weston and hardware keyboard

Daniel Stone daniel at fooishbar.org
Mon Jun 17 09:18:43 PDT 2013

Hi Marc,

On 17 June 2013 17:08, Marc Chalain <marc.chalain at gmail.com> wrote:
> I already wrote an email on this mailing list about the use of xkbcommon for
> embedded solution.
> I had the same problem as Robert, and I disabled the check of the keymap.
> The problem is that we have number of keyboards as input device, on the same
> platform, and it's not possible to define a keymap for each one or a global
> one.

If you don't have keymaps, how are you supposed to interpret the keys?

> My first observation is we need a PC keyboard support at the end ( often a
> virtual keyboard).
> The second one is that we have to handle key from keyboards without keymap.
> The disabling of xkbcommon make the first integration on board easier, but
> make an incomplete integration.
> I think that a solution is to find a new way for hardware keyboard to add
> with the xkbcommon.

Hardware keyboards raise a great deal of issues, and even if they do
not require the use of xkbcommon specifically, they require an API
very similar to that of wl_keyboard.  For instance, if you have
multiple clients; focus on client A; press Caps Lock; switch to client
B - in this case, client B must be aware that Caps Lock is currently
pressed.  And so on, and so forth.

If you only need to support non-keyboard input devices, then in this
case disabling xkbcommon and removing wl_keyboard is a totally valid
solution.  But for hardware keyboards, I'm not sure that reinventing
the wheel helps in this case.


More information about the wayland-devel mailing list