All those new things are great, but I've noticed that noone is realy thinking about accessibility anymore. Wayland, gtk4, flatpak, every version of qt... are barely usable with a screen reader. One reason why is it like that is probably the fact that orca (the only somewath good screen reader for linux) has only 1 fulltime maintainer... A lot of people using computers are blind or have low vision, so I think if linux improves its accessibility, we culd get more people to switch.
The problem is that most nobody cares that can do anything about it cares. Most open source voices and voice engines used on linux desktop are based on University projects that were abandoned and are only maintained by the few people that actually care about the blind or are blind themselves. Wayland isn't ready for prime time for people that can see, hear and have full mobility, complaining it's worse for the physically disabled while valid, is also pointless.
There's a lot of things that can happen in next 10 years. In 2012 Wayland was still a concept with some early implementations rather than anything functional and one might say that in 10 years it should be 100% complete and much better than X11, but to my observation, the whole graphics stack has transformed quite significantly. The development is pretty slow, but still 10 years will make a ton of difference. Current showstoppers doesn't seem impossible to fix, whether it's the NVIDIA, key binding protocol, color management or interface scaling issues. That's just my guess.
yeah but gtk5 is still gonna take a long time to come around, and wayland will get more support in enterprise for sure too, since it's redhat calling the shots in this and many other cases, sure the devs might talk about dropping support but if it's not ready, redhat won't ship it, just like fedora's plans to drop bios support got almost instantly abandoned
As someone who likes to be on the bleeding edge, I’ve been running Wayland on my main desktop since the release of Fedora 36 without any deal breaking issues (for me). I occasionally run into issues with screenshot tools, but for now I’m just using the built in tool from GNOME instead. But I’m all too aware that there are many different workflows right now that make Xorg a more viable option currently. Not to mention the huge amount of window managers and DEs only available for it.
Same here. I have encountered one or two issues but that's because I not only run Nvidia but it's on a laptop with 2 GPUs. Dual GPU laptops didn't get screen tearing fixes until like 2019. I'm a few days away from getting my first PC built tho. It's full AMD (except the Intel network card). It should have 105% Linux and Wayland compatibility :D Also, one of the best WMs (Qtile) has Wayland support. I think there are a couple more.
@@SirLancelotS I don't see how this is a display server issue rather than the Archie tool and file manager being the culprits here 🤔 Does the same setup work as expected under Xorg?
I've been running wayland with kde for months now and the only issue it's ever given me are context menus being out of place and some weird font issues with flatpaks. tbh if wayland didn't exist I might have never decided to daily drive linux because I have a 144hz monitor and 2 60hz monitors and xorg limiting my 144hz monitor to 60 is a pretty big turn off considering I game a lot
Ubuntu 22.04 and Fedora 36, two of the most used desktop distros, already shipped with Wayland as the default. Considering it's started in 2008, it already has more progress than IPv6 which started in 1995.
I have been using Sway for 6-7 months now and after the initial setup, it has been great! Many applications already support wayland natively, but the option just might not be enabled by default. For example, firefox supports wayland but doesn't enable support by default. You can use the wayland version by setting the MOZ_ENABLE_WAYLAND=1 environment variable.
As an XFCE user I will be stuck with xorg for a while. I don't mind though. Last time I checked, wayland was pretty bad for gaming, as there was no support for "unredirect fullscreen windows", which basically means the compositor is running at all times which adds input lag and lowers the frame rate
My only problem with wayland is that I don't have a viable replacement for my window manager, xmonad. Waymonad exists, but it's way to experimental and incomplete for daily use. I could switch to a fork of dwm, but that's a pain and I'm really used to the way xmonad works currently. I think that the current issue with wayland is just the ecosystem is decades less mature than xorg and doesn't have the laundry list of options that xorg has. This is an issue that can only be fixed with time, and when I have a wayland compositor that implements enough of my workflow, I'll be on it like butter on toast. If anyone knows of a good dynamic twm for wayland that I can get on Nixos, let me know, I'd love to try it.
dwl or river, dwl crashes often randomly for me(I have nvidia gpu), river I disliked because of the config file is basic using bash commands, then you would have to learn a programming language to then use your compositor.
In 7 or 8 year's time, I think it will be a very different story. I say this as a person who is going to use X11 for the foreseeable future (it's still super broken on my system).
@@miguelguthridge that'll probably be down to how many more spanners Nvidia throws in the works along the way. They finally caved on the GBM/EGLStreams thing, but Wayland is still unusable on the Nvidia blob for me.
I am actually trying out Wayland/Sway at the moment. There have been some stumbling blocks but all in all it seems to work. (with nvidia gtx-760) What I noticed is the screen output seems much more crisp using sway than it did using i3. Whether that is just my imagination I don't know. Of course I can't do without xorg-xwayland yet, as too many programs don't natively support Wayland.
I think this is a big advantage proprietary software has over open source. Once the top decides something needs to be done or changed, it is changed within a "small" amount of time. As far as I understand things like Wayland and pipewire are necessary for the future of the Linux desktop so I think it is a good thing that projects such as Fedora push through the changes even if it is still buggy.
openbsd already does maintain xorg somewhat, they have a project called xenocara which, although they claim is not a fork, seems to be effectively used to distribute the patches they need to xorg
It's mainly used to track the updates to Xorg, so if the current small amount of Xorg development falls out then they're left with a lot more work to do.
@@BrodieRobertson ah rip. I'm sure they'd pick it up as they've picked up huge projects prior but it would probably be at the state it is now if not worse, only occasionally pushing new drivers, fixing bugs/security issues and basically just maintaining it in its current state. I am honestly not counting on OpenBSD porting their existing WMs to Wayland and I expect they will hold out on support unless it is absolutely necessary (ie everybody else has moved on). In addition they are, for the most part, a server/embeded OS anyways so it's not the worst thing if their graphic software is outdated.
@@hostgrady OpenBSD IS a desktop OS. I don't understand why people always think it's primarily a server OS. The OpenBSD devs use OpenBSD on the desktop as a daily driver. They care about the desktop experience and won't drop it. It's going to be interesting to see what they will do. I think Theo himself still uses fvwm, which is probably going to be impossible to port to wayland. OpenBSDs own cwm would be easy to reimplement using wlroots.
I tried using wayland. I really, really tried, but I just can’t. My X11 setup is just too comfy no wayland setup comes close. Most my software just doesn’t work properly on wayland and the amount of visual customisation you can do on wayland compositors is severely limited compared to X11.
Until Wayland allows an app like a keyboard remapper to know which app window has the keyboard focus it is less than useless to me. Right now there is no standardized API for this on Wayland. On X11/Xorg it’s very easy to see things like the WM_CLASS and WM_NAME of the focused window and apply a specific keyboard shortcut overlay just for that app.
@@JahidulIslam Not in my experience, KDE has been working well for me on Wayland (Intel chipset). Even using its screen capture tool (Spectacle), had no issues. Gnome on the other hand I've had crashes.
If GTK5 will get mainstream a lot of years from now, like 8-10 years in the future, dropping X11 won't be a problem, Wayland will get mainstream and overtake X11 in the most popular DEs until that time. On Gnome, Wayland is already 80% ready if you are using Intel iGPU or AMD GPU/iGPU.
I've been maining wayland on Gnome for about a year now and I've rarely had issues. The only thing that's currently still broken is discord's / Team's screen share. (I’ve been on wayland ever since I got my 6900xt. My desktop experience with nvidia was horrendous)
Pierre wasn't talking about remote desktop software, he was talking about X forwarding. X forwarding is far, far more flexible than a VNC session or similar. To use an example from my work, I often have clients from several different remote machines all rendering on my computer under my window manager. There are several other users on these same systems, all running clients under their own logins at the same time. VNC simply cannot do that, it only does one session at a time, and even with something that can do multiple sessions like NoMachine, you have to run a full DE on each remote host in a window instead of just rendering the client(s) you need on your local X server. There is no way to run a Wayland client remotely, as far as I'm aware, so at best you're stuck with something like NoMachine that can at least do multiple sessions, and at worst you're stuck with the projects you showed off, which can only handle one user at a time.
I believe Waypipe is considered to be the de-facto Wayland equivalent to X application forwarding (not making any comments about its effectiveness or quality one way or the other; haven't tried it myself but have heard and read about it). Also, when full-DE remote desktop is required, I'm very happy that FreeRDP exists as a much nicer alternative to VNC on Wayland. I've been using the regular gnome-desktop-sharing daemon provided by GNOME which serves an ms-rd:// endpoint for connecting with standard Microsoft RDP client apps. It works pretty well for my use cases.
The other day I was installing a Linux distro on a virtual machine. The default was Wayland. I wasn't able to see the mouse pointer until I switched to Xorg. I'm not against Wayland, but it seems like it still has issues that needs to be resolved. They can't just simply say, "It works on my computer so it's fine." They must think of the millions of Linux users out there with different hardware and use cases.
