Re: Two questions

[Gediminas Paulauskas <menesis delfi lt>]

Carlos Perelló Marín wrote:

> El lun, 21-01-2002 a las 20:53, R.I.P. Deaddog escribió:
> 
>>On Tue, 22 Jan 2002, Yukihiro Nakai wrote:
>>
>>
>>>>I am the zh_CN translation coordinator. I have two questions:
>>>>
>>>>1. Is it ok to convert all zh_CN translation (gb2312 encoding) to
>>>>   UTF-8? I mean for both GNOME 1.4.x and GNOME 2.x. I don't want to
>>>>   break anything.
>>>>
>>>Hi, about Japanese team policy, .po is translated in
>>>
>>>GNOME 1.4 - ja_JP.eucJP (legacy locale)
>>>GNOME 2.0 - ja_JP.UTF-8
>>>
>>For Gnome 1.4.x, converting everything to UTF8 would break a lot of
>>things.


Converting po files to utf-8 won't break anything unless you are using an 
old version of glibc. Everything should work the same way as before. The 
only thing which affect strings which you get from gettext is your system 
locale.

> How it goes for Gnome 2.0 is unknown to me; anyone knows more on
>>this matter?
>>
> 
> 
> We must use UTF-8 for all .po files before GNOME 2.0 release.


No, you a wrong. There is no need to convert translations to UTF-8. Whatever 
  charset a .mo file is in, glibc automatically converts it to locale 
encoding, or, if the app say say to gettext so -- in UTF-8. All apps using 
Gtk+2.0 must do that, i.e. call

bind_textdomain_codeset (GETTEXT_PACKAGE, "UTF-8");

so that they get all text in UTF-8. Coverting string on the way from .mo 
file to app is a work of glibc or gettext, which do not change from Gnome 1 
to Gnome 2 in any way.

-- 
Gediminas