Linux’es diversity has never been found in the large fundamental pieces of software. Instead it’s typically been found in the nooks and crannies between them. We’ve typically had one or several of those and most have used those. It’s the kind of diversity you find between evolutionary differences between the same species, not revolutionary differences.
There’s an increasing amount of wayland compositors, so I don’t think diversity goes away.
Additionally, hyprland supports plugins which can do most things an X.org window manager could do. E.g. there’s a plugin to support river’s window layout protocol, which allows for creating custom window layout generator.
Diversity doesn’t just vanish, it’s replaced by new possibilities, created by solid protocol specifications with multiple implementations.
Similarily, nixpkgs and other repos continue to grow, just like flathub does too. These projects aren’t killing diversity, they’re enabling it.
I would argue they are all the same since most are based on wlroots and if wlroots doesn’t support something neither does the “increasing amount of Wayland compositors”.
To devils advocate a little in general with this topic: For wider spread adoption, Linux kinda needs to adopt around more standards. If you put yourself in the shoes of the average windows or Mac (even iOS/Android) user; it’s an overall standardized experience.
Linux now, is mostly a choice of DE and package manager. I still absolutely want distros like arch and Gentoo to still exists as they are.
Windows and Mac don’t have standards; they’re single solitary stand alone monoliths. The user experience is the same in their walled gardens because they are the same, not because those systems embrace standards. In particular Microsoft’s lack of standards has been a point of pain for Linux and FOSS users for decades. Linux has actual standards and that is exactly why there is so much diversity. That diversity would have crumbled into chaos long ago if the Linux community did not embrace standards.
If Windows users had to deal with the dependency issues, it likely would’ve never taken off. That’s kind of the problem I’ve seen around various Linux distros, though I wager it’s gotten a lot better in recent years. For the record, I’ve been out of the Linux game for a good 6 years, and I barely ever boot up my computer much. I’m able to run my business completely off my phone (except tax season), and I haven’t made the earnest effort to get back into it due to time constraints.
Flatpak doesn’t conform to the XDG home directory, and that upsets me. Also we have an ongoing dispute between SI and IEC units on their GitHub. But I like it otherwise.
It is by the juice of distro that thoughts acquire speed, the fingers acquire stains. The stains become a warning. It is by will alone I set my rig in motion.
Flatpak is good for diversity. Users don’t need to worry about whether the obscure distro they want to use has the software they want in its repos. If a distro supports flatpak it will work with most popular software out of the box.
Having run PostmarketOS on an old Samsung Galaxy tablet and now Arch on PineTab 2, Flatpak often works better than the native package manager. Especially with Wayland, many packages just work including touchscreen.
I may be misunderstanding flatpack, though I do understand the draw of all dependencies in one package.
One of the big things that drew me to linux some years ago was “oh, you don’t have to reinstall every dependency 101 times in a packaged exe so the system stays much smaller?” As well as in-place updates without a restart. It resulted in things being much much less bloated, or maybe that was just placebo.
Linux seems to be going in the flatpack direction which seems to just be turning it into a windows-like system. That and nix-like systems where everything is containerized and restarting is the only thing that applies updates seems to be negating those two big benefits.
Honestly anything that doesn’t get ported to wayland is probably old enough that it doesn’t really make sense to use as your primary desktop anyway. The most niche DE I regularly use is NsCDE, but it’s entirely FVWM scripts and FVWM is planning on adding wayland support. It’ll be a little sad to lose things like Trinity, WindowMaker, and Afterstep, but they were never amazing anyway and either way I doubt X will actually be unusable for a long time still.
I miss bspwm, none of the Wayland compositors work quite the same. Hyprland is close, but it’s just not quite as good. I moved to Wayland for the security benefits, but I miss X11/bspwm.
The worst part is there’s no standardization around screenshots/screen sharing/etc. so every DE/WM in Wayland has to be supported separately, or implement wlroots; which restricts how the software can be written.
I like the way standardisation is going, everything is going to be on the bee standard and that that isnt being updated too well too bad. What seperates us from the windows users is we can evolve if ya look at the distro tree it looks a lot like natrual selection to me
What’s so much better about Wayland than X? I mean, I’m not really a fan of X and the security nightmare that it is, but as a user it’s all pretty plug and play these days. What does a normal user get out of Wayland? Would they even know they’re using it?
I’d love to try it, but it currently won’t work with some software I use, so I haven’t bothered… And honestly I’m kind of confused about how everybody is talking about how amazing Wayland is (and how it seems to suddenly be the one true path for a bunch of distros) when my only experience with Wayland is people talking about how great it is and then not being able to screenshare or whatever… Which doesn’t make it seem great from the outside? That maybe sounds a bit flippant, but I genuinely don’t understand why “normal” people are so excited? I mean, I can see people caring about features like HDR and maybe that’s easier to build into Wayland than ancient X11, but I’d be more excited about the specific feature than Wayland itself which may make implementing these things easier?
Wayland cuts out all of the dead features and allows content to be drawn to the screen more directly. This means that there is a simplified architecture with great battery life.