KDE and Nvidia. When Wayland works flawlessly with these two, I'll switch. Occasionally I will try Wayland with KDE and it never works. I am not going out and buying a AMD card to replace my RTX 3080 Ti just to run Wayland. For embedded systems like the head unit in cars, Nvidia does support Wayland. Maybe now that Nividia open-sourced their drivers, maybe in the next few years things will be improved. As for now, it is Xorg. Wayland has yet to become a real boy.
My Fedora install came with Wayland enabled by default (Desktop: Ryzen 3600x, RX 580, Laptop: Thinkpad T480, no Nvidia graphics just integrated Intel). This is my first daily driver Linux desktop in years. I haven't noticed anything broken yet, everything seems to "just work". Even the few games I have tried seems to work well even the Hogwarts Legacy one.
A lot will need to be done for this sort of major shift to be worthwhile. If we can't get Wayland to be at least as feature-complete as Xorg, and at least as stable, and at least as accessible without losing stability, which is not fulfilled in either cases currently, then it won't be a worthwhile shift. Forcing it on users is not going to be worthwhile either as it undermines the premise of an open system. Currently it's too unstable, and isn't as compatible as it should be. It also isn't super great on older systems or Nvidia systems. If all of its current issues are resolved by the time GTK5 rolls around, and it remains backwards-compatible, then it'll be absolutely worthwhile. Currently, it is not. I find it quite concerning that Gnome is considering app exclusivity, which also goes against the open ethos. There's quite a bit of writing on the wall that needs to be addressed.
It has nothing to do with "app exclusivity" because its all about maintaining two parallel incompatible versions of their libraries, sometimes wholly separated in the case of static libs, this is a massive waste that you need to spend money and developers maintaining and wastes twice as much time developing any singular feature in the optimal case. (and 90% of cases are sub-optimal so generally expect around 350% slowdown for any functionality or solution) Its not like GTK3 or GTK4 will be dead in the sense of being completely unusable, it won't be actively maintained but you could still use it, it just means GTK5 features not seen in GTK3 and GTK4 will be exclusive to Wayland which 90% of the time isn't gonna change anything, if someone needs to make a GTK app they can just use GTK4, its extremely unlikely they'll need GTK5 features and since GTK is FOSS anyway if someone wants to bring certain GTK5 features to GTK4's Xorg implementation then they can fork GTK4 and do it themselves. I don't see how this is a problem. Do you expect GTK to work on every Windowing System all the time? That expectation is insane, you're hefting work on people for something that has no value, any scrub or company who built their own window system would still be excluded from GTK, X10 is also not supported by the more modern versions because what would the point be? X11 has been around for 40 years, most people who use computers now, even on Linux, weren't alive or weren't capable of doing anything with a computer when X11 was released, so why would you expect anyone to use it?
@@Spartan322 Well said and true to the point. I used both other the years what they see as features I see as vulnerabilities. They say well my screen capture is broken. Maybe that program doesn't work because they didn't want other program to screen capture private screens. Gnome screen capture toys work fine. If screen capture is harder on Wayland Go Wayland is what I say. one of the top reason I sopped using windows was they install the Xbox stream bar into windows making setting up screen capture to easy and hack-able. That and many more reasons. Even Nvidia are pushing Wayland drivers on now hardware, it will always be the old hardware that sticks to Xorg on new I have had no issue's on AMD or Intel, only shitty old Nvidia whose driver the last time I used my GTX1070 was crap. I'm selling that card soon as outside of Windows Gaming they are a crappy company to work with.
I tried going "full wayland" before sdl2 announcement about support for wayland. It's too early, and still wayland is buggy. Crashes on bookmarks, only sway was good (as wm) but i love dwm and i always miss so badly when i try alternatives. Maybe if suckless will release dwm, dmenu for wayland i will try again or idk. Replacements (some are just pain compared to simpler x11 counterparts), setting everything from scratch just to hit the wall is kinda pain and waste of time for now. Xorg/xenocara is stable like rock, i think everyone needs time before switching. Wayland have pros, like no need for additional stuff but im not ready yet. Anyone who made all these switch guides, big respect, thats a lot. For me, im waiting for ' the right time '.
It's already possible to use Wayland applications on X11 by running Weston in a window. This would be a solved problem if a better integrated compatibility layer was built using Weston, as a sort of counterpart to xorg-xwayland.
Linux is not ready for a wayland only desktop just yet.. bunch of current projects just aint there yet to name a couple barrier/synergy. maybe this threat from GTK will put a fire under them to do so but there are WAY more projects that are NOT ready yet also
Still using Xorg-X11 here, as 1. Wayland applications don't run over SSH, 2. my UI scripts depend on Xephyr, xdotool and xwd, and 3. my screen cat (oneko) does not work on wayland!
i mean the problem with X11 is that people are comfortable staying on it. so nobody really tries to move to Wayland seriously. If GTK5 forces everyone's hand - there will be more people getting around to fix Wayland. so, in this regard, i see this as a good idea.
@@ewetoo unless they are willing to maintain gtk4, they will have to jump ship eventually. either that or someone will have to maintain GTK5 non-wayland backend.
This switch is inevitable, but it is unlikely to be in 10 years, unless nvidia fixes their drives in 3 years max. I would think that projects the size of gtk5 do not decide to drop support for such an important part of the project in the middle of development with no guarantee that the nvidia users would be able to run wayland.
I disagree all wayland problems are not that major to not be fixed in the next 10 years. Both Gnome and Kde are pushing wayland, some of the biggest distros are already pushing wayland. I don't think that everyone will be on wayland in 10 years, but it's likely that the majority of users will use wayland.
@@ANGELRA the point here is not that those problems can't be fixed in 10 years. It is that it probably will be fixed in 10 years, but gtk5 is unlikely to abandon x11, if wayland is not rock solid in 3-4 years, because they can't just bet in the middle of development cycle that it would be rock solid by the time gtk5 is released
As a window manager only user this worries me, as I don't want to have to switch to another window manager. I've tried sway and qtile with Wayland, they both missed a lot of features I use, I could not get to change the keyboard layout or set the compose key, unable to set scaling to be larger, etc. And I'd also need to change a lot of my programs that are Xorg only.
I remember when Plasma 5 had come out and FreeBSD still only had a super old version of Plasma 4 and it was a total pain to install. Very interesting how it's starting to keep up with Linux technologies like Wayland
While I’ve been mostly happy with wayland and sway, it has issues with displaying submenus for VSTs I use for OBS filters/in my DAW, which is really annoying, so this is hopefully good news in that people will be incentivized to patch issues
Wayland is not fully ready although its ready at their side the window manager sides have not fully adopted and integrated with the separate api system of wayland for some more intricate settings
Wayland works in my Manjaro KDE (recent) with Nvidia 750 Ti, that I am going to replace with a 3060 Ti, as prices went down, and AMD still is not on par for gaming, but I do not know why, writing it is slow, and waydroid do not work, so I switched back to X11.
I'm using Wayland on my ASUS TUF GAMING F17 with an NVIDIA RTX 3050 Ti and I don't see any issues. So...It seems to be fine for me. Dunno why Ubuntu doesn't have that session as the default and Xorg as a backup.
While I think I can't comment on that, however, I have an Intel HD laptop, and nvidia within my PC, my laptop is running definitely worse, for example frame drops happen frequently on wayland than on x11, the wayland is less efficient than x11. This is a big surprise but this is the experience I have with Intel. With nvidia nothing works however, the desktop is definitely smoother compared to x11, but that's it. I just want to express that as of things right now I don't see huge benefits going wayland only, you have to definitely consider the compatibility issues. Although like I said I can't stand on either decision since I didn't use wayland on it's glory (AMD gpu).
I feel the pain on both sides, both wayland and gtk developers, even I have been in this situation. When people are using old software and you're writing something new (in their opinion, better) but the users, still use the old software. Dropping support left and right and forcing GTK or nothing type of mentality is not going to cut it. There is a hard thuth to swallow: if your program is good, people are going to use it, you don't have to threathen anybody or drop support for anything, people will eventually switch because they like it.
Seems a bit early but these tickets tend to stick arround forever so i guess its a good goal to have. Honestly the best way to get everyone onto wayland is to have a good backwards compatibility to old xorg apps. I know the devs are not thrilled about doing that but they will just have to suck it up if they don't want the fragmentation period to last forever. Get my xorg apps to run good enough on wayland and i will switch tomorrow.... once everyone is over new apps will natuerally be wayland but if they refuse we will have another python2 vs python3 situation...
I really want to use Wayland but I can't. Discord cannot stream on Wayland (if you a beginner and I want to recommend it to beginners). My some games not works well on Wayland. I.e.: on Doom 2016 I take 200 fps (it's a limit). But on Wayland I take 80-110 fps. I'll really wait until it's ready. I know waiting works everytime. First time I use Linux, I can't even use my gtx 1650 maxq GPU. But now I can use every hardware on linux.
Honestly, I'd use Wayland across the board if Wayland didn't have problems on NVIDIA (Gamescope not working, and assorted problems thanks to XWayland) and KDE (flickering bugs, vertical panels not working properly unless I sign out, and other things), and if FreeSync/GSync was supported on both NVIDIA and AMD/Intel. I feel like Fedora dropping X11 support altogether on their main branch on NVIDIA GPUs is a bit early right now. It's a bit of a mess, and I have to use X11 for now on my NVIDIA system. Looking forward to when Wayland gets support for HDR displays and GSync.
It took 10 years for them to move from GTK3 to GTK4 so by the time GTK5 comes, you probably don't have any reason to use X11 anymore. And Variable Refresh Rates work on Wayland
If I can get my hands on an Intel ARC card, I'd love to benchmark Wayland on it. My 1060 hangs after an hour in a Wayland session across multiple different DEs. Not knowledgeable enough to figure out why, but its happened on both Fedora and Ubuntu. I had a laptop years back that had Intel graphics and the framerate jumped when I switched to Wayland from Xorg, and because I've seen what Wayland CAN do, I actually am not against the idea of switching over years ahead once these kinks get worked out.
Lol! I didn't even knew that gtk5 is here. And to be honest, I have never seen a single application written in gtk4. Most of them are written in gtk3 and many are using gtk2
There just isn't enough push behind Wayland. Don't get me wrong, I'm not saying X get's the support it needs, but Wayland is nowhere near ready and at this pace of development I doubt it will be ready by 2030. Unless you get someone like Steam backing it I doubt it will be able to replace the X11 in the next 10-15 years, maybe more. When there is an interest, be it personal from a highly taletned technical person or corporate from a corporate agent that has the money to hire talented devs open source develops at a breakneck pace, but when those interests are lacking open source tends to develop at a snails pace.
The push only exists behind Wayland at this point, all of the corporate linux projects are focused on that. X11 isn't completely going away for a very long time, Steam for example is still going to rely on Xwayland or an Xwayland like tech for games that rely on Xorg but you don't need to be running Xorg directly on metal to acheive that.
@@BrodieRobertson Well, see, you say the push only exists behind wayland, yet at the glacial pace of development the push might as well not be there. Take the kernel as a counter example. If Intel or Samsung or Microsoft need a new function in the kernel they'll have it in the shortest amount of time possible. Wayland on the other hand doesn't really have that kind of push behind it. Not to offend the people working on Wayland but the best and brightest aren't being payed to develop it. The original X11 server dates back to 1987, while X dates back to 1984. In just 3 years doctoral students at MIT made the basis of what X has been for the past 35 years. Compared to the work that was done in that period, what the X.Org foundation has added over the past 20 years is paltry. Had wayland been university project 10 years ago it would have already supplanted X.Org. And that's because those doctoral students have something to prove. And outside uni, the kind of people that could make Wayland work in just a few years are getting payed a lot of money to work on other things corporations think are more important than a linux desktop renderer. The push behind wayland might as well not be there considering how little there is to show. I'm a solus user, a while back the looser in charge - Josh Strobl - quit. Why? Because he asked for something to be done and the person he asked didn't do it. Because apparently Josh didn't get the FOSS memo. People donate their time, know-how and abilities to a project, but they aren't tied to it. You can't force them to do anything. Stallman's FSF has been working on Hurd for over 30 years now. Linux sees more development in week than Hurd sees in a year. And it's not because Linux is open source and Hurd is not, it's because corporations have embraced linux and pay people to work on it. The kind of push behind wayland that you bring up flaky at best. The kind of push Wayland need is the kind the Linux kernel has. Where you have highly talented devs being payed not just to work on the kernel but payed to implement specific features into it. You need valve to hire people to work on wayland and ask those people: "what is needed to make it the default on the Deck? This, this and this? good, start with the first point, and keep going till it's implemented." That's the kind of push Wayland needs and is sadly lacking.
@@AlucardNoir you posted a load of text there man but after reading it all I'll say this Wayland isn't like the kernel, it isn't meant to be like the kernel, in fact the whole point of Wayland is to be a smaller protocol so it isn't a bitch to maintain like x is Wayland might need a "push" like you're saying but I honestly think that what redhat and the community are doing is enough many people including myself are already on Wayland, it's largely usable already. There are definitely some problems that need to be ironed out such as global keybindings and overall bugs. These problems can frankly be solved in ~5 years at current pace Wayland is already here and we've just gotta wait a little while longer for them to make it perfect
@@nilnailscrew4784 I'm a gamer. A linux gamer but still a gamer. Do you know what I do when a game doesn't work? I refund it. I don't bother trying to make it work. I'm not QA for the devs. I'm there to have fun. It either works or it doesn't. I get it, a lot of linux users are hobbyists and computer enthusiast but this BS has to stop. Those problems can frankly be solved in around 5 years? BS like that is why the "year of linux" is never going to happen. Android isn't just at feature parity with iOS, it surpasses it in certain respects. That isn't because Android isn't linux, it's because it was developed like every other corporate product, while desktop linux has been trying to go past 1% for what? 20 years now? Android succeeded because there was a clear goal in mind and the devs didn't get to chose if they were going to work on a feature or not. Android didn't succed because it was FOSS but in spite of being foss. People want stuff that works. Traditional development produces that, even when the end product is free and open source, but that is not how most open source projects are developed. Until stuff like Wayland and Pipewire start being developed under a more traditional model - like the Kernel is, which was my point by the way - I am highly skeptical. Nobody had any interest in maintaining the desktop icons in Nautilus but when they got removed a dev immediately jumped to implement the feature in the form of a Gnome Extension. Foss devs work under the auspices of "good enough". That's why foss development is so slow. Pulse and Jack work, no need to hurry with Pipe. X.Org still works, why put that much effort into Wayland? Also, if you have a problem with me using the kernel as an example, then replace that with Firefox. How many web browsers not based on Gecko or Chromium does the linux desktop have? And yet FF is what most distro's ship. Why? Because unlike the crap Gnome develps for example FF is an actually usable, secure, modern browser. Still not as good as chromium based browsers, but decades ahead of Epiphany, or whatever it is they'be been calling it for the last decade. KHTMLwas the basis for both Apple's WebKIT and Google's Blink, yet it's years behind both. There are only two kind of successful foss projects, those that are one man projects where one talented dev can singlehandedly maintain and develop something, and the ones that are made by large corporations. Anything in between tends to be mediocre at best. EDIT. and again, all this from someone who exclusively uses Linux on desktop, laptop and mobile.
@@AlucardNoir I think everyone would be happy if more corporations started building and funding FOSS but they aren't 100% necessary most of the linux toolchain wasn't a big project backed by millions of dollars like linux or firefox. If we want to compare wayland to something I think the most accurate thing to compare it to is systemd. Wayland does have some corporate support just like systemd had (redhat) and systemd took a fair bit to take over the market. In your original comment you said that you doubted that wayland will be ready in 8 years and after I said it could be done in 5 you said that wasn't good enough either. There is precedent for medium-sized projects to become ubiquitous (systemd). Believe me I think we all want this project to reach perfection as quick as possible and corporations do speed up the process. However we don't NEED them.
This is just one of those "evolution of software" conversations, where naturally X11 will be replaced at some point. Whether that is replaced by X12 or Wayland or something. People just seem to have a problem when the topic of sunsetting a project comes about, even when it is raised with the most logical and well-intentioned objectives. And yeah, once X11 is replaced and no longer actively used, it makes no sense to waste development time and resources into supporting it. The same could be said of X10.
Wayland is ready the moment xmonad switches to wayland and if that doesn't happen then I probably wont switch, unless there is some similar haskell based tiling wayland compositor to use as replacement
As long as my games load and run smoothly without having to juggle 3 terminals and 5 commands to launch them in awesome under Xwayland then I'll be fine switching to wayland. If not *rolls up sleeves I guess it'll be time to fork!
Can't switch without Night Light and gaming support on NVidia. NVidia is non-negotiable, since all ML libraries are written for CUDA with AMD not supported, or an afterthought at best.
One question that may be more important about future of Wayland or xorg,cause don't we forget, we are a minority, steam deck and its os will change it but mostly for the mini or handheld PC market How is Wayland on windows or MACos?, or for cellphones like Android or iPhone?
DirectX has nothing to with Xorg, they just both happen to have an X in the name, Xorg and Wayland are just Linux & BSD technologies. Wayland does have minor use on Windows in the context of WSL (Windows Subsystem For Linux) to allow you to render Linux applications as native Windows applications. Android or iOS have there own rendering stacks that also aren't related, except for the fact that OpenGL has been involved at one point or another.
I don't even wanna think about how I'd handle my discreet nvidia gpu in wayland, it's already a pain in the ass to use nvidia-prime in xorg and it gives me screen tearing until reboot unless I use vsync with picom, which I'd hate to do since I don't use compositing, (there's no force full composition pipeline option for hybrid systems). My dgpu will probably soon be legacy aswell since it's the oldest officially supported by the upstream driver.
If you use Nvidia or if you use a desktop like Mate Wayland is still not supported, I wonder if Gtk stops supporting Xorg in the short future, how many GUI programmers (not Gtk, but people that program GUIs) will switch to Qt, electron, etc. And is not like Gtk is a poplar library to begin with (excluding gnome apps, there are not that many people doing stuff with it).
I did some uni projects with GTK3 under Dlang. It was kind of ok but I think Qt is the superior framework. It has so much customisation and works better cross-platform.
On the subject of Nvidia cards, they are more like coprocessors for many people - CUDA, cuDNN etc. In this way, they are a forced choice when selecting a GPU for science and engineering applications.
Nvidia works fine on GNOME, and they don't care about anyone outside of GNOME users (honestly they don't even care much about them, they just hate them less).
@@encrypt3d587 It does or it doesn't. It depends on which generation of GPUs we're talking. On my Ampere graphics (RTX 3xxx) there's no usable compositor as every single one I've tried has its own set of issues, even though on Intel and AMD it has been pretty good for some time already.
Nvidia doesn't refuse to support, they've literally put out a list of points that they know they need to work on to bring Wayland support up to the state it should be. What they do is don't put as much effort into supporting it as we'd like.
@@BrodieRobertson Also, multimonitor seems to work better on Wayland... 144Hz and Freesync... On the one side, new KDE-update made some problems, with Latte and stuff. But that's what you get, when you want to be on a "Bleeding Edge" 😜
People don't stay on X11 because they're lazy though. They've stayed on X11 because replacing X11 with Wayland breaks lots of previously supported use cases and some hardware too apparently. Pulling the carpet from underneath won't help at all. The only reason to remove X11 support is that it will speed up development for GTK5, and that's a great reason if the missing pieces are covered before then.
Wayland works well enough for me. I run into occasional glitches, mostly minor. There is no missing functionality for my use. Not saying there is no functionality missing, just that I don't need that functionality. GTK5 is surely at least another 5 years out, likely more. In that time I believe all the legitimate complaints against Wayland will be resolved. It's already 90% there. I only see a few complaints consistently come up against it, one of which is misinformed. Everyone else just seems to hate on it out of principle, much like systemd. Doesn't make much difference to me either way though. I don't use Gnome, and I try to avoid GTK applications where possible.
Xorg users will have to grow up and start writing their own code or pay someone to do it for them it is not as if Wayland is new. Of course there is always Debian stable Slackware if they prefer but at some point x11 is going the way of DOS.
I don't know if I care too much , only for the fact that every GTK iteration breaks previous versions - so the fact that it is solely going for wayland support is immaterial to me as that will carry over to Wayland, and you end up with multiple GTK releases on your system. Terrible, absolutely terrible.
@@spicynoodle7419 true, but not all apps are current version of GTK , so I would be hard pressed to find a distro out there that doesn't come with multiple versions of GTK.
There's an argument to be made both ways, if support isn't difficult to maintain maybe it's best to keep it going for the small number of users and use cases which might need it, on the other hand if barely anyone is using it then wouldn't that developer effort be better used elsewhere in the project. I can't universally say what is right it really depends on the project and how big the user gap really is.
@@BrodieRobertson tbh if the program is designed more as a library that attaches to the GUI framework instead of blended spaghetti, changing GTK versions would be almost effortless. With that kind of design you could even make a Qt version easily. Unfortunately designing modular systems is tricky in the first place and there are plenty of Foss GUI projects that scream "fuck it, let's just get any GUI working asap"
I like nVidia cards, alternate window managers, AND alternative operating systems like Solaris and OpenIndiana. You can guess what my opinion about Wayland is. GNOME and GTK just keep getting crappier and more Linux-centric, and we all know Red Hat/IBM won't stop shoving their big fat Wayland and SystemD down everyone's throats.
odds are you won't think your awesome config is perfect after 10 years knowing myself and other twm users you'll probably have the urge to switch to something else within 5-10 years tops I really don't think that awesome is what'll hold you back not to mention there are Wayland versions of many of the big twms working or in progress so you might just be able to use the same config on awesomeland (name I made up) when the time comes
But the x11 complainers do have a choice it is open source so they can fork X11 and gtk5 and maintain it for themselves or hire someone to do it for them. If they have a problem with Wayland they can contribute code. It is rather their choice to whine and I choose not to listen to their trolling. The x11/Wayland developers gave their reasoning years ago and the x11 trolls are disrespectful of the developers.
@@BrodieRobertson Me neither… I guess it was just an assumption, as the only non-systemd distro I used lately was Devuan and it didn’t offer Wayland. Maybe I also read something somewhere about it, but I’m not sure.
All those new things are great, but I've noticed that noone is realy thinking about accessibility anymore. Wayland, gtk4, flatpak, every version of qt... are barely usable with a screen reader. One reason why is it like that is probably the fact that orca (the only somewath good screen reader for linux) has only 1 fulltime maintainer...
A lot of people using computers are blind or have low vision, so I think if linux improves its accessibility, we culd get more people to switch.
and the funniest part is that they're not all that usable for people with working eyes either
Come back for tomorrows video then because Redhat is thinking about that as well.
The problem is that most nobody cares that can do anything about it cares. Most open source voices and voice engines used on linux desktop are based on University projects that were abandoned and are only maintained by the few people that actually care about the blind or are blind themselves. Wayland isn't ready for prime time for people that can see, hear and have full mobility, complaining it's worse for the physically disabled while valid, is also pointless.
It's an "our way or the highway" mindset, and I'm sick of it.
There's a lot of things that can happen in next 10 years. In 2012 Wayland was still a concept with some early implementations rather than anything functional and one might say that in 10 years it should be 100% complete and much better than X11, but to my observation, the whole graphics stack has transformed quite significantly. The development is pretty slow, but still 10 years will make a ton of difference. Current showstoppers doesn't seem impossible to fix, whether it's the NVIDIA, key binding protocol, color management or interface scaling issues. That's just my guess.
remote sessions are really important for the enterprise sector. and the x11 remote solutions at the moment are just way more reliable
yeah but gtk5 is still gonna take a long time to come around, and wayland will get more support in enterprise for sure too, since it's redhat calling the shots in this and many other cases, sure the devs might talk about dropping support but if it's not ready, redhat won't ship it, just like fedora's plans to drop bios support got almost instantly abandoned
@@DMSBrian24 it was regarding the comment on gitlab brodie couldn’t quite relate to
RDP is so much nicer than VNC
@@nevoyu for Hacking 😬
Seriously, feature parity or don't bother getting out of bed Wayland.
As someone who likes to be on the bleeding edge, I’ve been running Wayland on my main desktop since the release of Fedora 36 without any deal breaking issues (for me). I occasionally run into issues with screenshot tools, but for now I’m just using the built in tool from GNOME instead. But I’m all too aware that there are many different workflows right now that make Xorg a more viable option currently. Not to mention the huge amount of window managers and DEs only available for it.
Same here. I have encountered one or two issues but that's because I not only run Nvidia but it's on a laptop with 2 GPUs. Dual GPU laptops didn't get screen tearing fixes until like 2019. I'm a few days away from getting my first PC built tho. It's full AMD (except the Intel network card). It should have 105% Linux and Wayland compatibility :D
Also, one of the best WMs (Qtile) has Wayland support. I think there are a couple more.
Not even basic things like drag and drop to extract a file works under Wayland. That thing Wayland is not ready YET
@@SirLancelotS I don't see how this is a display server issue rather than the Archie tool and file manager being the culprits here 🤔 Does the same setup work as expected under Xorg?
SteamVR currently only works on Xorg and crashes under Wayland. Steam has a lot of other issues with Wayland. Sad that they've not done much about it.
I've been running wayland with kde for months now and the only issue it's ever given me are context menus being out of place and some weird font issues with flatpaks. tbh if wayland didn't exist I might have never decided to daily drive linux because I have a 144hz monitor and 2 60hz monitors and xorg limiting my 144hz monitor to 60 is a pretty big turn off considering I game a lot
Wayland will be adopted as the industry standard about as quickly as ipv6 was.
I just wonder what it will do with XFCE?
Ubuntu 22.04 and Fedora 36, two of the most used desktop distros, already shipped with Wayland as the default. Considering it's started in 2008, it already has more progress than IPv6 which started in 1995.
I have been using Sway for 6-7 months now and after the initial setup, it has been great!
Many applications already support wayland natively, but the option just might not be enabled by default. For example, firefox supports wayland but doesn't enable support by default. You can use the wayland version by setting the MOZ_ENABLE_WAYLAND=1 environment variable.
As an XFCE user I will be stuck with xorg for a while. I don't mind though. Last time I checked, wayland was pretty bad for gaming, as there was no support for "unredirect fullscreen windows", which basically means the compositor is running at all times which adds input lag and lowers the frame rate
My only problem with wayland is that I don't have a viable replacement for my window manager, xmonad. Waymonad exists, but it's way to experimental and incomplete for daily use. I could switch to a fork of dwm, but that's a pain and I'm really used to the way xmonad works currently.
