Re: Tracer TRK 21 MTS mm kbd - patches

On Mon, Jan 31, 2005 at 12:42:49PM +0100, Radoslaw Warowny wrote:
>Hi,
>
>I have tested the keyboard with xfree86 4.4.0, and the result is the
>0xE2 (226 decimal) scancode is passed as 189 keycode by
>xf86PostKbdEvent. Here is output of xev with example multimedia 1
>keypress events:
>
>KeyPress event, serial 26, synthetic NO, window 0x1e00001,
>    root 0x60, subw 0x0, time 277978, (502,508), root:(506,556),
>    state 0x0, keycode 189 (keysym 0x0, NoSymbol), same_screen YES,
>    XLookupString gives 0 bytes:  ""
>
>KeyRelease event, serial 26, synthetic NO, window 0x1e00001,
>    root 0x60, subw 0x0, time 277978, (502,508), root:(506,556),
>    state 0x0, keycode 189 (keysym 0x0, NoSymbol), same_screen YES,
>    XLookupString gives 0 bytes:  ""
>
>KeyPress event, serial 26, synthetic NO, window 0x1e00001,
>    root 0x60, subw 0x0, time 277983, (502,508), root:(506,556),
>    state 0x0, keycode 234 (keysym 0x0, NoSymbol), same_screen YES,
>    XLookupString gives 0 bytes:  ""
>
>KeyRelease event, serial 26, synthetic NO, window 0x1e00001,
>    root 0x60, subw 0x0, time 278088, (502,508), root:(506,556),
>    state 0x0, keycode 234 (keysym 0x0, NoSymbol), same_screen YES,
>    XLookupString gives 0 bytes:  ""
>
>KeyPress event, serial 26, synthetic NO, window 0x1e00001,
>    root 0x60, subw 0x0, time 278091, (502,508), root:(506,556),
>    state 0x0, keycode 189 (keysym 0x0, NoSymbol), same_screen YES,
>    XLookupString gives 0 bytes:  ""
>
>KeyRelease event, serial 26, synthetic NO, window 0x1e00001,
>    root 0x60, subw 0x0, time 278091, (502,508), root:(506,556),
>    state 0x0, keycode 189 (keysym 0x0, NoSymbol), same_screen YES,
>    XLookupString gives 0 bytes:  ""
>
>My proposition, for implementing support for this (an maybe other
>similiar) keyboard is to add an option to "InputDevice" section of
>configuration file XF86Config, that will turn on ignoring of first
>scancodes greater than 0xE1. This option could be named for example
>like: "StripFirstCodeToE1". The option could than be checked in
>xf86PostKbdEvent, and make it to return if first scancode is greater
>than E1. This solution should be feasable for XFree86 4.3, 4.4 and
>perhaps other versions. There is also a chance it will fix problems
>with other keyboards, similiar to mine, that sends additional
>scancodes (maybe higher even than E2) while pressing multimedia
>keys, that could be safely ignored.

That sounds like a reasonable approach to start with.  If this is something
that can be automatically enabled when needed it would be even better.
Do you have a patch that does what you suggest for 4.4?

I don't know if it there is a mechanism for dropping keycodes at the XKB
level.  If so, that might be even better still.

David