Good to know. Guess I won’t be using wayland anytime soon.
There has recently been some work on implementing a “pure GTK3” Emacs: https://github.com/masm11/emacs
But other than that, I’m currently using Wayland and haven’t had any real issues with Emacs.
The argument here seems to be “emacs has a terminal UI and so it’s easy to make a voice reader for it”
I don’t disagree that text UIs can be good, but this does not feel at all unique to emacs…
Most terminal UIs are not programmable in the same way; they’re either “streams of bytes over stdio” or ncurses type stuff that tends to not be friendly to automation. Emacs has a high-level interaction-model/API with windows, major/minor modes, keymaps, buffers, and the minibuffer that the end user is intended to code against.
“It runs in a terminal” is a vast oversimplification; there’s a whole lot more going on here, even if the post doesn’t go into details describing it since it’s directed at Emacs users.
I am an emacs user (if only for a few months now). If emacspeak reads from a high-level middle layer instead of reading the text interface directly, that is interesting, but not clear from the post at all (since the post explicitly talks about running emacs in terminal mode as a reason why apps are built text-friendly).