Possible UTF-8 CJK Regressions in Terminal Emulators

[Alexander Winston]

Using the emacs21-nox and mule-ucs packages from the official Debian Sid
repository, I am able to view all the UTF-8 characters possible with the
available fonts in both GNOME Terminal and XTerm. (This includes CJK.)

However, after checking out the source code from CVS revision HEAD
approximately twelve hours ago and building with a simple "sh configure
&& make bootstrap", I am unable to view UTF-8 CJK characters. UTF-8
glyphs outside of CJK appear to display finely right out of the box (no
configuration required), including mathematical, Latin-based, and
Arabic.

I have "export LANG=en_US.UTF-8" in all of my shell start-up scripts,
and locale has verified that en_US.UTF-8 has indeed been generated.
GNOME Terminal and XTerm have no problems viewing UTF-8 CJK when it is
viewed via cat, less, grep, etc. 

A PNG screen shot of the situation is available at <http://kvota.net/
misirlou/cjk.png>. It includes a side-by-side comparison with VIM's UTF-
8 handling of three files: <file://cvs.gnome.org/cvs/gnome/nautilus/po/
ja.po>, <file://cvs.gnome.org/cvs/gnome/nautilus/po/ko.po>, and <file://
cvs.gnome.org/cvs/gnome/nautilus/po/zh_CN.po>.

signature.asc
Description: This is a digitally signed message part