[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: No german umlauts in console and xterm



Hi,

From: Miros/law `Jubal' Baran <baran@makabra.knm.org.pl>
Subject: Re: No german umlauts in console and xterm
Date: Wed, 6 Sep 2000 13:28:51 +0200

> 6.09.2000 pisze Tomohiro KUBOTA (kubota@surfchem0.riken.go.jp):
> 
> > There is 'language-env' package for German, Spanish, Japanese, Russian,
> > and Thai (for Woody).
> 
> Maybe we need some more general way to define default system localization
> (giving the users a possibility to change this default is very good,
> but I think we need a standardized way to change the defaults -- which
> I don't see)

Yes, I can imagine that 'language-env' will be a part of installer 
and the user is asked to choose mother tongue.  The settings can be
machine-wide (in /etc directory) or user-specific (by dot-files).

The point is, I think, the way of setting should be extremely extensible.
Database for LANG variable, console font, keymap, X font, and so on
is not sufficient.  For example, Linux console cannot display
multibyte characters.  Xterm and most of X terminal emulators 
also cannot.  Thus my .bashrc has

if [ "$TERM" = "linux" -o "${TERM%-*}" = "xterm" ]
then
  export LANG=C
else
  export LANG=ja_JP.ujis
fi

This is written by 'language-env'.  And more, what software is popular
is different from country to country.  For example, 'language-env'
Japanese setting has very complecated Emacs setting because Emacs
is one of a few Japanese-enabled editor and very popular in Japan.
There are many softwares which need special settings  (For example,
many softwares such as Window Maker have 'multibyte' configuration
item which has to be enabled for multibyte languages).  Since we can
hardly have settings for all languages and all softwares, we have to
select some of softwares.

And more, softwares to be localized have to be configurable by
/etc files.  Canna, one of Japanese input methods, cannot.

At last, required settings are very different from country to
country.  For German, settings so as to display umlaut characters
on console are required, while, for Japanese, settings for 
localized softwares to deny localization on console environment
is required.  Font setting for Xterm is needed for Polish while
invoking Hanterm instead of Xterm is needed for Korean.  Settings
and invocation of ne of input conversion servers is needed for 
Japanese.  I don't know what special settings are needed for 
right-to-left languages such as Hebrew and Arab.  Though Mule is 
the only software which I know can handle right-to-left languages,
is it appropriate to set EDITOR=mule ?

Thus it is difficult to have a *smart* way of language configuration.
'language-env' is what I think is best for current Debian system.
I think it evolve with the development of internationalization of
Linux and Debian.

However, I will be happy my 'language-env' would be a skelton for
internationalization of Debian installer...

---
Tomohiro KUBOTA <kubota@debian.org>
http://surfchem0.riken.go.jp/~kubota/


-- 
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org


Reply to: