What I am uncomfortable with is the inability to run Wayland on older video hardware. X11 will happily run with Vesa driver on older hardware which no longer has functional 3D accelerated drivers.
What is the way forward for the retro community to run a modern Wayland system on older hardware?
Hopefully someone more knowledgeable can chime in or correct me
If your Linux distribution is handling Mesa packages correctly, you will never lack OpenGL/Vulkan drivers.
The reason is that Mesa includes "software rendering" drivers for both OpenGL ("llvmpipe") and Vulkan ("lavapipe").
As the name(s) might suggest, they use LLVM to JIT shaders for your CPU (supporting SIMD up to AVX2, last I checked - although typical compositing shaders tend to get pattern-matched and replaced with plain `memcpy`s etc.).
So you should always be able to run a fully-featured Wayland desktop (albeit limited in performance by your CPU), on any unaccelerated framebuffer, nowadays (and I remember doing this even before Plasma 6 launched, it may be older than usable Wayland desktops tbh - the Mesa code sure is, but maybe distros hadn't always built those drivers?).
Software OpenGL rendering technically works, but is IME unusably slow for compositing. What does work okay is direct software rendering. It might work to configure OpenGL to hit all the fast paths in the software backend (which you might need to add first), but I'm really not sure if you can do it without at least some unnecessary data copying.
>What is the way forward for the retro community to run a modern Wayland system on older hardware?
I'm curious why that's something that should happen. Shouldn't the 'retro community' be happy with retro software and not expect cutting edge software to work on their hardware?
Software compositing is possible. Wayland doesn't strictly require hardware acceleration or operations that are hopelessly slow without hardware acceleration.
I don't think they're downplaying. Most users will be perfectly happy with the included KDE application for screenshots/screen capture (it is really fully featured, much better than the default options for Windows or macOS).
And they're not saying that no third party applications will work either, it is just that old screenshot applications that were created for X11 will need to be ported.
In the end that is not really much different from macOS dropping Rosetta in the next release of macOS (yes, I know there is a difference of effort). Old applications will stop working, applications that still get support will eventually support it (if they don't support Wayland already).
It's actually wild that they're trying to sell lack of basic desktop features as "security features". Try explaining to an average person that when they are playing a game and want to capture a cool moment to share on Discord with their friends that by design the window server they are using will not let them do that and it's "for their own good". Like God forbid I want my computer to be actually usable. We had the same situation in the past when Wayland lacked a way to lock in the cursor within a window and it took them like 8 years to develop a "protocol" or whatever they call it. Imagine that for each basic desktop usability/UX feature, we will have to wait another decade. At that rate Linux will never have widespread adoption on desktop. Wayland sometimes feels like an active effort to sabotage Linux on desktop.
I'd say it's completely irresponsible to not have such features, claiming it's about security, rather than implementing it behind a permission system that puts the end-user in charge.
It is implemented behind a permission system. KScreenshot works perfectly fine and so will most of the applications using PipeWire I guess, same that screen captures.
KDE is merely saying that some applications will have to be updated to use it so all of the current screenshot applications won't work out of the box.
No idea of why some commenters here are implying screenshots don't work in Wayland. It seems their knowledge is somehow stuck at the first proof of concept ten years ago.
I was talking in general. It's less about screenshots and screen recording, but more about drag-and-drop and global hotkeys like push-to-talk in Discord.
Screen recording like screen sharing works perfectly fine with PipeWire.
Global hotkeys are also supported perfectly fine. Applications just need to register with the compositor, which will transfer the key press to them. That's the feature working at it should. It prevents applications from hijacking keys when that's not what you want.
The issue is that Discord takes ages to ship anything on Linux and barely supports anything and the Linux community does it's best to keep supporting everything. Other plateformes would have just mandated the new way ages ago and be done with the transition by now.
It's a huge hairball with no easy fixes, but at the same time, that's of significant benefit to some specific players. You can have a very usable X11 desktop with positively pre-Cambrian software. But to keep up with Wayland's ever evolving omnishambles, you basically have to run KDE or GNOME, or maybe Sway.
That is not what the protocol is like at all. Wayland can't do certain things not because they are difficult to do, but because Wayland was initially pathologically minimalist (I say that as somebody who generally likes minimalism) and the approval process for extensions has only improved slowly over time. It turns out that there's much more than blitting rectangles to a (single) screen and routing input events to a desktop environment.