I think that the current issue with wayland is just the ecosystem is decades less mature than xorg and doesn't have the laundry list of options that xorg has. This is an issue that can only be fixed with time, and when I have a wayland compositor that implements enough of my workflow, I'll be on it like butter on toast.
If anyone knows of a good dynamic twm for wayland that I can get on Nixos, let me know, I'd love to try it.
dwl or river, dwl crashes often randomly for me(I have nvidia gpu),
river I disliked because of the config file is basic using bash commands, then you would have to learn a programming language to then use your compositor.
I've been seeing a lot of hyprland buzz on unixporn
While I like Wayland and use sway on my laptop, it has too much work to be done to it for a change this big.
In 7 or 8 year's time, I think it will be a very different story. I say this as a person who is going to use X11 for the foreseeable future (it's still super broken on my system).
@@miguelguthridge that'll probably be down to how many more spanners Nvidia throws in the works along the way. They finally caved on the GBM/EGLStreams thing, but Wayland is still unusable on the Nvidia blob for me.
I am actually trying out Wayland/Sway at the moment. There have been some stumbling blocks but all in all it seems to work. (with nvidia gtx-760)
What I noticed is the screen output seems much more crisp using sway than it did using i3. Whether that is just my imagination I don't know.
Of course I can't do without xorg-xwayland yet, as too many programs don't natively support Wayland.
I think this is a big advantage proprietary software has over open source. Once the top decides something needs to be done or changed, it is changed within a "small" amount of time.
As far as I understand things like Wayland and pipewire are necessary for the future of the Linux desktop so I think it is a good thing that projects such as Fedora push through the changes even if it is still buggy.
Wayland and pipewire are given. Almost all big distro are switching to those tech. Nvidia should accelarate fixing wayland related issue.
Impossible technologically, a fork will 100% be made with xorg support if it is not official even
The future is bloat.
@@xrafter well technically speaking Wayland better follows the Unix philosophy with its display server model but nobody ever mentions that
@@spaceguybob
Yeah, Unix is also bloat. Thanks for reminding me!
@@xrafter so therefore the future is Unix
@@xrafter life is bloat. Embrace the microwave background radiation of death
openbsd already does maintain xorg somewhat, they have a project called xenocara which, although they claim is not a fork, seems to be effectively used to distribute the patches they need to xorg
It's mainly used to track the updates to Xorg, so if the current small amount of Xorg development falls out then they're left with a lot more work to do.
@@BrodieRobertson ah rip. I'm sure they'd pick it up as they've picked up huge projects prior but it would probably be at the state it is now if not worse, only occasionally pushing new drivers, fixing bugs/security issues and basically just maintaining it in its current state.
I am honestly not counting on OpenBSD porting their existing WMs to Wayland and I expect they will hold out on support unless it is absolutely necessary (ie everybody else has moved on). In addition they are, for the most part, a server/embeded OS anyways so it's not the worst thing if their graphic software is outdated.
@@hostgrady OpenBSD IS a desktop OS. I don't understand why people always think it's primarily a server OS. The OpenBSD devs use OpenBSD on the desktop as a daily driver. They care about the desktop experience and won't drop it. It's going to be interesting to see what they will do. I think Theo himself still uses fvwm, which is probably going to be impossible to port to wayland. OpenBSDs own cwm would be easy to reimplement using wlroots.
@@freesoftwareextremist8119 no I know people USE it as a desktop, but it's a server OS man.
@@hostgrady It's not a server OS. Why would it be? If you really want to go that route, then just like Linux, it's actually a mainframe OS.
I tried using wayland. I really, really tried, but I just can’t. My X11 setup is just too comfy no wayland setup comes close. Most my software just doesn’t work properly on wayland and the amount of visual customisation you can do on wayland compositors is severely limited compared to X11.
Until Wayland allows an app like a keyboard remapper to know which app window has the keyboard focus it is less than useless to me. Right now there is no standardized API for this on Wayland. On X11/Xorg it’s very easy to see things like the WM_CLASS and WM_NAME of the focused window and apply a specific keyboard shortcut overlay just for that app.
That will probably never be standardized but it may make it's way to wlroots
I tried Wayland recently with KDE..
I won’t try it again within the next 5 years.
KDE is unstable in wayland. Right now gnome has the better support.
@@JahidulIslam Not in my experience, KDE has been working well for me on Wayland (Intel chipset). Even using its screen capture tool (Spectacle), had no issues. Gnome on the other hand I've had crashes.
0:05 lmao "YOU WILL USE WAYLAND AND YOU WILL LIKE IT"
If GTK5 will get mainstream a lot of years from now, like 8-10 years in the future, dropping X11 won't be a problem, Wayland will get mainstream and overtake X11 in the most popular DEs until that time.
On Gnome, Wayland is already 80% ready if you are using Intel iGPU or AMD GPU/iGPU.
I've been maining wayland on Gnome for about a year now and I've rarely had issues. The only thing that's currently still broken is discord's / Team's screen share. (I’ve been on wayland ever since I got my 6900xt. My desktop experience with nvidia was horrendous)
@@BrodieRobertson AMD. I used to be an nvidia user but that was rubbish.
Wayland is still nowhere near ready enough to make mandatory.
Pierre wasn't talking about remote desktop software, he was talking about X forwarding. X forwarding is far, far more flexible than a VNC session or similar. To use an example from my work, I often have clients from several different remote machines all rendering on my computer under my window manager. There are several other users on these same systems, all running clients under their own logins at the same time. VNC simply cannot do that, it only does one session at a time, and even with something that can do multiple sessions like NoMachine, you have to run a full DE on each remote host in a window instead of just rendering the client(s) you need on your local X server. There is no way to run a Wayland client remotely, as far as I'm aware, so at best you're stuck with something like NoMachine that can at least do multiple sessions, and at worst you're stuck with the projects you showed off, which can only handle one user at a time.
I believe Waypipe is considered to be the de-facto Wayland equivalent to X application forwarding (not making any comments about its effectiveness or quality one way or the other; haven't tried it myself but have heard and read about it).
Also, when full-DE remote desktop is required, I'm very happy that FreeRDP exists as a much nicer alternative to VNC on Wayland. I've been using the regular gnome-desktop-sharing daemon provided by GNOME which serves an ms-rd:// endpoint for connecting with standard Microsoft RDP client apps. It works pretty well for my use cases.
The other day I was installing a Linux distro on a virtual machine. The default was Wayland. I wasn't able to see the mouse pointer until I switched to Xorg. I'm not against Wayland, but it seems like it still has issues that needs to be resolved. They can't just simply say, "It works on my computer so it's fine." They must think of the millions of Linux users out there with different hardware and use cases.
May be it is because of your virtual tool. I have used Gnome boxes and Libvert GUI. No mouse pointer issue on those tools.
KDE and Nvidia. When Wayland works flawlessly with these two, I'll switch. Occasionally I will try Wayland with KDE and it never works. I am not going out and buying a AMD card to replace my RTX 3080 Ti just to run Wayland.
For embedded systems like the head unit in cars, Nvidia does support Wayland. Maybe now that Nividia open-sourced their drivers, maybe in the next few years things will be improved.
As for now, it is Xorg. Wayland has yet to become a real boy.
Wayland KDE 5.25 is usable on my Geforce. You need very current libraries.
My Fedora install came with Wayland enabled by default (Desktop: Ryzen 3600x, RX 580, Laptop: Thinkpad T480, no Nvidia graphics just integrated Intel). This is my first daily driver Linux desktop in years. I haven't noticed anything broken yet, everything seems to "just work". Even the few games I have tried seems to work well even the Hogwarts Legacy one.
A lot will need to be done for this sort of major shift to be worthwhile.
If we can't get Wayland to be at least as feature-complete as Xorg, and at least as stable, and at least as accessible without losing stability, which is not fulfilled in either cases currently,
then it won't be a worthwhile shift.
Forcing it on users is not going to be worthwhile either as it undermines the premise of an open system.
Currently it's too unstable, and isn't as compatible as it should be.
It also isn't super great on older systems or Nvidia systems.
If all of its current issues are resolved by the time GTK5 rolls around, and it remains backwards-compatible, then it'll be absolutely worthwhile.
Currently, it is not.
I find it quite concerning that Gnome is considering app exclusivity, which also goes against the open ethos.
There's quite a bit of writing on the wall that needs to be addressed.
