Hi,
I can regularly segfault xterm-158 with the following setup:
xrdb -remove
XMODIFIERS="@im=XIMSERVER" \
LC_CTYPE=xx_YY.UTF-8 \
xterm -fn -gnu-unifont-medium-r-normal--16-160-75-75-c-80-iso10646-1 \
-fw -gnu-unifont-medium-r-normal--16-160-75-75-c-80-iso10646-1 \
-fb -gnu-unifont-medium-r-normal--16-160-75-75-c-80-iso10646-1
(the gnu-unifont being the only free unicode font I have found so far containing
a large number of Chinese characters.)
XIMSERVER was one of | and xx_YY was one of
xcin-gb2312 | zh_CN
xcin-big5 | zh_TW
kinput2 | ja_JP
I have generated UTF-8 locales for the above language-country combinations
using (as Markus Kuhn suggested, thanx!)
localedef -c -i xx_YY -f UTF-8 /usr/lib/locale/xx_YY.UTF-8
LC_CTYPE=xx_YY.UTF-8 locale charmap
says, the charmap is UTF-8.
xterm will *not* segfault,
if the given XIMSERVER is not up and running
if Preedit-Type is set to "Root" (i.e. don't try OverTheSpot)
if xx_YY is a locale, for which no input method is available (e.g. en_US)
The segfault occurs in
XFontsOfFontSet() called in charproc.c:5009
with
term->screen.fs = 0x0
because
XCreateFontSet() called in charproc.c:5003
returned 0x0.
Does this happen to other people, too, or is it just me
doing something wrong?
[ I'm using Linux with X11R6 V4.1.0 and glibc 2.2.4 ]
And also, is it possible to display half-width/full-width characters correctly
using gnu-unifont and xterm in UTF-8 mode? (My experiments so far have either
resulted in overlapping Chinese characters or full-width Latin-1 characters.)
Cheers,
Kerstin
.
_______________________________________________
I18n mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/i18n
