On Tue, 2 Aug 2016 22:08:08 +0200
Silvan Jegen <s.jegen_AT_gmail.com> wrote:
> As far as I can tell, the goal of the Wayland devs is to keep the
> required protocols to a minimum and graduate prooven protocol
> extensions to official Wayland ones.
It sounds good on paper, but really turns out to be a horrible mess
in reality.
> So theoretically, as long as you implement the Wayland protocol (and
> it's assumptions) correctly, any compatible Wayland-speaking client
> should work just fine.
Yes, the clients are not the problem. We are talking about the
compositor here.
> Since Wayland is only a protocol, as long as both the client and the
> server follow it closely enough both the clients and the server will
> be happy. What is crucial is that the protocol is minimal and strictly
> defined however. I am still cautiously optimistic that this is and
> will be the case...
It's not only about client-server interaction, it's about how you for
instance should capture input in a compositor. You could use libinput,
or a gazillion other libs out there with different levels of device
support. I can already see the bug reports because this and that
joystick, touchpad, whatever does not work in a specific compositor.
And even clients have to do their own font-antialiasing. Sounds like a
lot of fun! Please stop repeating the propaganda spread on the web,
Wayland is not DoA without reason, and there is also a reason why
nobody uses it nowadays other than to play around with it. It's a
horrible mess and the wayland devs expect us to boil the ocean without
any clear benefits at hand.
Cheers
FRIGN
--
FRIGN <dev_AT_frign.de>
Received on Tue Aug 02 2016 - 22:16:06 CEST