It has nothing to do with "app exclusivity" because its all about maintaining two parallel incompatible versions of their libraries, sometimes wholly separated in the case of static libs, this is a massive waste that you need to spend money and developers maintaining and wastes twice as much time developing any singular feature in the optimal case. (and 90% of cases are sub-optimal so generally expect around 350% slowdown for any functionality or solution) Its not like GTK3 or GTK4 will be dead in the sense of being completely unusable, it won't be actively maintained but you could still use it, it just means GTK5 features not seen in GTK3 and GTK4 will be exclusive to Wayland which 90% of the time isn't gonna change anything, if someone needs to make a GTK app they can just use GTK4, its extremely unlikely they'll need GTK5 features and since GTK is FOSS anyway if someone wants to bring certain GTK5 features to GTK4's Xorg implementation then they can fork GTK4 and do it themselves. I don't see how this is a problem. Do you expect GTK to work on every Windowing System all the time? That expectation is insane, you're hefting work on people for something that has no value, any scrub or company who built their own window system would still be excluded from GTK, X10 is also not supported by the more modern versions because what would the point be? X11 has been around for 40 years, most people who use computers now, even on Linux, weren't alive or weren't capable of doing anything with a computer when X11 was released, so why would you expect anyone to use it?
@@Spartan322 Well said and true to the point. I used both other the years what they see as features I see as vulnerabilities. They say well my screen capture is broken. Maybe that program doesn't work because they didn't want other program to screen capture private screens. Gnome screen capture toys work fine. If screen capture is harder on Wayland Go Wayland is what I say. one of the top reason I sopped using windows was they install the Xbox stream bar into windows making setting up screen capture to easy and hack-able. That and many more reasons. Even Nvidia are pushing Wayland drivers on now hardware, it will always be the old hardware that sticks to Xorg on new I have had no issue's on AMD or Intel, only shitty old Nvidia whose driver the last time I used my GTX1070 was crap. I'm selling that card soon as outside of Windows Gaming they are a crappy company to work with.
Gnome devs have been going rogue since Gnome3, been a decade. Yet Ubuntu & co stick with them, reinforcing their behaviour
I tried going "full wayland" before sdl2 announcement about support for wayland. It's too early, and still wayland is buggy. Crashes on bookmarks, only sway was good (as wm) but i love dwm and i always miss so badly when i try alternatives. Maybe if suckless will release dwm, dmenu for wayland i will try again or idk.
Replacements (some are just pain compared to simpler x11 counterparts), setting everything from scratch just to hit the wall is kinda pain and waste of time for now. Xorg/xenocara is stable like rock, i think everyone needs time before switching.
Wayland have pros, like no need for additional stuff but im not ready yet.
Anyone who made all these switch guides, big respect, thats a lot. For me, im waiting for ' the right time '.
It's already possible to use Wayland applications on X11 by running Weston in a window. This would be a solved problem if a better integrated compatibility layer was built using Weston, as a sort of counterpart to xorg-xwayland.
wlroots compositors can be run in a window, too. Although the more important thing is running x11 apps in wayland or porting them from x11 to wayland.
Linux is not ready for a wayland only desktop just yet.. bunch of current projects just aint there yet to name a couple barrier/synergy. maybe this threat from GTK will put a fire under them to do so but there are WAY more projects that are NOT ready yet also
Still using Xorg-X11 here, as 1. Wayland applications don't run over SSH, 2. my UI scripts depend on Xephyr, xdotool and xwd, and 3. my screen cat (oneko) does not work on wayland!
What are your thoughts on the new .jxl photo format by the JPEG group? When will it/will it ever have widespread Linux support?
I haven't looked into it but I want to see the Quite Ok Image Format become standard. It's way better than PNG.
I use X because it just werks, a display server is not something I give much consideration about as long as it's doing its job
i mean the problem with X11 is that people are comfortable staying on it. so nobody really tries to move to Wayland seriously.
If GTK5 forces everyone's hand - there will be more people getting around to fix Wayland. so, in this regard, i see this as a good idea.
No it just creates a lot of pissed-off people looking for a distribution that doesn't have
GTK5.
@@ewetoo unless they are willing to maintain gtk4, they will have to jump ship eventually.
either that or someone will have to maintain GTK5 non-wayland backend.
I love how most people didn't even bother to watch the first 2 minutes of the video before posting useless comments.
That's why one must choose wisely the words to use in an introduction. If you mislead your viewers from the beginning, don't complain later.
This switch is inevitable, but it is unlikely to be in 10 years, unless nvidia fixes their drives in 3 years max. I would think that projects the size of gtk5 do not decide to drop support for such an important part of the project in the middle of development with no guarantee that the nvidia users would be able to run wayland.
I disagree all wayland problems are not that major to not be fixed in the next 10 years. Both Gnome and Kde are pushing wayland, some of the biggest distros are already pushing wayland. I don't think that everyone will be on wayland in 10 years, but it's likely that the majority of users will use wayland.
@@ANGELRA the point here is not that those problems can't be fixed in 10 years. It is that it probably will be fixed in 10 years, but gtk5 is unlikely to abandon x11, if wayland is not rock solid in 3-4 years, because they can't just bet in the middle of development cycle that it would be rock solid by the time gtk5 is released
@@ЯрославФ-ы7ж I agree with you, but still 3-4 years it's a long time, we will see what happens.
As a window manager only user this worries me, as I don't want to have to switch to another window manager. I've tried sway and qtile with Wayland, they both missed a lot of features I use, I could not get to change the keyboard layout or set the compose key, unable to set scaling to be larger, etc. And I'd also need to change a lot of my programs that are Xorg only.
I remember when Plasma 5 had come out and FreeBSD still only had a super old version of Plasma 4 and it was a total pain to install. Very interesting how it's starting to keep up with Linux technologies like Wayland
While I’ve been mostly happy with wayland and sway, it has issues with displaying submenus for VSTs I use for OBS filters/in my DAW, which is really annoying, so this is hopefully good news in that people will be incentivized to patch issues
Wayland is not fully ready although its ready at their side the window manager sides have not fully adopted and integrated with the separate api system of wayland for some more intricate settings
Wayland works in my Manjaro KDE (recent) with Nvidia 750 Ti, that I am going to replace with a 3060 Ti, as prices went down, and AMD still is not on par for gaming, but I do not know why, writing it is slow, and waydroid do not work, so I switched back to X11.
Current sway user here, love it and Wayland, all my games feel way smoother, and VR works so much better on Wayland than it does xorg
Great, I'm happy with that too.
I'm using Wayland on my ASUS TUF GAMING F17 with an NVIDIA RTX 3050 Ti and I don't see any issues.
So...It seems to be fine for me. Dunno why Ubuntu doesn't have that session as the default and Xorg as a backup.
While I think I can't comment on that, however, I have an Intel HD laptop, and nvidia within my PC, my laptop is running definitely worse, for example frame drops happen frequently on wayland than on x11, the wayland is less efficient than x11. This is a big surprise but this is the experience I have with Intel. With nvidia nothing works however, the desktop is definitely smoother compared to x11, but that's it. I just want to express that as of things right now I don't see huge benefits going wayland only, you have to definitely consider the compatibility issues. Although like I said I can't stand on either decision since I didn't use wayland on it's glory (AMD gpu).
I daily use Fedora 37 under KDE in wayland mode and have no issue with my RTX 2080 Ti, I can use all applications and even play games normaly =O
I feel the pain on both sides, both wayland and gtk developers, even I have been in this situation. When people are using old software and you're writing something new (in their opinion, better) but the users, still use the old software. Dropping support left and right and forcing GTK or nothing type of mentality is not going to cut it. There is a hard thuth to swallow: if your program is good, people are going to use it, you don't have to threathen anybody or drop support for anything, people will eventually switch because they like it.
Seems a bit early but these tickets tend to stick arround forever so i guess its a good goal to have. Honestly the best way to get everyone onto wayland is to have a good backwards compatibility to old xorg apps.
I know the devs are not thrilled about doing that but they will just have to suck it up if they don't want the fragmentation period to last forever.
Get my xorg apps to run good enough on wayland and i will switch tomorrow.... once everyone is over new apps will natuerally be wayland but if they refuse we will have another python2 vs python3 situation...
Hopefully Wayland on NVIDIA is stable by then I'm not 100% against it if so
mainly feelin for all the small niche desktops whose development may not survive, but the way of this place ig
Feels like when the new guy who just got hired starts trying to change everything.
I really want to use Wayland but I can't. Discord cannot stream on Wayland (if you a beginner and I want to recommend it to beginners). My some games not works well on Wayland. I.e.: on Doom 2016 I take 200 fps (it's a limit). But on Wayland I take 80-110 fps. I'll really wait until it's ready. I know waiting works everytime. First time I use Linux, I can't even use my gtx 1650 maxq GPU. But now I can use every hardware on linux.
Honestly, I'd use Wayland across the board if Wayland didn't have problems on NVIDIA (Gamescope not working, and assorted problems thanks to XWayland) and KDE (flickering bugs, vertical panels not working properly unless I sign out, and other things), and if FreeSync/GSync was supported on both NVIDIA and AMD/Intel. I feel like Fedora dropping X11 support altogether on their main branch on NVIDIA GPUs is a bit early right now. It's a bit of a mess, and I have to use X11 for now on my NVIDIA system.