Other than that, it doesn’t really bring much to the table currently. Not everyone needs (or wants) HDR and many of the other features that I would like to have are still in the works, so… I don’t really see a reason to use it, at least not now.
Support for HDR, variable refresh rate, direct draw and battery improvements sound like a very good list to have, other than the overall leaner build. You personally not caring about it doesn’t change the fact that it’s good to not stagnate when it comes to things like this.
Variable refresh rate has become the de facto standard of modern gaming now. They aren’t referring to the direct draw API, but the fact that Wayland does not have extra baggage to draw to the screen through a display server. Wayland just draws to the screen directly, saving time and performance.
First of all, X is not a security nightmare. There were 0 cases of someone getting hacked because of X exploit. It’s a FUD.
Now Wayland is a fad (haha). It’s not that much better than X and when it was drafted 10 years ago everyone just ignored it. Over the decade it became clear that X is stuck and at some point it will become obsolete so people started looking at alternatives and Wayland started getting some traction. Over time different tools started getting Wayland support, some people started getting exited about it and a kind of new meme developed where using Wayland meant that you’re ahead of everyone else (just like using Arch BTW). In the end it’s just a nice PR stunt. Ask people what specifically is so great about Wayland and they will mention some obscure features most people don’t need and features that it will have ‘soon’. In the long term the move will hopefully be a good thing but as of now if you don’t specifically need the few features it has you can keep ignoring it.
It is not a 'fad". Major distros have defaulted to Wayland (Ubuntu, Fedora, Red Hat, Debian, Manjaro etc).
X11 is old and designed for use cases in the 1980s. A lot of features have gradually moved out of X11 into the kernel or into other compositor systems. But the core X11 system is still limited by legacy design decisions and needing work arounds (which are complex to build and maintain).
Wayland is built to be the modern system that is built for current usage and needs. A lot of the benefits are not immediately obvious to the end user - a desktop is a desktop. But desktop interface projects like KDE who build user interfaces are hitting X11s limitations all the time, and a lot of effort goes in to working around X11s limits compared to working with Wayland. Effort spent working to work around X11 is time and work that could have been spent elsewhere on other fixes or new features and innovations.
The push to Wayland is deliberate and necessary, but was not always inevitable. Now that it’s being adopted so widely as the default by big distros and projects it is likely inevitable. It has essentially reached critical mass.
I think a lot of people asking “what’s the point” are not the ones working to build systems and distros at the back end. It’s easy for us as end users to take for granted all the work behind the scenes that make our desktops “just work”. But if you’re a volunteer building a compositor fit for 2024, I can see why it’d be frustrating working around the limitations of a system built for 1984.
X11 has served us incredibly well and is a hugely important project. But Wayland is the way forward.
While I don’t think X11 is great, I do not think wayland compositor is made to be easier to develop with. Wlroots had to be made to make things easier for compositor devs.
Sorry, I used the term “fad” to make a pun on X flaws being a ‘FUD’ (haha). It’s not a fad in the sense that it will soon disappear. What I meant is that the excitement around it is not funded in actual benefits and it just recently became fashionable to support it.
When Wayland can do and run everything X11 can, without problems, plus everything it promisses it can do, then I’ll make the switch. Till that time comes, I’m sorry, but it’s just not for me 🤷.
There are some really major deficiencies in Xorg that aren’t present in Wayland. The main one that made me switch was proper support for variable refresh rate, and the ability to mix and match any fixed or variable refresh rate displays you want.
It’s a super common use case to have a primary monitor with high refresh rate and VRR, plus one or two cheaper monitors that don’t. Xorg doesn’t really support that at all without some really hokey tricks that severely impede usability.
Proper sync support is another one. Yes, you can set tearfree in X but the implementation is crap. You’ll still get tearing in a lot of programs and at least in my experience, it introduces a pretty significant and perceptible input lag, far more than needed to eliminate tearing.
It’s a super common use case to have a primary monitor with high refresh rate and VRR, plus one or two cheaper monitors that don’t. Xorg doesn’t really support that at all without some really hokey tricks that severely impede usability.
I wish Wayland shills would stop spreading this lie. It literally just works. In fact, I’m doing it now on my laptop with a 144Hz 1080p monitor, and an external 60Hz 1440p monitor connected with Thunderbolt, with a dual-GPU setup (iGPU + nVidia, which Wayland doesn’t properly support, yet this is nVidia’s fault somehow even though Wayland compositors run entirely in user space, without interacting with the driver directly).
With VRR? Xorg definitely did not support this as of a year or so ago without running a separate xorg screen for each monitor which prevents you from doing stuff like moving windows between your displays.
Mixed refresh rates worked okay-ish but VRR definitely did not work well in multi monitor setups.
That’s why it doesn’t make sense arguing about it with Wayland fans. They always find this one obscure feature that X is missing and then claim it’s absolutely essential for everyone to have it. Most people have just one monitor, two equal/similar monitors, a handheld device with one screen or (and that’s the vast majority) simply don’t give a fuck that one of their monitors is working on a lower refresh rate. I’m glad Wayland finally found some traction with gamers obsessed with those things and is being adopted but the constant BS about everyone needing it is getting boring.
