GUI frameworks should implement this, just like any app built on GTK, Qt, Iced or possibly others have native wayland support.
But yes I agree this is not a good situation. There should be something like “accessibility permission” on Android, where apps can basically read anything.
That’s one of the huge problems with Wayland. The core protocol is super minimalistic so it falls to each and every individual app to (re)implement everything: accessibility, clipboard, keyboard, mouse, compositing etc. etc.
The fact this was done in the name of security is a solution looking for a problem. Inter-window communication was never a stringent security issue on Linux.
It’s like advising people to wear helmets in their everyday life. Sure, in theory it’s a great idea and would greatly benefit those who slip and fall, or a flower pot falls on their heads, or are in a car accident and so on. But in practice it would be a huge inconvenience 99.99% of the time.
The largest part of all Linux apps out there will never get around to (re)implementing all this basic functionality just to deal with a 0.01% chance of security issues. Wherever convenience fights security, convenience wins. Wayland will either come around or become a bubble where 99% of Linux userland doesn’t work.
it falls to each and every individual app to (re)implement everything: accessibility, clipboard, keyboard, mouse, compositing etc. etc.
I haven’t read so much nonsense packed in a single sentence in a while. No, apps don’t implement any of these things themselves. How the fuck would apps simultaneously “implement compositing themselves” and also neither have access to the “framebuffer” (which isn’t even the case on Xorg!) nor information about other windows on the screen?
Please, don’t rant about things you clearly don’t know anything about.
So rather than having one single app that deals with screen reading, it’s now down to every individual application to make accessibility a priority.
Huge retrograde step.
We can all agree that authors should all value accessibility, but we also all know that they won’t.
GUI frameworks should implement this, just like any app built on GTK, Qt, Iced or possibly others have native wayland support.
But yes I agree this is not a good situation. There should be something like “accessibility permission” on Android, where apps can basically read anything.
So because they won’t, those who need accessibility, will require x.org… forever?
That’s one of the huge problems with Wayland. The core protocol is super minimalistic so it falls to each and every individual app to (re)implement everything: accessibility, clipboard, keyboard, mouse, compositing etc. etc.
The fact this was done in the name of security is a solution looking for a problem. Inter-window communication was never a stringent security issue on Linux.
It’s like advising people to wear helmets in their everyday life. Sure, in theory it’s a great idea and would greatly benefit those who slip and fall, or a flower pot falls on their heads, or are in a car accident and so on. But in practice it would be a huge inconvenience 99.99% of the time.
The largest part of all Linux apps out there will never get around to (re)implementing all this basic functionality just to deal with a 0.01% chance of security issues. Wherever convenience fights security, convenience wins. Wayland will either come around or become a bubble where 99% of Linux userland doesn’t work.
I haven’t read so much nonsense packed in a single sentence in a while. No, apps don’t implement any of these things themselves. How the fuck would apps simultaneously “implement compositing themselves” and also neither have access to the “framebuffer” (which isn’t even the case on Xorg!) nor information about other windows on the screen?
Please, don’t rant about things you clearly don’t know anything about.