Looking forward to when Wayland gets support for HDR displays and GSync.
It took 10 years for them to move from GTK3 to GTK4 so by the time GTK5 comes, you probably don't have any reason to use X11 anymore. And Variable Refresh Rates work on Wayland
steam and proton are X11 apps(maybe?), so yeah, I'll use what my most used applications do.
I'll switch to wayland as soon as something compatible with awesomewn for wayland gets available
If I can get my hands on an Intel ARC card, I'd love to benchmark Wayland on it. My 1060 hangs after an hour in a Wayland session across multiple different DEs. Not knowledgeable enough to figure out why, but its happened on both Fedora and Ubuntu. I had a laptop years back that had Intel graphics and the framerate jumped when I switched to Wayland from Xorg, and because I've seen what Wayland CAN do, I actually am not against the idea of switching over years ahead once these kinks get worked out.
I have a Zotac NVIDIA Geforce GTX 1050 TI, and i was wondering if there was a way for me to switch over to wayland despite me using an NVIDIA card?
I mean, by the time GTK5 becomes reality, Wayland should be fully capable of replacing X in every workload.
Lol! I didn't even knew that gtk5 is here. And to be honest, I have never seen a single application written in gtk4. Most of them are written in gtk3 and many are using gtk2
GTK5 is a very long way away still
"Over time the Linux desktop's shifting more and more towards Wayland"
And yet there still is no reliable way of disabling VSync.....
There just isn't enough push behind Wayland. Don't get me wrong, I'm not saying X get's the support it needs, but Wayland is nowhere near ready and at this pace of development I doubt it will be ready by 2030. Unless you get someone like Steam backing it I doubt it will be able to replace the X11 in the next 10-15 years, maybe more. When there is an interest, be it personal from a highly taletned technical person or corporate from a corporate agent that has the money to hire talented devs open source develops at a breakneck pace, but when those interests are lacking open source tends to develop at a snails pace.
The push only exists behind Wayland at this point, all of the corporate linux projects are focused on that. X11 isn't completely going away for a very long time, Steam for example is still going to rely on Xwayland or an Xwayland like tech for games that rely on Xorg but you don't need to be running Xorg directly on metal to acheive that.
@@BrodieRobertson Well, see, you say the push only exists behind wayland, yet at the glacial pace of development the push might as well not be there.
Take the kernel as a counter example. If Intel or Samsung or Microsoft need a new function in the kernel they'll have it in the shortest amount of time possible.
Wayland on the other hand doesn't really have that kind of push behind it. Not to offend the people working on Wayland but the best and brightest aren't being payed to develop it.
The original X11 server dates back to 1987, while X dates back to 1984. In just 3 years doctoral students at MIT made the basis of what X has been for the past 35 years. Compared to the work that was done in that period, what the X.Org foundation has added over the past 20 years is paltry. Had wayland been university project 10 years ago it would have already supplanted X.Org. And that's because those doctoral students have something to prove. And outside uni, the kind of people that could make Wayland work in just a few years are getting payed a lot of money to work on other things corporations think are more important than a linux desktop renderer.
The push behind wayland might as well not be there considering how little there is to show. I'm a solus user, a while back the looser in charge - Josh Strobl - quit. Why? Because he asked for something to be done and the person he asked didn't do it. Because apparently Josh didn't get the FOSS memo. People donate their time, know-how and abilities to a project, but they aren't tied to it. You can't force them to do anything. Stallman's FSF has been working on Hurd for over 30 years now. Linux sees more development in week than Hurd sees in a year. And it's not because Linux is open source and Hurd is not, it's because corporations have embraced linux and pay people to work on it. The kind of push behind wayland that you bring up flaky at best. The kind of push Wayland need is the kind the Linux kernel has. Where you have highly talented devs being payed not just to work on the kernel but payed to implement specific features into it. You need valve to hire people to work on wayland and ask those people: "what is needed to make it the default on the Deck? This, this and this? good, start with the first point, and keep going till it's implemented." That's the kind of push Wayland needs and is sadly lacking.
@@AlucardNoir you posted a load of text there man but after reading it all I'll say this
Wayland isn't like the kernel, it isn't meant to be like the kernel, in fact the whole point of Wayland is to be a smaller protocol so it isn't a bitch to maintain like x is
Wayland might need a "push" like you're saying but I honestly think that what redhat and the community are doing is enough
many people including myself are already on Wayland, it's largely usable already. There are definitely some problems that need to be ironed out such as global keybindings and overall bugs. These problems can frankly be solved in ~5 years at current pace
Wayland is already here and we've just gotta wait a little while longer for them to make it perfect
@@nilnailscrew4784 I'm a gamer. A linux gamer but still a gamer. Do you know what I do when a game doesn't work? I refund it. I don't bother trying to make it work. I'm not QA for the devs. I'm there to have fun. It either works or it doesn't.
I get it, a lot of linux users are hobbyists and computer enthusiast but this BS has to stop. Those problems can frankly be solved in around 5 years? BS like that is why the "year of linux" is never going to happen.
Android isn't just at feature parity with iOS, it surpasses it in certain respects. That isn't because Android isn't linux, it's because it was developed like every other corporate product, while desktop linux has been trying to go past 1% for what? 20 years now? Android succeeded because there was a clear goal in mind and the devs didn't get to chose if they were going to work on a feature or not. Android didn't succed because it was FOSS but in spite of being foss.
People want stuff that works. Traditional development produces that, even when the end product is free and open source, but that is not how most open source projects are developed. Until stuff like Wayland and Pipewire start being developed under a more traditional model - like the Kernel is, which was my point by the way - I am highly skeptical.
Nobody had any interest in maintaining the desktop icons in Nautilus but when they got removed a dev immediately jumped to implement the feature in the form of a Gnome Extension. Foss devs work under the auspices of "good enough". That's why foss development is so slow. Pulse and Jack work, no need to hurry with Pipe. X.Org still works, why put that much effort into Wayland?
Also, if you have a problem with me using the kernel as an example, then replace that with Firefox. How many web browsers not based on Gecko or Chromium does the linux desktop have? And yet FF is what most distro's ship. Why? Because unlike the crap Gnome develps for example FF is an actually usable, secure, modern browser. Still not as good as chromium based browsers, but decades ahead of Epiphany, or whatever it is they'be been calling it for the last decade. KHTMLwas the basis for both Apple's WebKIT and Google's Blink, yet it's years behind both.
There are only two kind of successful foss projects, those that are one man projects where one talented dev can singlehandedly maintain and develop something, and the ones that are made by large corporations. Anything in between tends to be mediocre at best.
EDIT. and again, all this from someone who exclusively uses Linux on desktop, laptop and mobile.
@@AlucardNoir I think everyone would be happy if more corporations started building and funding FOSS but they aren't 100% necessary
most of the linux toolchain wasn't a big project backed by millions of dollars like linux or firefox. If we want to compare wayland to something I think the most accurate thing to compare it to is systemd. Wayland does have some corporate support just like systemd had (redhat) and systemd took a fair bit to take over the market. In your original comment you said that you doubted that wayland will be ready in 8 years and after I said it could be done in 5 you said that wasn't good enough either. There is precedent for medium-sized projects to become ubiquitous (systemd).
Believe me I think we all want this project to reach perfection as quick as possible and corporations do speed up the process. However we don't NEED them.
This is just one of those "evolution of software" conversations, where naturally X11 will be replaced at some point. Whether that is replaced by X12 or Wayland or something. People just seem to have a problem when the topic of sunsetting a project comes about, even when it is raised with the most logical and well-intentioned objectives.
And yeah, once X11 is replaced and no longer actively used, it makes no sense to waste development time and resources into supporting it. The same could be said of X10.
Wayland is ready the moment xmonad switches to wayland and if that doesn't happen then I probably wont switch, unless there is some similar haskell based tiling wayland compositor to use as replacement
As long as my games load and run smoothly without having to juggle 3 terminals and 5 commands to launch them in awesome under Xwayland then I'll be fine switching to wayland. If not *rolls up sleeves
I guess it'll be time to fork!
FreeBSD and nixOS (if systemd was removed) are looking like better operating systems.
Can't switch without Night Light and gaming support on NVidia. NVidia is non-negotiable, since all ML libraries are written for CUDA with AMD not supported, or an afterthought at best.
how the hell this stuff is so coupled together?
One question that may be more important about future of Wayland or xorg,cause don't we forget, we are a minority, steam deck and its os will change it but mostly for the mini or handheld PC market
How is Wayland on windows or MACos?, or for cellphones like Android or iPhone?
May be mkin a stupid question but I assume at least in windows some form of xorg is used, or directx is related to it in some form or another
DirectX has nothing to with Xorg, they just both happen to have an X in the name, Xorg and Wayland are just Linux & BSD technologies. Wayland does have minor use on Windows in the context of WSL (Windows Subsystem For Linux) to allow you to render Linux applications as native Windows applications. Android or iOS have there own rendering stacks that also aren't related, except for the fact that OpenGL has been involved at one point or another.
