The problems with having different keyboard
types attached to the machine and using them in raw mode seems to emphasize
that using unique keycodes for raw mode in the kernel might be a win.
I don't understand the problem. You ask for scancodes, you get them.
You ask for keycodes, you get those. What you want depends on the
application and is not a kernel matter.
What special purpose is attached to keycode 0 ?? (just interested, I
noticed that using scancode 0 -> keycode 0 is bad for loadkeys and
remapped the ADB scancode 0 (a) in the Mac driver).
Entry 0 on a keymap gives information about the allocation status
of the keymap. (Once we had 16 maps always allocated. Now we
allocate dynamically, so that most people have 6 or 7 maps,
but one can have 8 different modifier keys and 256 keymaps.)
For example, for the KDSKBENT ioctl, a value 0 means that the
corresponding keymap must be deallocated.
The question is: do we want to have the kernel pass unique, portable
keycodes also in raw mode?
In raw mode you get what you ask for: raw scan codes.
I'm not current on the difference between
raw and mediumraw mode - does mediumraw mode report up/down state?
Yes, it does. (That is the reason that there are only 127 keycodes,
and not 255.)
Any reason X wouldn't work in mediumraw mode, Geert?
No reason. But again, this is an X conversation and does not
belong on the kernel list. And since Linux and FreeBSD etc
all have a different system of keycodes, I can imagine that
the XFree people preferred to start with the raw scancodes.
Andries
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu