I am very open to Wayland in theory but I legit worry because it's been in the air for years and years and it still feels so far away. And some of the Wayland-y points around things like security just never felt relevant to me.
Though the real pain points for me have been around the various containerization things that just cause things like my IME to barely function.
At least Python 3 if you started your project on Python 3 you mostly could just get where you wanted to be.
I’m not a fan of Wayland or a detractor of X. I’m not personally invested in either of them. I couldn’t care less what’s pushing pixels at my screen as long as it works.
That said, what you need to know about Wayland is that the X.org devs mostly migrated to it en masse because they said X had hit an evolutionary dead end and couldn’t be dragged into the present, by the design of it. These were the people who were already maintaining X and presumably liked it. I doubt there were a lot of haters in that crowd.
It’s not like someone invented Wayland and shoved the X devs out of the way. The X devs largely became Wayland devs because they believe it’s the better path forward. I don’t really know how I could argue against them.
They're way too late now. The question is not about promotion anymore, just about when will the alternative to away. This page kinda made sense 2 years ago.
Without necessarily supporting or opposing the statements (because I have no opinion about Wayland after all), the first public mention of this website seems to be [1].
I have emailed with ProbonoPD, AKA Simon Peter. He is smart, motivated, and knows his stuff. He is the man behind Appimage (which I personally prefer to either Snap or Flatpak) as well as the "Hello System" FreeBSD distro.
As much as X11 is an overly complex and dated protocol, this article hits its mark well. My current desktop is actually running Wayland, but I still need X11 for a variety of reasons.
I see the development advantages of Wayland, but not the practical advantage as a user. And even as a developer, X11 is stable and well known (albeit definitely weird in places).
At the end of the day, things worked perfectly on X11 and my audio and video and various apps still glitch a lot on Wayland even after all these years. Most of that is not exactly Wayland's fault, but it highlights the advantage of X11. It's the devil you know (and everyone has worked out a lot of edge cases for).
The primary argument is to avoid fragmentation sprinkled with some other not-well-thought-through points. Distros aren't rolling back to X11. That ship has sailed. You can either help move things along, or drag your feet and increase fragmentation - the thing being complained about..
Everything under the "security" section of this article is so unbelievably wrong, that I'm not sure if it's worth anybody's time to refute one by one.
> Hardware Support: Nvidia users face particular challenges, and many specialized hardware configurations simply don't work reliably.
> The XLibre project continues this legacy with active community development.
Guess what? Nvidia users also "face particular challenges", and many normal hardware configurations simply don't work reliably under XLibre either. I guess Linus was right, don't use Nvidia.
I literally don't have the words to look down my nose at this. Former X11 contributor, compositor contributor, active distro maintainer. I just can't. I laughed. I scoffed. I incredulously looked at how much time someone with some braincells put into this.
I don't care, I'm saying it. Fuck your fragmentation whining. USERS WANT THEIR SHIT TO WORK. They expect high refresh monitors to work. They expect to be able to use a fking external monitor alongside their hidpi laptop. They need to be able to use actual fractional scaling, again with varying dpi monitors.
They want their shit to work reliably and not tear. Wayland delivers on this. Now. In a way X11 never EVER will.
God Jesus fuck i cannot believe how much people's time has been wasted on this bullshit useless banal conversation.
> I don't care, I'm saying it. Fuck your fragmentation whining. USERS WANT THEIR SHIT TO WORK.
...You understand that this is the strongest pro-X11 argument, right? As a user, I don't care that the backend is messy, I care that my stuff works on Xorg and doesn't work on Wayland.
This reads as significantly AI-assisted and repetitive.
And some of the points are really questionable, like claiming “Color management becomes inconsistent” when X11 doesn’t even support HDR, claiming that Linux containers and namespaces somehow mitigate X11’s security issues (that’s not how it works!), and talking about some outdated issues like NVIDIA support and SSH forwarding (both of which shouldn’t be an issue anymore from what I’ve heard).
