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

Re: Re xkb-data with support for Mac keyboards into unstable

On Thu, Sep 28, 2006 at 09:36:07PM +0900, Charles Plessy wrote:
> Bonjour Denis,
> I have downloaded xkb-data 0.8-14 and played with commands such as:
> setxkbmap -rules "xorg" -model "macintosh" -layout "jp,ca" -variant ",multi" -option compose:rwin -option grp:lwin_toggle

Excellent, this is a very good example of what is possible with these
new layouts ;)

But there is one problem though: Japanese and Canadian layouts are the
most complex ones, they had been initially written using 2 groups.
Both have been splitted into 2 variants in X.Org.  Try:
  $ setxkbmap -model macintosh -layout jp -option -print
  xkb_keymap {
        xkb_keycodes  { include "macintosh+aliases(qwerty)"     };
        xkb_types     { include "complete+numpad(mac)"  };
        xkb_compat    { include "complete+japan"        };
        xkb_symbols   { include "pc(pc105)+jp(latin)+jp:2+inet(apple)" };
        xkb_geometry  { include "macintosh(macintosh)"  };
When jp layout is selected, 2 variants are in fact selected: first one
is jp(latin), second one is jp, which defaults to jp(jp106).
As you can imagine, jp(latin) is a slightly modified "us" version.

It is likely that Japanese people do not use the second group, but
instead an input method engine.  With this setup, I believe that
"eigo" and "kana" keys can be used to switch between those groups.

> I am using an iMac with a Apple A1048 keyboard.
> (see http://charles.plessy.org/mac/A1048.html or
> http://developer.apple.com/technotes/tn/images/tn1152_001.gif for
> something similar in spirit)
> This is a Japanese keyboard, and the default in Debian seems to make it
> output japanese syllabic characters (katakanas). While it seems to make
> sense, I have to say that I have never seen a japanese person inputing
> japanese that way. I think that it would really make sense to make it
> output ascii characters by default.

You can then replace -variant ",multi" by -variant "latin,multi"
But I have no idea why jp(jp106) is the default and not jp(latin), this
should certainly be fixed.

> Nevertheless, I checked that the input and output were corrsponding, and
> found the following discrepancies:
> - The "ro (ろ)" key (code 211) does not input a JIS underscore when
>   pressed in combination with shift.
> - The "mu (む)" key (code 35) outputs "\" or "|" depending wether it is
>   pressed with shift or not, as if it were from keyboard made for the
>   USA.

No idea for these ones :(

> - The comma key of the numeric pad (code 134) does not work at all.

There is indeed no definition for this keycode.

> - Also, what is printed on the keyboard is hiragana, not katakanas (they
>   have a 1 to 1 correspondance, but their apparance differs, for
>   instance ro and mu are ロ and ム in katakana.

It seems that one can switch between hiragana and katakanas, but I do
not know exactly how.


Reply to: