From mboxrd@z Thu Jan  1 00:00:00 1970
From: Dan Cross <cross@math.psu.edu>
Message-Id: <200102060511.AAA21802@augusta.math.psu.edu>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] azerty [french] keyboard support
In-Reply-To: <038f01c08fa7$74fe30e0$0ab9c6d4@cybercable.fr>
References: <20010205144434.189C3199F3@mail.cse.psu.edu>
Cc:
Date: Tue,  6 Feb 2001 00:11:44 -0500
Topicbox-Message-UUID: 5a0bca3a-eac9-11e9-9e20-41e7f4b1d025

In article <038f01c08fa7$74fe30e0$0ab9c6d4@cybercable.fr> you write:
>i was of that opinion, but that would mean 3 zillion different
>keyboard types which would increase manufacturing costs.  all
>you they do (in general) is to stick the right set of keytops
>on and a scan-code/character table is set up to match the
>keytops.  you have to do the translation anyway and a table
>lookup costs nothing.  in general it's done by some black art.
>
>the fix charles sent me is small and elegant and should
>work for some subset of japanese keyboards i think.  if i
>remember, some have a shift, left shift and right shift keys
>(ie. 4 potential characters on each key).

This is exactly the sort of thing I'm talking about; making
a table driven translation machine driven by a little language.
What I don't like is the hardcoded tables in devcons.c.

I'll go out on a limb and say that I'd rather not see it in
the kernel at all, though.  Maybe a very basic kernel based
driver for basic I/O, but then hook into a user level program
to do all the real processing.  kbdfs, anyway?  :-)

>however, japanese is a more complicated case.  what i've seen
>is that you have a piece of code that sorts the kanji you use
>on a frequency basis, you type the stem of the prononciation
>and cycle through them till you get the one you want.

Ouch, that seems cumbersome, to put it mildly.  :-(  No offense
to the Japanese members of 9fans, but I'm glad I don't have to
type the language.  :-)

	- Dan C.