@@BrodieRobertson thanks for clarifing
because of similar name and function i assume a connection
Whats the alternative to xkill on wayland?
There is no need for X11 drivers if you are using KMS and Mesa. X11 modesetting driver can continue working fine.
There is working and working well
I don't know why but I got this feeling that "I will use Wayland and I will like it".
Simple screen recorder does not have a great support for Wayland. :(
I didn't know it had any wayland support
@@BrodieRobertson I think it does. But on wayland it's glitchy for me. So I have to use xorg!!!
I don't even wanna think about how I'd handle my discreet nvidia gpu in wayland, it's already a pain in the ass to use nvidia-prime in xorg and it gives me screen tearing until reboot unless I use vsync with picom, which I'd hate to do since I don't use compositing, (there's no force full composition pipeline option for hybrid systems). My dgpu will probably soon be legacy aswell since it's the oldest officially supported by the upstream driver.
I feel like a lot of people forget and I even forget about hybrid systems now that I don't use one, I remember it even being a pain under Xorg.
GTK2 only died because of security bugs. GTK3 might be okay now but it took 5 or more years to finally get to the stability of GTK2.
I wonder if GIMP will reach GTK7 while it's still active
What if your PC says it doesn’t support Wayland?
Hopefully it does by then
If you use Nvidia or if you use a desktop like Mate Wayland is still not supported, I wonder if Gtk stops supporting Xorg in the short future, how many GUI programmers (not Gtk, but people that program GUIs) will switch to Qt, electron, etc. And is not like Gtk is a poplar library to begin with (excluding gnome apps, there are not that many people doing stuff with it).
I did some uni projects with GTK3 under Dlang. It was kind of ok but I think Qt is the superior framework. It has so much customisation and works better cross-platform.
To be fair a lot of GUI app developers don't even use GTK4 at this point so the point where it'll actually be a concern is a long way away.
On the subject of Nvidia cards, they are more like coprocessors for many people - CUDA, cuDNN etc. In this way, they are a forced choice when selecting a GPU for science and engineering applications.
What the point of making wayland only when nivdia refuses to support it
Nvidia works fine on GNOME, and they don't care about anyone outside of GNOME users (honestly they don't even care much about them, they just hate them less).
Eh, just a matter of time
@@encrypt3d587 It does or it doesn't. It depends on which generation of GPUs we're talking. On my Ampere graphics (RTX 3xxx) there's no usable compositor as every single one I've tried has its own set of issues, even though on Intel and AMD it has been pretty good for some time already.
Nvidia doesn't refuse to support, they've literally put out a list of points that they know they need to work on to bring Wayland support up to the state it should be. What they do is don't put as much effort into supporting it as we'd like.
the only reason i use wayland is because pinch to zoom on firefox
How long will Wayland bashers keep ignoring wayvnc?
When people actually remember it exists is a good start
Pipewire and Wayland. I use both of them.
Maybe, because I don't do any video editing or such. Basic use and gaming...
I use pipewire on my system
@@BrodieRobertson Also, multimonitor seems to work better on Wayland... 144Hz and Freesync...
On the one side, new KDE-update made some problems, with Latte and stuff. But that's what you get, when you want to be on a "Bleeding Edge" 😜
I WANT to use Wayland, but it's flickering like crazy on NVidia and KDE. also the menu button doesn't work
What distro
@@BjornsTIR Garuda Linux which is an Arch based distro
@@MarkMann1 plasma is not a distro.
It's not Wayland that's flickering, it's your graphics card XD (jk)
@@donaldmickunas8552 sorry, I misread it as DE since I was glancing over some other stuff. I'm running Garuda
I think this change is very important to make Wayland adoption faster.
People don't stay on X11 because they're lazy though. They've stayed on X11 because replacing X11 with Wayland breaks lots of previously supported use cases and some hardware too apparently. Pulling the carpet from underneath won't help at all. The only reason to remove X11 support is that it will speed up development for GTK5, and that's a great reason if the missing pieces are covered before then.
@@pastenml sure
@@Eyuphuro Wayland needs to import a pile of standards from X11 or make their own equivalent.
I don't use Wayland because I have a Nvidia GPU and I don't want to drop bspwm and all the other great window managers.
Wayland works well enough for me. I run into occasional glitches, mostly minor. There is no missing functionality for my use. Not saying there is no functionality missing, just that I don't need that functionality.
GTK5 is surely at least another 5 years out, likely more. In that time I believe all the legitimate complaints against Wayland will be resolved. It's already 90% there. I only see a few complaints consistently come up against it, one of which is misinformed. Everyone else just seems to hate on it out of principle, much like systemd.
Doesn't make much difference to me either way though. I don't use Gnome, and I try to avoid GTK applications where possible.
Xorg users will have to grow up and start writing their own code or pay someone to do it for them it is not as if Wayland is new. Of course there is always Debian stable Slackware if they prefer but at some point x11 is going the way of DOS.
I don't know if I care too much , only for the fact that every GTK iteration breaks previous versions - so the fact that it is solely going for wayland support is immaterial to me as that will carry over to Wayland, and you end up with multiple GTK releases on your system. Terrible, absolutely terrible.
Dropping support is the right thing to do. We don't want to be Microsoft here, supporting GUIs from Win95, XP, 7, 8.1, 10, 11 simultaneously.
@@spicynoodle7419 true, but not all apps are current version of GTK , so I would be hard pressed to find a distro out there that doesn't come with multiple versions of GTK.
@@breadmoth6443 GTK3 and 4 are basically the same visually. I think having both is ok. The GTK2-themed XFCE and Mint programs are low-key disgusting.
There's an argument to be made both ways, if support isn't difficult to maintain maybe it's best to keep it going for the small number of users and use cases which might need it, on the other hand if barely anyone is using it then wouldn't that developer effort be better used elsewhere in the project. I can't universally say what is right it really depends on the project and how big the user gap really is.
@@BrodieRobertson tbh if the program is designed more as a library that attaches to the GUI framework instead of blended spaghetti, changing GTK versions would be almost effortless. With that kind of design you could even make a Qt version easily. Unfortunately designing modular systems is tricky in the first place and there are plenty of Foss GUI projects that scream "fuck it, let's just get any GUI working asap"
Wayland is still a mess. Wayland users hate this one fact.
Linux just got 30 years old and Brodie is already making predictions 20 years into the future
am i the only one who is confused why he uses ublock on brave?
Ublock offers fastly more control
I like nVidia cards, alternate window managers, AND alternative operating systems like Solaris and OpenIndiana. You can guess what my opinion about Wayland is. GNOME and GTK just keep getting crappier and more Linux-centric, and we all know Red Hat/IBM won't stop shoving their big fat Wayland and SystemD down everyone's throats.
X is frigging ancient, like crazy ancient, and an absolute dogs breakfast under the hood.
Yet X does the job reasonably well. Whereas Wayland implementations still have irritating gaps in functionality.
Maybe somebody will make "wxorg" and we'll have some compatibility between Wayland and Xorg
Well you'll find out seen
🥺 i really don't want to remove my perfect awesome config
and awesome doesn't have any plan to switch
and i really need GTK after 10 year
Looks like Xwayland will support entire DEs and WMs in the future. So you will be able to use awesomewm inside wayland
odds are you won't think your awesome config is perfect after 10 years
knowing myself and other twm users you'll probably have the urge to switch to something else within 5-10 years tops
I really don't think that awesome is what'll hold you back
not to mention there are Wayland versions of many of the big twms working or in progress so you might just be able to use the same config on awesomeland (name I made up) when the time comes
But the x11 complainers do have a choice it is open source so they can fork X11 and gtk5 and maintain it for themselves or hire someone to do it for them.
If they have a problem with Wayland they can contribute code.
It is rather their choice to whine and I choose not to listen to their trolling. The x11/Wayland developers gave their reasoning years ago and the x11 trolls are disrespectful of the developers.
Wayland on BSD? I thought Wayland was bound to systemd. Interesting to know, that it’s not.
I don't even know where you heard that
@@BrodieRobertson Me neither… I guess it was just an assumption, as the only non-systemd distro I used lately was Devuan and it didn’t offer Wayland. Maybe I also read something somewhere about it, but I’m not sure.
I cany make wayfire work on my pc, it just crashes my whole pc
also people that use tiling window managers use xorg
X11 is the only thing nvidia users can run, I hope that wayland fixes it by the time gtk5 comes out...
No, the current NVIDIA drivers support stuff needed by Wayland compositors. I ran Sway and KDE Plasma/Wayland on my PC with an NVIDIA GPU.
I still haven't even found one gtk4 only app that i rely on, super happy with my gtk 3 apps, so yeah whatever
Unless you live in Gnome it pretty easy to miss
This is such a nothing-burger. GTK5 is 10 years away. Anyone here thinks that (if this even happens) Wayland won't be ready in 10 years?!
I use XFree86 and MWM btw
Haha!