Re: Language and locale setting in /root/dbootstrap_setting

[Petter Reinholdtsen <pere@hungry.com>]

[Junichi Uekawa]
> Er... no, it will cause problems, if it was used for anything that
> is not supported by the console by default.
> 
> I don't think that many languages are supported by the default console.
> That kind of attitude is found pretty often in latin-charset people.

Eh, I think you misunderstood me.  I believe my patch is a step in the
right direction, not the ultimate fix.  It is only activated if the
installer passes a LANG variable.  This is not done by default, and
thus no one will per default run into problems if my patch is included
in base-config.  If someone want to use this functionality, we will
need to test if the locale is usable in the Linux console, and fix any
remaining problems.  I believe my patch will work correctly in some
cases where the console charset matches the charset in the locale, and
fail in other cases when the charsets are different.  But I do not
think this is a problem with my patch as such, just that it is missing
some functionality needed in some locales.  These locales do not work
during install at the moment anyway, so I do not see the problem
created.

To summarize:

 - installing using locale and translations do not work at the moment.

 - adding my patch make it work with some locales and translations
   using the same charset as the console

 - it is still broken for locales and translations using a different
   charset than the console

 - I believe my patch give a better starting point to fix
   installations with languages and locales using a different charset
   than the console

Fixing the install for non-latin charsets must be done by someone
understanding the special issues when using non-latin charsets, ie not
me. :-)

> Try checking for ISO-8859-1.  I think trying to display anything
> other than that might be problematic on some systems.

Even that will be problematic on serial console terminals.

How is the charset/font information in the langs/*.src files used?
There are references to charset, font and acm, but what is it and how
is it used?

Example:

<language name="Japanese" charset="euc-jp" font="LatArCyrHeb-16" acm="iso01">
    <name>Japanese</name>
    <hint>ja - ÆüËܸ줬ÁªÂò¤µ¤ì¤Þ¤·¤¿¡£¥¨¥ó¥¿¡¼¤ò²¡¤·¤Æ¤¯¤À¤µ¤¤</hint>
    <list>
        <name>¸À¸ì¤ÎÁªÂò</name>

        <item arch="i386 powerpc arm alpha" locale="ja" acm="iso01"
           font="LatArCyrHeb-16" keymap="i386/qwerty/us" msgcat="ja">
            <name>ÆüËÜ¸ì ±Ñ¸ì¥­¡¼ (QWERTY)</name>
        </item>
        <item arch="i386 powerpc arm alpha" locale="ja" acm="iso01"
           font="LatArCyrHeb-16" keymap="i386/qwerty/jp106" msgcat="ja">
            <name>ÆüËܸì 106¥­¡¼ (QWERTY)</name>
        </item>
    </list>
</language>

The locale part is by the way broken.  it should be 'ja_JA' or something
other valid variation.