A lot of the complaints are about supposedly missing features when X11 and Wayland aren't the same things, IIRC. X11 is a big ball of mud with all kinds of features. Wayland is an attempt to do one thing, and only one thing. You must bring in the other features (like window management and remote networking) as adjacent components.
"This Phillip's-head screwdriver can't cut cheese or open a wine bottle."
> Wayland is an attempt to do one thing, and only one thing. You must bring in the other features (like window management and remote networking) as adjacent components.
It's fine to separate a small core protocol from extensions that flesh out the rest of the features. Wayland, however, did the small core and didn't bother to actually implement the rest, so 16 years on it still doesn't have feature parity.
> "This Phillip's-head screwdriver can't cut cheese or open a wine bottle."
Then stop handing me a screwdriver and telling me I should use it instead of my multitool.
They mention macOS as "one platform with consistent APIs and behavior". But macOS (or NeXTStep, as it started out as), an actual UNIX system (certified!), chose not to use X11 to achieve that, but to entirely roll their own.
Unfortunate that this page largely shills XLibre, the X fork by the guy who got ejected from contributing to Xorg when he landed a bunch of code that was blatantly untested, leading to tons of his commits getting reverted. Like, I am not a fan of Wayland, still using X on my machine, but from what I've seen of his posting, this guy isn't great PR for the "X is fine actually" camp.
The ship sailed when Keith Packard, brilliant programmer and X.org project lead, decided that the era of X11 is over, and no further development will happen to the Xserver, and that the rootless Xserver for Wayland, in the near future, will be the only actually supported version.
This lead to the weird fork of Xlibre that is more about politics than it is about technological solutions.
Everyone has chosen to abandon the security nightmare that is X11, somehow it took 15 years instead of 5. But hey, we finally made it.
Accessibility is dead to me on Wayland... Until emulate key, mouse presses in other windows, registering global hotkeys that isn't a hack and Enumerating window title and executable.
Wayland fragmented development efforts on Linux. It's gotten a little better as some compositors consolidate, but still pretty awful to cobble together a stack that works for the features above.
If Wayland would have simply provided the API, but not the underlying implementation (compositors Could have handled type implementation). They could have sidestepped this whole issue of fragmentation.
What if someone doesn't want KDE or GNOME, or any tiling WM?
What if we want some of the other desktops out there?
Examples:
I like Macs and macOS, with a global menu bar. I like Ubuntu's Unity desktop. It doesn't work on Wayland and won't. Unity8, AKA Lomiri, does but it is intended for handhelds with touch and is badly compromised. No global menu bar, for instance.
What if we want something NeXT-like? I like GNUstep. Neither GSDE nor NEXTSPACE supports Wayland.
I like Xfce. It has an experimental version but keyboard window management is broken. Xfce's compatibility with standard Windows keystrokes is one of the reasons I like it. It seems to me Wayland folks don't know these and I've yet to see a Wayland environment that supports them. E.g. Alt+Space, X to maximise. Alt+Space, N to minimise.
I want full title bars, so I can use the scroll wheel on them to roll them up, as in Windowshade.
I also want to be able to middle-click the title bar to send that window to the back of the stack. Once _every_ Linux desktop did this, but GNOME started software rot by getting rid of title bars in favour of "client side decorations" and the feature started to disappear.
What if I want one of the lightweight Windows-alike desktops? MATE, Trinity, EDE, XPde? No Wayland support.
What if I want a Chinese desktop with a bit of Win7 bling? DDE and UKUI both don't support Wayland.
What if I want a different trad Unix environment?
CDE, nsCDE, Maxx Interative, OLWM... none work on Wayland.
You could use most of those arguments verbatim and unironically create a page titled "Please don't promote X11". Meanwhile, some of the Wayland-specific complaints were solved a while ago and some weren't even Wayland's fault to begin with.
