[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [sorry] problems with the omsha cdroms



Hi,

At Wed, 9 May 2001 10:36:17 +0900,
JC Helary <helary@cerise.niji.or.jp> wrote:

> hehehe :-) the fact is that i did not set anything myself. if i
> remember well, i read the upgrade english info included on the
> cdrom where of course nothing is written about japanese locale
> (which does make sense). so i was extrememely surpsised to see
> all the mojibake. the apt process being extremely straightforward
> if you follow the instructions you absolutely have no worry. only
> when the upgrade was over (or seemed to be) did i use user-ja-conf.

You may have invoked dselect (or other scripts) under Japanese locale,
because your had already had Japanese configuration before the upgrade.


> you are right. sorry i was not more explicit. it seems that user-ja
> writes blocks of info starting with a #(beginning comments) line,
> intructions and ending with a #(end comment). the beginning of the
> block in .bashrc was not written on its own line but at the end of
> a line already present in the file. the rest of the block was
> correctly inserted. the lines before that in .baschrc were the
> alias section, and it seems the #(beginning comments) of the
> user-ja conf were not considered as a comments since all the
> lines below (that include the user-ja configuration) were considered
> as attempts at defining aliases. so the messages i had on the
> screen after logging in were taking all the configuration
> code lines one by one and saying that those were not aliases.

I see.  The #(beginning comments) line was written just after the
end of the last line (which did not have LF).  I will check whether
the current version of language-env still has the same trouble or not.
(I won't fix the trouble of potato version because potato is already
released and this trouble is not related to security.)


> i noticed as well that nvi-canna had the same problems in 2.2
> than in 2.1, maybe user-ja does not work well with it or maybe
> it is another problem. anyway, it does not seem to recognize
> the japanese input toggle. whatever i typed (^k, ^o, ^\, ^@ etc,
> i tried pretty much all the keyboard) it did not work. so anyway,
> i opted for jvim that works all right.

Hmm, strange.  Please check whether the alias (function) "vi" works
well or not.

function vi {
  if [ -x /usr/bin/nvi-m17n-canna ]
  then
    NEXINIT=`cat ~/.nexrc-m17n-canna` nvi-m17n-canna $*
  elif [ -x /usr/bin/nvi-m17n ]
  then
    NEXINIT=`cat ~/.nexrc-m17n` nvi-m17n $*
  else
    /usr/bin/vi $*
  fi
}


> in the case of mc it seems that depending on whether a japanese
> display is on or off i get the english version of mc or the japanese
> one. this is extremely smart and helpful.

Does mc itself have the feature?  It may be because of $TERM-dependent
setting of $LANG in .bashrc .  Translated messages based on gettext
are sensible to $LANG.  (more accurately, LC_MESSAGES locale category).

> now, is there a way to emulate the same behavior for lynx and mutt
> for ex ?

Lynx and mutt has to do much more i18n-related work than mc.  The encodings
of the contents of mail and web are independent from the environment
where lynx and mutt are to be invoked.  However, I think it is a good
idea to report a wishlist bug for lynx and mutt for such feature or to
discuss directly with upstream developers.


> is there a way to overcome the conflicts and set up a configuration
> depending on the display ? (ok, i suppose the answer is yes ;-)

Like this?  (though I didn't test)

function mutt-i18n {
  if [ "$LANG" = "ja_JP.ujis" -o "$LANG" = "ja_JP.eucJP" ]
  then
    mutt-ja
  else
    mutt
  fi
}


> i am a very poor student, and i accept book donations :-)

I guess your university has a library...


> my orignal idea was to start a (not too complex) project from scratch
> to understand what are the problems related to i18n/m17n on console.
> eventually a small text editor. if it is too big a project for a
> beginner i accept suggestions :-)

Very good.  If you develop a i18n/m17n console, it will certainly
make many people happy.  However, I suggest you to look for some
preceding projects, instead of researching from scratch.  For example,
please study "jfbterm".  I also heard that internationalized Debian
installer is recently being developed and it seems to be able to
display various languages on console.  I also suggest you to study
ISO 2022 and Unicode.  My document on i18n will help you to study
encodings.

And more, if your "i18n/m17n" includes East Asian languages, you will
need to study about doublewidth character.  To support Thai you will
need combining character.  Arab/Hebrew needs bi-direction.


> ps:(i forgot to turn kon on so i write in english... sorry) there
> seem to be a setonaikai linux user group, i subscribed to the list
> but it looks like there is no traffic. i was thinking of organizing
> a udonkuu taikai in my 'hometown' of takamatsu. anybody interested ?

Then I recommend you to announce your plan to some mailing lists and
news groups where many Japanese people join.  Unfortunately 
debian-japanese@lists.debian.org is not a good one for this purpose.
If your target is Debian users, I recommend debian-users@debian.or.jp .
I also recommend you to use fj.os.linux newsgroup.

---
Tomohiro KUBOTA <kubota@debian.org>
http://www.debian.or.jp/~kubota/
"Introduction to I18N"  http://www.debian.org/doc/manuals/intro-i18n/



Reply to: