Re: accented chars. shown as question marks in non-browser tools, sarge
Hi,
Thanks for the detailed suggestions and explanation. I can't test this
imemdiatley, but from what I can check, I think it's going to work.
On Mar 08 2007, H.S. wrote:
> Arlie Stephens wrote:
>
> >
> >I've got the same basic problem with just about every tool I use,
> >notably my email client mutt. Other versions of linux have somehow
>
> Do you use mutt in xterm? If so, it will be very difficult to get this
> right. Try this:
I do use xterms by preference. I also frequently read mail remotely,
using either putty (from a pc) or the xterm client of some other
system.
> 1. Use gnome-terminal or konsole (at least it out)
I'll try konsole; I don't use gnome. (If I wanted a windows XP lock
and feel, I could install windows XP ;-( though kde isn't much better
on that score.)
> 2. Make sure your locale is set to UTF-8 (see further below).
Is it reasonable to expect that random mailing list messages are UTF-8
these days?
I'm guessing that it is, because I believe PCs can't manage anything
else, and there's so many of them that they tend to set the common
standard.
> 3. Make sure the relevant language fonts are installed.
It looks like this is the problem. Not the fonts but the locales - the
only locales I have are 'C' and 'POSIX', at least on the system I use
most frequently.
Interestingly, another sarge system reports
$ set | grep LANG
LANG=en_US.UTF-8
LANGUAGE=en_US:en_GB:en
The obvious difference between the two is that the second system was
installed at sarge, whereas the first was unintentionally upgraded
from woody. (Debian's installer kindly put 'stable' rather than
'sarge' in /etc/apt/sources.list, and havoc resulted when sarge became
'stable'.) I've got a third system that was upgraded from woody to
sarge in a more controlled fashion, without having half of its
packages randomly uninstalled by berserk dependency checks, but I
can't check it remotely.
> Give the "set" command on a command prompt and see that you get for
> these variables (I have the following):
> $> set | grep LANG
> LANG=en_CA.UTF-8
> LANGUAGE=en_CA:en_US:en_GB:en
$ set | grep LANG
LANG=C
> If your LANG is not set to UTF-8, you need to set it right. Make sure
> you have the various locales install:
> $> locale -a
> C
> en_CA
> en_CA.iso88591
> en_CA.utf8
> en_GB
> en_GB.iso88591
> en_GB.iso885915
> en_GB.utf8
$ locale -a
C
POSIX
> If you do not a list similar to the one above, you need to generate the
> locale choices:
> <become root>
> #> dpkg-reconfigure locale
>
> (or is it locales? try both). It will give you an ncurses based list of
> locale choices to generate, Select the one you want and press OK. Next
> screen will ask you to set a default locale. Set a UTF-8 locale. After
> this configuration, logout and log in again. Use the "set" command again
> to make sure you have a UTF-8 locale now.
I'll try this at lunch. I'm logged in remotely, and much prefer to be
physically at the console when doing anything as root.
> Next, make sure you have relevant fonts install. I guess it is good to
> have the ttf-* fonts installed, at least the ones you think you might need.
>
> Finally, open gnome-terminal or konsole and fire up mutt. You should see
> various language characters in all their glory. BTW, xterm does not
> support UTF-8 properly yet.
Any idea whether putty can deal with it? It seems to me it could deal
with iso88591 (?) back when that got set up correctly more or less by
default.
> >Etch _claimed_ to default to UTF-8 - not my preference, but any
> >consistent and working setup is better than nothing - and I need to
> >check whether _that_ encoding actually works. (How can I find some
> >text that's definitely encoded in that format?) But what I want is the
> >ability to read anything. Well, anything in any European language,
> >with emphasis on french, german, and icelandic, plus a few related
> >dead languages.)
> >
> >Using a windows XP box for all my non-english language work is *not*
> >making me happy.
>
> True. I wanted the same thing. It turns out that making international
> languages work is easier in Linux (I have done this in Debian Etch, Sid
> and Ubuntu also) than in Windows! At least for me.
As far as I know, winblows forces one to use UTF-8, along with a
variety of special PC-only characters, like their "smart" quote - and
doesn't give the unfortunate user any control or insight. But it *is*
at least possible to use the system, out of the box, to communicate
with other similar systems.
Of course given my observations of differences between systems,
perhaps that's also true of debian - in the absence of a botched
upgrade from 'stable' to 'stable' ;-) - except for the problem of
using xterms.
--
Arlie
(Arlie Stephens arlie@worldash.org)
Reply to: