Re: man-db locale support
On Wed, Sep 29, 1999 at 08:38:02PM +0300, Fabrizio Polacco wrote:
> See bug#39281
>
> I think we need to define a policy.
> Some people wants to use only the language code, other seems to need the
> full range of options.
> Sometimes it depends on the local HOWTO author, and on his preferences
> :-)
Hmm... Let me add few thoughts here. My native language is Russian. :) Here,
in Russia, we have a really nice problem with encoding schemes. At the moment
there is a whole lot of them: KOI8-R, ISO-8859-5, Windows-1251, CP866. (There
could be others, but I do not remember them at the moment :) As result, people
(depending on their's tastes) use something like:
LANG=ru_RU.KOI8-R
LANG=ru_RU.CP1251
LANG=ru_RU.IBM866
or just
LANG=ru
The latter means what? :) Yes, something that is standardized (sp?). ISO (I believe
the most standards are from there) says that the latter case should be
interpreted as
LANG=ru_RU.ISO-8859-5
That is: ISO-8859-5 is the standard encoding for Cyrillic (Russian ABC).
OK. What I want to say? All these encodings are quite different. If I want
to install a manual page prepared in KOI8-R (which is pretty common in Unices
world) I must use /usr/share/ru_RU.KOI8-R/manN/ directory and not
/usr/share/ru/manN as otherwise it will be unreadable for those (who follow the
above mentioned standard and) who do not have proper settings for displaying
unit.
Returning to the bug #39281, I believe that a patch is needed. :) But not the
one in the BTS. The correct code should check for all possibilities. If one
has LANG=ru_RU.KOI8-R and she issues `man command', man should look into all
directories:
/usr/share/ru_RU.KOI8-R/manN
/usr/share/ru_RU/manN
/usr/share/ru/manN
etc (according to MANPATH and current locale settings)
What do you think?
Cheers,
--
Mike
Reply to: