Kubuntu finally removes support for X11 in new installs | Hacker News Reader

Kubuntu finally removes support for X11 in new installs

neowin.net

35 pointsbundie17 comments

And so it begins; the forced migration to Wayland brought about from users not willingly switching to a half-baked solution from their existing working X11 environment.

Curious timing with everyone starting to drop X11 couple of weeks after a fork of X11 was announced. Almost like the announcement put a fire under some behinds to quickly get rid off X11 before the fork starts eating Wayland's lunch.

> X11 before the fork starts eating Wayland's lunch.

I'm sitting on my porch just absolutely cackling. Y'all are hilarious.

Wayland verion 1.0 was released back in 2012, how is that a half-baked solution 13 years later?

The publishing of v1.0 of a protocol has little to do with the stability of software implementing that protocol. The software in question is in fact KWin, the KDE window manager. It is the most mature and featureful Wayland compositor, and the KDE devs have done a lot of work over the past few years to get it competitive with its X11 backend. However even they themselves still consider this a transitional phase: https://www.phoronix.com/news/KDE-Wayland-Is-The-Future

Also, time is in any case less relevant that developer hours. Were I feeling snarkier, I might point out that GNU/Hurd was first released in 1990 - perhaps we can finally switch to that...

Because it's unfinished and still missing crucial features 13 years later

I don't understand why anyone thinks this is a good thing. It's been THIRTEEN YEARS and Wayland is not even close to a viable replacement for X.

I'm still baffled that on a platform where fragmentation is a serious issue it was decided the best course of action is to have every DE implement it independently. We are going from every DE sharing certain settings and configs to them all doing it differently, maybe not even implenting certain aspects even on the same OS.

I just don't get it.

It's explained here: https://www.youtube.com/watch?v=RIctzAQOe44

The presenter (an X11 dev) argues that it was sorta already the case. That, nowadays, X11 WMs are thick, do much of the work themselves, don't use most of the X server's capabilities, and it's just a weird man in the middle between applications and the WM.

I don't have the experience to confirm or deny this, though

Yeah, this is the biggest headache: so many things have gone from "you can do this with basically any WM and DE" to "only works on GNOME, or KDE, or this obscure compositor, hope you don't want some combination where the intersection is zero!"

I'm mostly seeing this sentiment from people who have little to no experience using Wayland. I've been using Sway as my primary environment for 5+ years now and it's been working great. What are the actual specific features that you believe are missing? Copy/pasting works, screenshots work, video recording works, screensharing works, before you try to parrot some ancient stuff. Post-Sway Wayland (even if you don't use Sway itself) is really a whole different beast. Drew and emersion and co. really sped things up and a lot of new compositors have sprung up since then with similarly good or better support for things. Splitting out wl-roots from Sway so that other compositors didn't need to build as much from scratch probably helped, but some compositors are still doing their own thing successfully enough. Hyprland seems to be pretty popular these days. Even Pewdiepie is using it, last I heard. They even added a tearing protocol for people who think they need to have tearing to get just a bit more performance out of their games and such. On the note of games, the Steam Deck's Game Mode uses Valve's Wayland compositor, Gamescope, and has been greatly successful as well. Some people are even running Gamescope on top of other environments on desktops because of how well games work in it.

The only thing I've heard isn't working in the last year or two is stuff related to accessibility for blind users. Screen readers and related software aren't working as well as they should yet. I don't mean to dismiss that as unserious, by the way, though I would wager that the vast majority of people don't use/need those features and should've been fine to switch over years ago. I hope accessibility improves ASAP as well for those that need it (and sorry if you're one of those people).

Maybe your use of Linux is limited to those things that happen to work. As far as I understand, Wayland is deliberately incomplete and a number of problems people have stem from design decisions.

Kicad, e.g., reports problems related to this. To quote https://www.kicad.org/blog/2025/06/KiCad-and-Wayland-Support...

> These problems exist because Wayland’s design omits basic functionality that desktop applications for X11, Windows and macOS have relied on for decades—things like being able to position windows or warp the mouse cursor. This functionality was omitted by design, not oversight.

Wayland might work for your or my personal selection of work cases, and currently my stance is "when it's done it's done, let's wait and see". But it might also end Linux support for entire classes of software and use cases.

We recently swapped RPIs with Windows mini-PCs due to xdotool disappearing, and whatever the alternative is called... segfaulting. Keep going like this.

Does Wayland actually work reliably now? That would be a nice surprise.

Hey Xorg, your code is like, the worst in the entire galaxy spanning the entire history of mankind will you let contributors fix it?

"No" -Xorg

Hey Xorg, it's 2025 and people have really powerful GPUs so fancy-smanchy effects can be implemented that use like 1% of a GPU's horsepower but would have been considered impossible fantasy beyond the power of every supercomputer in the world combined when you were written, can you implement something to handle stuff like that?

"No" -Xorg

Hey Xorg, your code is so insecure that any instance of it should be considered a critical vulnerability of the highest severity level will you let more than the dozen or so people you have work on it?

"No" -Xorg

Hey Xorg why do my windows get all jaggedy and messed up when I wiggle that around?

"No, uh, I mean, that's called screen tearing and it's a feature" -Xorg

Hey Xorg I have three monitors and I want to run them at different resolutions, refresh rates, and fractional scaling levels. Can I do tha..

"What the hell kind of frame buffer do you have that lets you do that?" -Xorg

Hey Xorg I was going to ask if I could do that without touching any config files or using the command line because I ain't got time for that. Wait. Frame buffer? Is this 1993? Are you developing on a single headed TurboGX-equipped Sun SparcStation running the SPARC port of Linux and you have absolutely no clue about either the state of the art or the march of progress?

"Yes" -Xorg

"Wait, why are people abandoning me?" -Xorg

My understanding is that most of those issues stem from Xorg's core design not really being compatible with modern hardware and the only way to fix it requires breaking all clients.

> not really being compatible with modern hardware

Yes. That's correct. You are correct. It is no longer fit for purpose and hasn't been for a long time. But inertia.

> the only way to fix it requires breaking all clients

Do it.

Rip the OpenVMS, SVR4, and HP/UX roots out of the ground and throw them in the trash, next to telnet and SysV filesystem support in the kernel.

>> the only way to fix it requires breaking all clients

>Do it.

You get Wayland that way. (Wayland actually started out as X12)

Reply on news.ycombinator.com