users that use different keyboard layouts for English/native
> text input.
> For example in Ukrainian 'Proceed with installation? [Y/n]' can be
> translated as 'Почати інсталяцію? [Т/Н]' but to enter Т or Н chars
> user must to switch keyboard layout.
> So, for Cyrilic languages it's better to stick with Y/N but with
> translated question.

We really have little to do with this- It is up to the translators to
make that decision. We simply put both the questions AND the answers
in gettext (ex _("Y") and _("N")). If you want to leave it, there is
no problem at all with that- just don't translate the [Y/n] part of
the question or the literal Y and N strings.

-Dan