Mixed VRR is not an obscure feature for one. Most of my friends with gaming rigs have a primary monitor with VRR and use their old fixed rate monitors as secondary displays. Does it make a massive difference to run fixed refresh rate? No but it is noticeable and nice to have. Windows can do it and I paid for the hardware. Without parity on this kind of stuff, Linux is a hard sell to the people who do care about it.
Does it matter to Joe Schmoe? Probably not, but Joe Schmoe probably doesn’t care about Linux to begin with. You have to go for the tech enthusiasts first before you can get it to the masses.
1.6% of gamers use Linux. 25% of developers use Linux. Typical tech enthusiast is not gamer. Just because in your bubble people use VRR doesn’t mean it’s important to majority of users. Most Linux users don’t care.
1.6% of gamers is still millions of people. Entire industries exist on the back of much smaller customer bases than that. Might as well say we should stop caring about desktop linux completely since the server market dwarfs it.
I’m not saying we should just ignore it. I’m saying that it took the time it took (a decade) for Wayland to become a thing because most people don’t need it. Some people do and it’s not getting traction but most people can still safely ignore it.
Here’s the sad truth that Wayland haters hate: Wayland is way more performant and streamlined. X11 is an overly patched mess.
Everytime I had to install a distro, EVERYTIME I had to do some textfile hacking to avoid screen tearing with X11. Turns out in Wayland that is a virtually impossible bug.
Forget about making touchscreens work properly in X11, specially with a secondary screen.
I also remember all the weird bugs that appear in X11 when you have 2 screens with different scaling. No issue at all with Wayland.
Pretty basic stuff in any modern setup.
Wayland performs perfectly on platforms like KDE Plasma or Gnome. I miss no feature. It just requires that some propietary apps realise its potential. And that is what is already happening and will happen throughout 2024.
What does it do on new hardware? Not a lot of people are running normal desktop Linux on phones / tablets, are they? Which, totally cool if it works better on those things… but I guess I’m just surprised by how much hype there is for Wayland when X just works for me and would presumably just work for most people’s use cases. Like… who are all of these people that are emotionally invested in display servers, and what am I missing?
I mean, 20 years ago or whatever there was always the pain of black screens and X configs… but it just kind of works now in my experience?
Well that’s confusing because the meme is complaint text with Hulk saying that he sees this as an absolute win and you titled the post “I don’t” which means you don’t see this as absolute win and therefore agree with complaint text in the image.
Killing is overly dramatic, but it’s putting a burden on certain projects if they want to convert to it and not all have the resources to tank it. I don’t see Window Maker porting their toolkit to Wayland, for instance.
But XWayland exists so I don’t see what’s the fuss.
I don’t mind Wayland but I sure hope flatpack will not become the default way to distribute packages. Most packages I tried so far didn’t work. I just avoid it now.
That’s strange? I’ve never come across a single broken Flatpak across multiple computers with Linux installed. Do you have examples of broken Flatpaks?
X11 is already dead, and it will not become more or less usable it will always stay the way it’s and wayland will gwt better. that’s the difference and flatpak is just an option it doesn’t try to replace what’s already availible. spreading distrust and miainformation about these softwares doesn’t help
How do you mean that? I’ve been using X11 for like 17 years. i3 uses X11, and I will most likely not use another WM if I can help it. It’s perfect for me. X11 is available in the core repositories of all the big distros.
Just because they don’t do full releases doesn’t mean it isn’t developed anymore. They switched to updating modules individually, with three updates made this month. Doesn’t sound very abandoned to me.
I actually used Sway for a while. Can’t remember why I switched back though. What would X11 “going end-of-life” entail? Not being distributed/packaged anymore? Is there an official timeline for that or something?
Essentially they’re not doing feature work on the core codebase. I’m not sure if that’s true or not, but the packaging of it wouldn’t be up to the developers but the distro maintainers.
I’ll say that while it still has features that Wayland doesn’t it’s not dead, it doesn’t get updates yes but it still used by a lot of people for the fact that Wayland just doesn’t support some stuff that x11 does. Great example I have is TeamViewer and Nvidia+KDE
While TeamViewer is definitely neglected I use it often on Wayland and it works well actually!
In the past year or so it doesn’t shut down correctly. But the core functionality works well.
I’ve been experimenting with Rustdesk as an alternative because I doubt they’ll update the Linux client anytime soon. The Windows version looks like an entirely different application at this point
In terms of feature parity. I believe the only thing left is global hotkeys, which hyprland proved it can be done.
after a brief glance it looks like it, but that’s the same case as x over ssh. otherwise there’s things like vnc that wouldn’t care what each side is running
It’s such a niche feature that I bet most people under 30yrs never heard or used that it’s become too cringey that everyone keeps mentioning it.
But there’s the solution already mentioned.
I’d just like that some people would look a bit at themselves and realize that almost nobody wants or cares about that single weird feature. There are many remote desktop solutions more known to end users that need that kind of interaction.
Actual Unix users care. Maybe people that just jumped ship from Microsoft don’t, but I think that’s just because they don’t know what’s possible and how convenient it is.
Add comment