[superkuh]

As an aside, you talk about wayland as if it were one thing. But the wayland protocol is intentionally minimal. Each wayland compositor picks and chooses between different third party libs to support various features. So you never know if something will actually work on the wayland compositor you use. If you stick within your ecosystem, yes, but it's not unified like X11 linux is. It's very fragmented and one's personal experience definitely doesn't say anything about other people's experience. Unlike with X11 where everyone uses the same thing.

For example, mouse and keyboard support and libei, libinput, or nothing (looking at you, weston). You never know what you're going to get and so applications that need to do basic keyboard/mouse things have to guess. It doesn't work all the time. In X11 it does.

Another example, accessibility features. The only wayland compositor that supports screen reading is GNOME's. They invented two new protocols, incompatible with all existing linux accessibility libraries. Only GNOME's wayland compositor and userspace applications use them.

So, in summary: one's experience can't be extrapolated with wayland because there is no single wayland.

[shmerl]

It makes sense for something like accessibility to be part of the protocol because it almost always needs access to stuff that Wayland restricts by design.

[Fire-Dragon-DoL]

Still with this situation,it will be maybe 10 years before we get accessibility again

[o11c]

Let's do the math. If each Wayland implementation supports an independent 95% of what users need, then:

* With 0 implementations, Wayland is good for 100.0% of users

* With 1 implementations, Wayland is good for 95.0% of users

* With 2 implementations, Wayland is good for 90.2% of users

* With 3 implementations, Wayland is good for 85.7% of users

* With 4 implementations, Wayland is good for 81.5% of users

* With 5 implementations, Wayland is good for 77.4% of users

* With 6 implementations, Wayland is good for 73.5% of users

* With 7 implementations, Wayland is good for 69.8% of users

* With 8 implementations, Wayland is good for 66.3% of users

* With 9 implementations, Wayland is good for 63.0% of users

* With 10 implementations, Wayland is good for 59.9% of users