I get it that some people don't want give up the old systems in favor of the shiny new technologies. I'm also on that side for some software like systemd (not debating. Systemd does offer many compelling benefits). But this article feels like complaining for the sake of it, rather than making reasonable detractions that could perhaps be addressed in Wayland. Attacking a software for just existing isn't a good strategy in FOSS. It is fundamentally about having choices.
I help (or administrate) a couple of elderly relatives with their Linux desktops. The systems use Fedora Kinoite (an older version which still runs X), a immutable distribution which updates itself without any necessary user action (flatpack and rpm-ostree). This setup is simply perfect.
From time to time I log into those machines to check if everything is ok (updates ran through, backups were made and so on) or at times where the relative needs help with something. I use anydesk or rustdesk, both work pretty great with X but don't with wayland. rustdesk has an experimental wayland mode but it's unusable choppy and slow. Anydesk doesn't work with wayland at all.
As mentioned above I run an older version of Fedora Kinoite because newer versions run with wayland only.
What are my options in this case? I don't care if the system runs X or wayland, it just have to support this use case.
33 comments
[ 4.0 ms ] story [ 60.4 ms ] threadThough the real pain points for me have been around the various containerization things that just cause things like my IME to barely function.
At least Python 3 if you started your project on Python 3 you mostly could just get where you wanted to be.
That said, what you need to know about Wayland is that the X.org devs mostly migrated to it en masse because they said X had hit an evolutionary dead end and couldn’t be dragged into the present, by the design of it. These were the people who were already maintaining X and presumably liked it. I doubt there were a lot of haters in that crowd.
It’s not like someone invented Wayland and shoved the X devs out of the way. The X devs largely became Wayland devs because they believe it’s the better path forward. I don’t really know how I could argue against them.
> Training Overhead
> $5,000-$15,000 per IT staff member for multi-compositor environment training
When's the last time your IT department gave you "multi-compositor training"?
> Software Replacement
> $500-$2000 in software replacement costs when tools don't work under Wayland
> Multi-Monitor
> X11: Reliable, consistent behavior
> Wayland: Compositor-dependent, often buggy
> Distro wars intensified: Choice of display server becomes a major differentiator
Definitely written by AI
[1] https://x.com/probonopd/status/1955387873659850828
"Think twice before abandoning X11. Wayland breaks everything!"
https://gist.github.com/probonopd/9feb7c20257af5dd915e3a9f2d...
This could be a web-ified version.
I have emailed with ProbonoPD, AKA Simon Peter. He is smart, motivated, and knows his stuff. He is the man behind Appimage (which I personally prefer to either Snap or Flatpak) as well as the "Hello System" FreeBSD distro.
I see the development advantages of Wayland, but not the practical advantage as a user. And even as a developer, X11 is stable and well known (albeit definitely weird in places).
At the end of the day, things worked perfectly on X11 and my audio and video and various apps still glitch a lot on Wayland even after all these years. Most of that is not exactly Wayland's fault, but it highlights the advantage of X11. It's the devil you know (and everyone has worked out a lot of edge cases for).
One guy.
Everything under the "security" section of this article is so unbelievably wrong, that I'm not sure if it's worth anybody's time to refute one by one.
> The XLibre project continues this legacy with active community development.
Guess what? Nvidia users also "face particular challenges", and many normal hardware configurations simply don't work reliably under XLibre either. I guess Linus was right, don't use Nvidia.
I don't care, I'm saying it. Fuck your fragmentation whining. USERS WANT THEIR SHIT TO WORK. They expect high refresh monitors to work. They expect to be able to use a fking external monitor alongside their hidpi laptop. They need to be able to use actual fractional scaling, again with varying dpi monitors.
They want their shit to work reliably and not tear. Wayland delivers on this. Now. In a way X11 never EVER will.
God Jesus fuck i cannot believe how much people's time has been wasted on this bullshit useless banal conversation.
