by Hugo » Sat, 07 Apr 2001 18:03:51
We have 3 servers with US character set (default) enabled.
We need to use en_CA
How to get the good libraries, install them and make them work ?
Thanks for your help, it's really urgent...
_________________________________
Hugo Heitz
Societe Divento
tel : 01 42 21 01 08
fax : 01 40 26 28 13
24, Rue Du Mail
75002 PARIS
_________________________________
by Sven Maschec » Mon, 09 Apr 2001 07:52:24
"locale -a" will show you if you have installed it at all.
See also /usr/lib/locale/.
But perhaps you meant this with "good libraries"?
If you have no locales installed, "iso_8859_1" IMHO is always
available as a last resort, for LC_CTYPE only.
If you have it, what about just modifying /etc/default/init?
"en_CA" and other "*iso88591*" use a well supported characterset
anyway. But you might want to verify, that you're actually using
a latin1 font when having problems.
I don't know what more is changed when _installing_ a box with
the preferred locale, which according to some sun document also
makes it the system default. I guess it might affact X11/CDE,
but you talked about a server...
Sven
--
http://www.uni-ulm.de/~s_smasch/Locale/
1. Character set assigments (Re: How do I display IBM PC characters?)
[ Added comp.os.linux.developments.system as receiver because terminal driver
is part of kernel -- right? Added comp.terminals as receiver, because that
is terminal (or terminal emulation) issue. ]
?: I'm trying to figure out how to display IBM PC characters. I know it's
?: possible, but doing a simple printf() with the value gets it masked to
?: 7-bits, and when I tried ncurses, it put the wrong character up... Any
?: pointers to more info on this?
? Before you do your printf, print this: "\033(U" and it will switch
?to the DOS character set. "\033(B" will switch back. Or vice versa.
Just a comment (and some surprising notes :-))
These ESC ( U is guite odd code in standards view as
far I understands. ESC ( assigns bank G0. And bank G0 is on accessible
in range 128-255. That is GR (right side; characters (128)160-255) can newer
point to to bank G0. Only to banks G1-G3.
It should be more understandable if code is
ESC - A Assing Latin/1 (area (128)160-255) to G1
ESC - U Assign DOS character set (area 128-255) to G1
But it isn't that way :-)
And codes 'ESC ( U', 'ESC ) U', 'ESC * U' and 'ESC + U' have already another
standard meaning (see later).
(both ESC - and ESC ) assigns G1 -- charset names are different.
Hmm. ESC - can assign areas 160-255 (32-127),
ESC ( can assign area 161-254 (33-126) -- yeas these are very confusing.)
By to way -- from where that ident "U" comes for DOS character set?
Just curious.
Oops. Letter "U" is reserved for Latin-greek-1 (iso-ir-27) according
of RFC 1345 (that is informal RFC).
RFC 1345 lists following codes:
ESC ( U Assigns iso-ir-27 to G0
ESC ) U Assigns iso-ir-27 to G1
ESC * U Assigns iso-ir-27 to G2
ESC + U Assigns iso-ir-27 to G3
RFC 1345 don't list codes
ESC - U Assign {something} (160-255 (32-127)) to G1
ESC . U Assign {something} (160-255 (32-127)) to G2
ESC / U Assign {something} (160-255 (32-127)) to G3
[ Hmm. Perhaps I comment some other issues later. ]
3. How can I view eastern character set under Sun Solaris 2.6 x86
6. Reading is much more interesting than TV (0883/1708)
7. character sets on solaris 2.3
9. Solaris N/W flakey? and Asian character sets
11. Objet: Net pbs under Solaris 8
12. HELP: Non-Latin character sets under X
13. HELP! Cyrillic character set!