If all the expected stuff was there, Sway (chosen because it's not tied to a desktop environmen) could be your X server equivalent.
If you're not familiar, the way X11 works is comparable to "multiplayer notepad" for your pixels ("multiplayer MSPaint"?).
All your monitors are combined into a large canvas where every pixel can be written and read by any X11 client.
Screenshots (and screen sharing) could be silently performed with zero user feedback (or any good way to even detect when apps might be doing this maliciously, AFAIK).
This is one of the big "security implications" that motivated Wayland (and somewhat similarly, the Flatpak sandbox and the XDG Portal infrastructure that has by now outgrown it).
The infrastructure is already there, for 3rd party apps to request these abilities (with the user getting the choice of following through, or denying the request), e.g.:
Keep in mind that any action which doesn't require interactivity every time (e.g. restarting the screensharing of a previously-chosen window/display) could have "user gave permission" be remembered, but that seamless case still only applies to that combination (so that client can't peek at anything else than what it was offered).
Anyway, what the blog post is talking about is really just 3rd party apps that haven't been updated (and e.g. might already not function properly inside Flatpak).
> feature parity
If we are being honest, screenshots/screensharing was never a "feature" of X11, it was a big security hole abused as a feature.
> The infrastructure is already there, for 3rd party apps to request these abilities (with the user getting the choice of following through, or denying the request), e.g.:
Why can't the infrastructure include the user simply allowing the app to continue to have access without a specialized infrastructure of app request (that requires updating the app)?
> really just 3rd party apps that haven't been updated
What's with the downplaying "just" again? Have all the best or most popular apps been updated to ensure no disruption?
> screenshots/screensharing was never a "feature" of X11, it was a big security hole abused as a feature.
So what was the screenshot feature of X11? Also classifying use as abuse isn't that honest, only the lack of security is
In case I wasn't clear enough, there was never a screenshot feature. Any X11 client could read any pixel written by any other X11 client.
> Have all the best or most popular apps been updated to ensure no disruption?
I am not familiar with DE-agnostic "screenshot apps" for Linux, they always seemed more common on other OSes, and I've always used the DE-specific apps (which were the first to support such mechanisms, some of them even using more direct DE-specific private protocols instead of XDG Portals).
But I spent a few seconds googling for general screenshot apps, found Flameshot (which makes sense as a cross-platform app), and it turns out that support for the XDG Portal approach was added to it almost 5 years ago:
And if you peek around the diff, you can tell that KDE/GNOME-specific support, on Wayland - using DBus but not the XDG Portals protocol - already existed, in early 2021, in fact...
That's the commit that added KDE/GNOME-specific Wayland screenshot support.
8 years ago, in a 3rd-party app!
> Why can't the infrastructure include the user simply allowing the app to continue to have access without a specialized infrastructure of app request (that requires updating the app)?
I'd forgotten that this happened, but for screensharing from a X11 client, someone already went through the trouble of emulating it (on top of the XDG Portals + PipeWire infrastructure):
It's only a temporary hack, and it only matters for X11 clients running under XWayland - if an app can run as a native Wayland client, it should have XDG Portals-based implementations of relevant features.
> What's with the downplaying "just" again? Have all the best or most popular apps been updated to ensure no disruption?
Am I downplaying, or are you describing a vague category of "the best or more popular apps" without giving examples?
I feel like it's too easy for some of this stuff to end up in FUD-like arguments without considering the objective reality (of how far we've come in the past few years etc.).
Anyway, my subjective take is that X11 took a decade or two too long to die, and most (if not all) gripes users might have with Wayland can be traced back to X11 outliving its UNIX Workstation origins and having never been designed as a Personal Computing graphical environment.
The 1996 extension had severe limitations. Untrusted clients have no clipboard, but also no GPU acceleration at all and other features were barely tested using it so it was somewhat random if they would work. It breaks a ton of applications and was therefore used by approximately no one.
Ok, so instead of a couple UAC style prompts for screen readers, macro recording, desktop sharing, etc, and some tweaks to GDK, we got what? An entire new backend GDK windowing system, and a pile of broken applications? And its been decades?
And its not like actual flaws people found couldn't be fixed.
Did you consider that maybe when you hold an opinion different than the people actually knowledgeable about a topic - like the people developing desktop environments and the former developers of X building Wayland - it might be because you are wrong and have a poor understanding of the field and not because they want to annoy you?
