Hi, From: "Petr Kolar" <[EMAIL PROTECTED]> Subject: Re: Gettext and LC_MESSAGES Date: Mon, 12 Jun 2000 09:16:15 +0200
> I admit I have not choosen the best examples. But currently the name > of the capital can be placed only in LC_MESSAGES (if needed). And AFAIK > LC_MONETARY contains only int_curr_symbol and currency_symbol (the > abbreviations), so full names of the currencies must be also given in > LC_MESSAGES (also only if needed by some program). If the messages contains > word "colour" or "theatre", there should be also more catalogs for english > (some with "colour" and some with "color", some with "theater" and some > with "theatre") - and these catalogs will be in en_US en_GB etc., not in > en. I have never thought about "color" and "colour"... I agree with you. Then the behavior of gettext that LANG=de gives German messages should be changed, doesn't it? (I think catalogs can be installed in "en" directory if messages of en_US, en_GB, and etc are identical.) > It seems it is due to lack of documentation and experience. Like most > (American) programmers did not write programs that worked with 8 bit > characters in old days, many European programmers kill the program for > Japanesse today, as they don't know, how to write it correctly, and > also many Japanesse don't realize that there are other charsets (eg. > ISO-8859-2) between ASCII or ISO-8859-1 and multibyte charsets. > I don't think it is for bad intention... I am writing a document on internationalization at Debian Documentation Project (http://www.debian.org/doc/ddp). However, it is intended to be read by developers, not by users, and does not mention the way to set LANG variable and so on. Japanese people don't use ISO-8859-1. Though it is true that ja_JP LC_CTYPE locale uses ISO-8859-1, we use only 7-bit part of it (same as ASCII). I don't understand what should we care about to handle ISO8859-n (n>1) character sets... (For example, xterm can display ISO8859-2 charset by specifying font name.) Petr, Could you please write something and contribute to my document on i18n? I prepared chapters for each language and each example of i18n-related improvement of softwares. I also would like you to contribute my 'language-env' package. To handle Czech (I guess you are a Czech speaker from your mail address... Sorry if I am wrong.) language, I think you need some configuration. For example, a line in ~/.Xresources to display ISO8859-2 font in xterm. 'language-env' is a Debian package to set up users' personal environment for the users' native language. I would like 'language-env' to support Czech. Please read http://surfchem0.riken.go.jp/~kubota/linuxwork/language-env for detail. --- Tomohiro KUBOTA <[EMAIL PROTECTED]> http://surfchem0.riken.go.jp/~kubota/
