Revised OLPC XKB definitions

Bernardo Innocenti bernie at codewiz.org
Tue Jun 19 12:13:43 PDT 2007


Sergey Udaltsov wrote:

> The patch looks clean and reasonably ok to me (especially symbols/inet
> part:). A couple of minor questions:
> 
> 1. You're using AB11 as a group switcher, quite frequently - would it
> make sense to put it as a separate option into symbols/group and
> include whenever you need it?
> 2. Why do you redefine AB11 for keycode 219 (this is the only point of
> having olpc keycodes file)? Is using plain I219 from evdev not
> suitable for you (especially taking the fact that in many cases that
> key is used as a rather special key, not just Shift or smth..

I'm also wondering why we couldn't just go on and use I219.

Actually, I cheated and removed these controversial bits from our
patchset:


diff -u -p -r1.4 evdev
--- keycodes/evdev      31 Aug 2006 21:44:13 -0000      1.4
+++ keycodes/evdev      18 Jun 2007 22:57:54 -0000
@@ -147,7 +147,8 @@ default xkb_keycodes "evdev" {
 
        <HZTG> =  93;   // Hankaku/Zenkakau toggle
        <HKTG> = 101;   // Hiragana/Katakana toggle
-       <AB11> = 211;   // backslash/underscore
+       // FIXME: conflicts with OLPC definition
+       //<AB11> = 211; // backslash/underscore
        <HENK> = 100;   // Henkan
        <MUHE> = 102;   // Muhenkan
        <AE13> = 132;   // Yen
@@ -242,7 +243,8 @@ default xkb_keycodes "evdev" {
        <I216> = 216;   // #define KEY_FASTFORWARD         208
        <I217> = 217;   // #define KEY_BASSBOOST           209
        <I218> = 218;   // #define KEY_PRINT               210
-       <I219> = 219;   // #define KEY_HP                  211
+       // FIXME: conflicts with OLPC definition
+       //<I219> = 219; // #define KEY_HP                  211
        <I220> = 220;   // #define KEY_CAMERA              212
        <I221> = 221;   // #define KEY_SOUND               213
        <I222> = 222;   // #define KEY_QUESTION            214


I'm Cc'ing Mark Foster, who knows better than me, but I have
a theory: the two keys we're removing are exactly 8 apart from
each other.  And evdev keycodes are skewed by 8 with respect to
the raw PC keyboard keycodes.

So, maybe in the beginning Mark didn't know of this obscure
shift and started using AB11 when in fact he just wanted the
211 *PC* keycode.

-- 
   // Bernardo Innocenti
 \X/  http://www.codewiz.org/



More information about the xorg mailing list