Wayland is flawed at its core and the community needs to talk about it(old.reddit.com)
old.reddit.com
Wayland is flawed at its core and the community needs to talk about it
https://old.reddit.com/r/linux/comments/1pxectw/wayland_is_flawed_at_its_core_and_the_community/
13 comments
I'm pretty sure there's going to be a bunch of comments here taking the article seriously just so the usual people can rant over Wayland more.
This is my first time learning about this, so it's not like I hate Wayland and want confirmation on it. But the comments did not refute his main claims. It's a debate on whether the restrictions are a good trade off, but it seems to me that what he said is factual mostly.
If the debate is on the trade offs then his entire argument collapses when he fails to highlight a single legitimate trade off. His examples were absolutely refuted.
I dont know how so many people read that thread and agreed with the OP. When I read through his post I thought that this person cannot even accurately identify what is a wayland issue and what is an application issue. He is looking at an application called xkill and getting mad that it doesnt work outside of x. Its like me getting mad that windows disk cleaner doesnt run on linux and blaming linux for it.
I dont know how so many people read that thread and agreed with the OP. When I read through his post I thought that this person cannot even accurately identify what is a wayland issue and what is an application issue. He is looking at an application called xkill and getting mad that it doesnt work outside of x. Its like me getting mad that windows disk cleaner doesnt run on linux and blaming linux for it.
Where exactly are the corrections?
This again feels like someone is confusing X11, the implementation, with X, the protocol, and makes an analogous inference that there is a single Wayland reference implementation and that everything Wayland is Wayland11, when what's really happening is that he chose a specific compositor, and the compositor just happens to implement Wayland.
The idea of a second implementation doesn't exist with X and this leads to a culture shock.
Each compositor is equivalent to an independent X11 implementation. This means most of the Wayland problems have nothing to do with Wayland and more to do with the immaturity of competing implementations. It's akin to blaming the HTTP protocol for YouTube changing the design again.
The primary complaint seems to be that compositors are struggling to catch up with all the new portals that are being created to support features outside of displaying window contents. You'd have thought that this is a good thing. It means you're being listened to.
There is a reason for the negativity though and the reason people are complaining is that they were taking X11 maintenance for granted, which had concentrated every single complaint we are hearing now into a single code base. Now that there are multiple competing compositors the free lunch is over.
The idea of a second implementation doesn't exist with X and this leads to a culture shock.
Each compositor is equivalent to an independent X11 implementation. This means most of the Wayland problems have nothing to do with Wayland and more to do with the immaturity of competing implementations. It's akin to blaming the HTTP protocol for YouTube changing the design again.
The primary complaint seems to be that compositors are struggling to catch up with all the new portals that are being created to support features outside of displaying window contents. You'd have thought that this is a good thing. It means you're being listened to.
There is a reason for the negativity though and the reason people are complaining is that they were taking X11 maintenance for granted, which had concentrated every single complaint we are hearing now into a single code base. Now that there are multiple competing compositors the free lunch is over.
The problem is that the "competing" compositors have to invent new protocols to do stuff that users need to do, but wayland has decided the users shouldn't want. So everyone comes up with their own protocols to create an actually functional desktop, but in incompatible ways.
The issue here is that this never changed for Linux.
> If I want an application to see other windows, that should be MY decision. If I want to run automation scripts, thats MY choice. If I want to accept a theoretical security risk in exchange for functionality I actually need, that should be up to ME.
They still can do all of those things if they wanted to. Unlike MacOS or Windows they can write and run whatever they want. The entire rant is just self entitled demands on on-source developers demanding that other people have to make the features they want.
Linux owes you nothing.
> If I want an application to see other windows, that should be MY decision. If I want to run automation scripts, thats MY choice. If I want to accept a theoretical security risk in exchange for functionality I actually need, that should be up to ME.
They still can do all of those things if they wanted to. Unlike MacOS or Windows they can write and run whatever they want. The entire rant is just self entitled demands on on-source developers demanding that other people have to make the features they want.
Linux owes you nothing.
No software is perfect, and if we really want to discuss flawed then IMO we should talk about systemd instead, which took one of UNIX' (and later on GNU/Linux) most famous core architecture principles - namely "do one thing and do it well" - and took it out to a kill site.
And yet they'll have to pry my OpenRC, cronie, and sysklogd from my cold dead hands...
And yet they'll have to pry my OpenRC, cronie, and sysklogd from my cold dead hands...
Systemd seems prime for a rethink like X11 has received with Wayland. I hope that systemd becomes the next target for cross-distro development and collaboration once Wayland has settled in a bit more.
This seems doable. SystemD is a bit heavy for embedded systems. Unit files provide a clean declarative syntax that a simpler implementation could accept. Admin commands and config syntax remain the same while the init process itself could be < 1% of the disk space.
Its funny that every single thing listed is incorrect and corrected by the comments, do you think OP will reflect on this and adjust the confidence of his opinion?