"Po Lu" <luangruo@yahoo.com> writes:

> Eli Zaretskii <eliz@gnu.org> writes:
>
>> I think Alexander says that the bindings are reversed, i.e. C-Δ
>> invokes the binding of C-δ and vice versa.  That's the bug he is
>> complaining about.  Your description seems to suggest that both keys
>> should have invoked the same command, not two different commands
>> exactly reversed from how they were bound.
>
> If so, I can't reproduce the bug by binding keys to the characters he
> specified in his e-mail.  Alexander, please send the lines in your XKB
> keymap which bind those characters, or alternatively run:
>
>   xkbcomp :0 test.xkb
>
> and attach the test.xkb file produced.  Thanks in advance.

The line for, for instance, ω and Ω is

    key <AB09> { [ w,                       W,
    EuroSign,                U03C9,                   9,
    U221E,                   U03A9,            U1F61F                 ]
    };

Here, lowercase Greek symbols are on the fourth layer and uppercase on
the seventh. I also included the test.xkb file that you mentioned. I
just had the thought that maybe the behavior can be explained by the way
higher layers are accessed through modifiers: I looked it up, and
usually the fourth layer is accessed through CapsLock and RightShift and
the seventh through CapsLock and RightAlt, but I made it so that the
fourth layer can also be accessed through Backslash and Grave, and if I
use either combination to type C-ω it acts as if I was typing C-Ω.