The flaws were not limited to the 1996 poor security extensions. These kind of half broken extensions are everywhere in X11. At some point, if the tweaks you have to do is basically rewriting the whole rendering pipeline and adding new APIs for the most significant systems, what you are doing is strictly équivalent to writing a new piece of software which is exactly what the people behind Wayland did.
And don't worry, the change adverse people you see here complaining about limitations fixed years ago would be complaining the same if the effort was on rewriting part of X11. That's life. Armchair complainers and keyboard warriors will complain while actual doers push things forward.
> Which is a load of FUD, the X11 security extensions from (checks google) 1996, restrict this.
Wait, what ? X11 has extensions ? As in can be "extended" ? And has the same thing since ( for the sake of dialogue) 1996 ?
That't why it must die. We need a monolith window system, with clear versions, all incompatible with each other. Only then, real progress can be made. /s
I love kde and it is what i use but still having the bug from time to time that the panels dissapear and have to relaunch plasmashell, also i wish they merged the virtual desktops and activities into one concept and allowed different wallpapers on each.
The theme settings is also confusing because of gtk apps, global theme etc. Feels everything around theming could be made nicer.
Prob some more nitpicks but overall it is a really great desktop environment.
> I love kde and it is what i use but still having the bug from time to time that the panels dissapear and have to relaunch plasmashell
I have that too, on Wayland. I also had a bug I mentioned recently in another comment with transparent terminals flickering, but that seems to depend on what is behind them and I think is a bug specific to Konsole.
Overall, its very close compared to a few years ago when several things were problematic with Wayland, but I do slightly feel its not quite there yet.
> i wish they merged the virtual desktops and activities into one concept and allowed different wallpapers on each.
> Prob some more nitpicks but overall it is a really great desktop environment.
Interesting, I always heard that x11 was not getting dropped until Plasma 7. While I am now fully on wayland myself, still not sure how to feel about this, feels a bit soon.
The x11 will be supported for 1.5 more years in latest state. And then there will still be support from LTS distribution for bug fixes. Only new features won't be there. It seems very glacial to me still.
> This is a perfect use case for long term support (LTS) distributions shipping older versions of Plasma. For example, AlmaLinux 9 includes the Plasma X11 session and will be supported until sometime in 2032.
Yeesh. Not optimistic about this. Fingers crossed but it's really frustrating when people feel this rush to move off the old way that's working in favor of a new way that doesn't yet do everything the old one does.
I love Wayland. I just absolutely love hard-rebooting my computer twice a day. Oh, need to walk away from the desk? Better switch into a TTY before unplugging the dock so I can force restart SDDM when it crashes. Oh no! Walked out of WiFi range? Your DE is hard frozen and the only operable button is the power button.
Wayland truly saves me so much time at work. It's forced me to re-up my ctrl+s reflex. Now I never lose work when it randomly crashes!
We've also been playing a fun new game. Every morning when I walk into the office my giant 5Kx1K monitor might be reset to literally any imaginable resolution and I have to figure out how to navigate to the display configuration menu to manually reset it because of course replugging the screen doesn't work.
I love Wayland. It's so easy to use and reliable. But MOST IMPORTANT it's newer than X11. Thank god I don't have to use gross old software written in uncool languages. God for-fucking-bid we have to run old software that works instead of new software that-- while not as good as the old software-- is still new
What's that? Wayland is almost 20 years old and still not to feature parity with any other OS? Well, it's newer than X11!
Nobody is developing x.org because it's a hot mess. It's just been half-baked solutions smacked on top of each other. That's why hardware acceleration barely works and artifacts are just a fact of life.
It's just that Wayland is newer, it's that x.org is a dead end and has been for many decades. If other people want to develop it, then by all means. But they're not, and ultimately I can't put a gun to someone's head and force them to work on a codebase. So here we are.
I use lemurs because as far as I can tell there's no good graphical display managers (lightdm may be the closest as SDDM doesn't show battery level and GDM doesn't allow changing the mouse cursor theme).
Bluetooth is still broken beyond belief and here we are with another post from KDE about Wayland. If Wayland will be as buggy as the KDE bluetooth implementation, it’s an easy pass.
What is the way forward for the retro community to run a modern Wayland system on older hardware?
Hopefully someone more knowledgeable can chime in or correct me
reply