peter.hutterer at who-t.net
Wed Dec 17 14:58:08 PST 2008
On Wed, Dec 17, 2008 at 05:42:42PM +0200, Marius Gedminas wrote:
> On Wed, Dec 17, 2008 at 09:21:47PM +1100, Daniel Stone wrote:
> > Hi,
> > On Wed, Dec 17, 2008 at 12:02:34PM +0200, Marius Gedminas wrote:
> > > On Wed, Dec 17, 2008 at 12:04:16PM +1000, Peter Hutterer wrote:
> > > > To send that key from an input driver you'd have to init the key class and
> > > > send the keycode up to the DDX. But being > 255 the server would ignore it.
> > >
> > > Are there any plans about dealing with keycodes > 255? I keep noticing
> > > this problem in various contexts recently, and I'd like to know if
> > >
> > > * there's no known solution
> > Not for core-only clients.
> Are workarounds (e.g. allowing users to remap keycodes > 255 to keycodes
> < 255) possible, or too ugly to be allowed to see daylight?
the latter. You could get around it by assigning a 3 group xkb layout on the
device and then - in the driver - switch the group before sending a keycode
above 255. The theory then goes that for groups 3 and above you can do keycode
% 255 and it'll be mapped to the correct thing.
This however assumes that
- the driver knows about xkb and manages the keymaps accordingly. which it doesnt.
- the clients don't touch the xkb map for the device. which they do.
More information about the xorg