Re: How to setup the console to greek ?
Στις 29/Δεκ/2004, ημέρα Τετάρτη και ώρα 22:24, ο/η Michelle Konzack
έγραψε:
> # I am not on the List. Please CC me.
>
> Hello Simos,
>
> Thanks for the fast answer...
>
> Am 2004-12-29 21:48:07, schrieb Simos Xenitellis:
> >
> > Hi,
> > Have a look at
> > http://members.hellug.gr/djart/articles/grlinux/grlinux-4.html
>
> OK, I will check out.
>
> > Specifically, you need to execute
> >
> > $ echo -ne "\033(K\033)K" >/dev/tty
> > $ consolechars -f /usr/share/consolefonts/iso07.f16.psf.gz
> > $ loadkeys gr
>
> This give me a:
>
> Loading /usr/share/keymaps/i386/qwerty/gr.kmap.gz
> findfile(): timeout waiting for undead child(ren) ?
Hmm, it's a "known" issue. I though it got fixed by now. Debian GR
anyone?
Simply uncompress the file. The gr.kmap file calls other files and there
is some sort of race condition there. The same command (loadkeys gr)
works if the file is uncompressed.
>
> > for Debian Linux, and you are done.
> > It implies you use the system encoding ISO-8859-7. In addition, one can
>
> Right.
>
> > add these commands to the system files so that you do not need to run
> > manually.
>
> I run it from .bash_profile because it is a multiuser/
> multilanguage environement. - A killer under WOODY.
>
> > A rant about Unicode now.
>
> But 'mc' does not work with Unicode
There has been a patch circa June 2004. I do not use mc myself though,
so cannot tell about the quality of this. You may google on "midnight
command utf-8" for more.
> > Unfortunatelly, you cannot use Unicode while in console as you are
> > unable to write characters with accents by pressing a dead key
> > ("composing"). For example, for Greek, you cannot type ";"+"a" to
> > generate ?? when in Unicode mode. It's an issue that is being
> > investigated
> > (https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=143014) and
> > affects any language that needs dead keys. You would be able to input
>
> I know... because I use "de-latin1"...
> "de_de@euro" works fine but de_DE.utf8 drives me crazy.
>
> > Unicode if you would patch your kernel and recompile. See bugzilla
> > report for more.
>
> Hm thats interesting.
> Why it is not standard in the Kernel ?
The patch is a real hack, so the maintainers are not happy to include.
Check out the thread on LKML (see Bugzilla links). I am not sure how
comfortable you are to recompile the kernel. The patch is really small.
> > You can view documents in Unicode on the console, documents with several
> > scripts such as Latin (all western), Cyrillic, Arabic and Hebrew. The
>
> I was using "LatArCyrHeb" as font with Unicode and have tried to filter
> in 'mutt' with "libfribidi" for Arabic... but it is the hell !!!
>
> It lakes the shaping.
Shaping? You mean in arabic where characters get joined together? I
suppose it's one of the limitations of the console. For Greece there are
Unicode codepoints for every combination of greek characters
with/without accents, so it's easy.
> > Linux console can show around 512 characters simultanously when in
> > Unicode mode and allows "console unicode fonts" where you can choose
> > which character sets you feel like using (package "kbd"). This
> > functionality is default to other distros, such as Fedora.
>
> Already tried and it was a thread on "debian-user-german".
> There is nobody which can help with latin/arabic/farsi.
I am not familiar with Debian. Does Debian support the "kbd" package as
other distros do? I think Debian supports "lct" (Linux Console Tools),
which are rather unmaintained and non-utf8 friendly.
The way to go is Unicode. In Fedora you simply touch the files
/etc/sysconfig/i18n and /etc/sysconfig/keyboard and the rest is
automatic. The console input/dead keys is problematic and unless someone
writes a proper Unicode input mechanism, it's not possible to go all
Unicode for both console and X.
Oh, you could vote for
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=143014
Regards,
Simos
Reply to: