On Fri, 11 Jul 2003, Wu Yongwei wrote:

> S***, it seems I made a mistake.  The font selection in Windows 2000 is not
> at all as flexible as Java; it's more like Linux.  Just that the default
> font in the Simplified Chinese version is still Tahoma instead of Song Ti.

 Thanks for checking that out. You saved me some tinkering :-)


> Jungshik must be right that I could change the default font in locale zh_CN
> to make ASCII characters appear nicer.

  With Gtk2 and fontconfig, I don't have to tinker with the  font
configuration as much as before because it looks all right to me.
As for CSS-style font list specification, the infrastructure is already
in place (fontconfig), but the 'UI' part needs some catch-up to do.
For instance, most GUI programs and window managers don't have UI to
let multiple (ordered-list of) fonts be specified (although it's
possible to do so by editing configuration files manually in _some_
cases.)


>  The only problem is that the
> standard locale for Simplified Chinese in Red Hat 8.0 (which I use) is
> zh_CN.GB18030.  I was told that it was possible to change that to
> zh_CN.UTF-8, but I did not find the motive/time to do that.

  It's rather easy. See
<https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=75829>.


> Regarding the 'A' APIs in Windows.  Do you mean that there should be some
> API to change the interpretation of strings in 'A' APIs (esp. regarding file
> names, etc.)?  If that were the case, the OS must speak Unicode in some form
> internally.

  Yes, that's what I meant.  Beni already gave some details.

Beni>  win2k does have the option of
Beni> witching the encoding used in the 'A' APIs, it's just global and
Beni>  requires a reboot.

 Yup, I frequently do to test Mozilla under different locales.
Having to reboot is really painful. On POSIX systems, we can just
run   a program under any supported locale at the command line. Under Win2k/XP,
'chcp' works inside a 'command prompt'  (even setlocale() works), but I haven't 
checked out
if there's 'SetACP' (the opposite of 'GetACP').


> remount the partition in an appropriate encoding; if it is on an EXT2/3

  As you found out, there's a tool or you can easily make one as many other
have done.  Once you switch to UTF-8 locale, there's no need to look back.

> partition or on a CD-ROM, then I am out of luck.  Maybe the mount tool
> should do something to handle this? :-)

   In case of CD-ROM, it's not much of an issue. See mount(8) man page and other
man pages referred there.

   Jungshik


P.S. A word of caution. A lot of _text-mode_ programs still assume that a single octet
takes a single screen 'cell', which holds for most legacy single byte and double
byte encodings. This assumption breaks down for UTF-8 and three byte sequences of
EUC-JP and four byte sequences of GB18030 (and eight byte sequences of EUC-KR).
Some of them are modified to cope with two-byte UTF-8 sequences (U+0100 - U+07FF),
but don't work with U+0800 and beyond. Needless to say, combining characters
are not handled in those programs.


--
Linux-UTF8:   i18n of Linux on all levels
Archive:      http://mail.nl.linux.org/linux-utf8/

Reply via email to