[Fwd: Re: [Nautilus-list] [Fwd: About wchar support.]]

[James Su <suzhe gnuchina org>]

-------- Original Message --------
Subject: Re: [Nautilus-list] [Fwd: About wchar support.]
Date: Thu, 14 Jun 2001 13:19:15 +0000
From: James Su <suzhe gnuchina org>
To: Darin Adler <darin bentspoon com>
References: <200106140509 f5E59C911795 gnuchina org>



Hello,
   If we use UTF-8 or char*, we cannot handle the string directly. Like 
my patch, we must do lots of mbs<->wcs convertion in many functions. It 
will increase the program's complexity and affect its performance.

Regards,
James Su

Darin Adler wrote:

> On Thursday, June 14, 2001, at 05:11  AM, James Su wrote:
>
> >    I found that nautilus and eel treat almost all string as char *, 
> > it's not an i18n compatible method.
> > May be it's better to use wchar_t * (GdkWChar *) instead of char * 
> > (gchar*
> > ) as internal string type.
> > It would simplify the string handle and will achieve better i18n 
> > compatiblity.
>
>
> Thanks for your suggestion.
>
> The approach planned for GNOME 2 is to use UTF-8 and char * rather 
> than UTF-16 or local 16-bit character sets and GdkWChar *. This is not 
> something that's specific to Nautilus, but rather something that's 
> going to be done this way across GNOME.
>
> Saying that treating strings as char * is not an i18n-compatible 
> method is overstating the case. You can make functions that handle 
> multibyte characters correctly without always using wide characters, 
> and this is what the GNOME hackers are planning for GNOME 2.
>
> Perhaps if I was doing my own program I'd want to do what you suggest, 
> but I want to use the same approach used in the rest of the GNOME 
> project.
>
>     -- Darin