Hi, so as it seems revision 1.8 of the xf86-input-keyboard/src/bsd_kbd.c broke usual XkbLayout setup (like "us,ru" and such) in a way that it *always* picks whatever you have in the keyboard.encoding regardless of what is configured in the xorg.conf.
Could be that this chunk in the diff is an accidental removal? - if (xf86findOption(pInfo->options, "XkbLayout") != NULL) - return TRUE; piece of log: [974988.501] (**) Option "CoreKeyboard" [974988.501] (**) Keyboard0: always reports core events [974988.501] (**) Option "Protocol" "standard" [974988.501] (**) Keyboard0: Protocol: standard [974988.502] (--) Keyboard0: using wscons layout us [974988.502] (**) Option "XkbRules" "xorg" [974988.502] (**) Keyboard0: XkbRules: "xorg" [974988.502] (**) Option "XkbModel" "pc104" [974988.502] (**) Keyboard0: XkbModel: "pc104" [974988.502] (**) Option "XkbLayout" "us" [974988.502] (**) Keyboard0: XkbLayout: "us" [974988.502] (**) Option "XkbOptions" "grp:ctrl_shift_toggle" [974988.502] (**) Keyboard0: XkbOptions: "grp:ctrl_shift_toggle" [974988.502] (**) Option "CustomKeycodes" "off" [974988.502] (**) Keyboard0: CustomKeycodes disabled [974988.502] (II) XINPUT: Adding extended input device "Keyboard0" (type: KEYBOARD) Unfortunately I don't have a decent i386 machine at hand to compile this, but I'm willing to try a binary (I'm running the latest snapshot). Cheers, Mike