...You understand that this is the strongest pro-X11 argument, right? As a user, I don't care that the backend is messy, I care that my stuff works on Xorg and doesn't work on Wayland.
And some of the points are really questionable, like claiming “Color management becomes inconsistent” when X11 doesn’t even support HDR, claiming that Linux containers and namespaces somehow mitigate X11’s security issues (that’s not how it works!), and talking about some outdated issues like NVIDIA support and SSH forwarding (both of which shouldn’t be an issue anymore from what I’ve heard).
"This Phillip's-head screwdriver can't cut cheese or open a wine bottle."
It's fine to separate a small core protocol from extensions that flesh out the rest of the features. Wayland, however, did the small core and didn't bother to actually implement the rest, so 16 years on it still doesn't have feature parity.
> "This Phillip's-head screwdriver can't cut cheese or open a wine bottle."
Then stop handing me a screwdriver and telling me I should use it instead of my multitool.
The ship sailed when Keith Packard, brilliant programmer and X.org project lead, decided that the era of X11 is over, and no further development will happen to the Xserver, and that the rootless Xserver for Wayland, in the near future, will be the only actually supported version.
This lead to the weird fork of Xlibre that is more about politics than it is about technological solutions.
Everyone has chosen to abandon the security nightmare that is X11, somehow it took 15 years instead of 5. But hey, we finally made it.
Wayland fragmented development efforts on Linux. It's gotten a little better as some compositors consolidate, but still pretty awful to cobble together a stack that works for the features above.
What if we want some of the other desktops out there?
Examples:
I like Macs and macOS, with a global menu bar. I like Ubuntu's Unity desktop. It doesn't work on Wayland and won't. Unity8, AKA Lomiri, does but it is intended for handhelds with touch and is badly compromised. No global menu bar, for instance.
What if we want something NeXT-like? I like GNUstep. Neither GSDE nor NEXTSPACE supports Wayland.
I like Xfce. It has an experimental version but keyboard window management is broken. Xfce's compatibility with standard Windows keystrokes is one of the reasons I like it. It seems to me Wayland folks don't know these and I've yet to see a Wayland environment that supports them. E.g. Alt+Space, X to maximise. Alt+Space, N to minimise.
I want full title bars, so I can use the scroll wheel on them to roll them up, as in Windowshade.
E.g. https://www.windowmizer.com/
I also want to be able to middle-click the title bar to send that window to the back of the stack. Once _every_ Linux desktop did this, but GNOME started software rot by getting rid of title bars in favour of "client side decorations" and the feature started to disappear.
What if I want one of the lightweight Windows-alike desktops? MATE, Trinity, EDE, XPde? No Wayland support.
What if I want a Chinese desktop with a bit of Win7 bling? DDE and UKUI both don't support Wayland.
What if I want a different trad Unix environment?
CDE, nsCDE, Maxx Interative, OLWM... none work on Wayland.
Amiga style: amiwm doesn't work.
Classic Mac style: mlvwm doesn't work.
I get it that some people don't want give up the old systems in favor of the shiny new technologies. I'm also on that side for some software like systemd (not debating. Systemd does offer many compelling benefits). But this article feels like complaining for the sake of it, rather than making reasonable detractions that could perhaps be addressed in Wayland. Attacking a software for just existing isn't a good strategy in FOSS. It is fundamentally about having choices.
The ability to run X applications over the network is crucial to me.
The real hazard is that you won't be able to get drivers for X11 and XLibre in the future, because only BSD would be using it.
From time to time I log into those machines to check if everything is ok (updates ran through, backups were made and so on) or at times where the relative needs help with something. I use anydesk or rustdesk, both work pretty great with X but don't with wayland. rustdesk has an experimental wayland mode but it's unusable choppy and slow. Anydesk doesn't work with wayland at all.
As mentioned above I run an older version of Fedora Kinoite because newer versions run with wayland only.
What are my options in this case? I don't care if the system runs X or wayland, it just have to support this use case.