1. 77

    1. 22

      What a solid call to action! The writing so successfully evokes the unfliching conviction of inevitability, making it clear that inertia still holds stable platforms while simultaniously grasping the chance to present itself as The Moment– and indeed, so much is happening!

      The XFCE devs are discussing Wayland support as well, potentially hitting the GNOME2 market and catching some of the customizability-minded folks who bounced off of Plasma. Seems like tiling folks could hardly be better served than by Sway (not that I don’t love the alts :p), and GNOME/mutter support is old news. What an exiting time!

    2. 12

      Really, the opinion I view with the most suspicion is this:

      “By the time Ubuntu LTS removes X11 too, everything will be fully ready.”

      which involves at least three assumptions about the future with an indefinite horizon.

    3. 17

      As always, it was not a good idea to assume the color management+HDR was something that could be skipped for later. How are you going to convince creators & those working with creators to give up a vital part of the process/workflow just to move to Wayland when X11 is still working?

      1. 4

        Not trying to be a troll, genuinely asking - is there all that many creators on linux? Like, there are a few very great open-source software (krita, blender) for artists, but almost every industry standard only runs on windows and mac, and the latter has perhaps the biggest name among creators - a big part of which is color management.

        Could linux ever properly do this latter?

        1. 11

          It’s a bad thing if someone who wants to create art with Krita or Blender or some ither tool finds themselves pushed towards using a proprietary OS to do so because it has the color management features they need and Linux doesn’t. Creators will be motivated to use Linux if they can do their creative work on it.

        2. 5

          It’s a chicken-egg problem. If Linux can’t handle professional workflows, then professionals won’t use it.

          HDR is also useful for gaming and watching movies. However, the whole industry is still struggling with it. Monitors with fake HDR that doesn’t do anything are too common. Studios guard their HDR movies under the worst DRM, and there are too many competing half-baked HDR content standards.

        3. 4

          It doesn’t matter in relative terms how many folks are running proprietary OSs–absolutely there are a lot of folks creating on these platforms & you totally see like YouTubers talking about libre software & doing all of their editing on with those platforms. The kicker is that if you want to a platform to be an alternative to the status quo controlled by publicly-traded, US megacorps, you need to have the fundamental properties these users expect. You can choose darktable over Lightroom & learn a new tool without sacrificing much (if anything), but you can’t just be missing a core feature like color management. With the recent Creative Cloud price hikes, I’m sure there are amateurs & even professionals seeking free alternative to their workflow–and for now those folk will find better refuge in X11.

        4. 4

          There are definitely FOSS-related content creators who use and run Linux. Baby Wogue, BugsWriter, Novaspirit Tech are some names if you’re curious. Don’t treat this as an endorsement though, this “content creator” format is… really not my thing, so I’m really not familiar with their content soecufucakkt. I never managed to understand why anyone would ever put themselves through this, let alone how they could find it entertaining or useful. I am interested in these things by proxy (I know lots of people who work in marketing and a bunch of them who do social media specifically) but I couldn’t tell you what’s good.

          Linux does actually see a lot use in VFX though. That one’s pretty big.

          Could linux ever properly do this latter?

          If you mean colour management, it probably can. Getting one Wayland compositor to do it properly is just a technical struggle. Getting all of them to do it properly is a non-technical thing and I’m not sure it’ll ever happen in a productive format (i.e. one that actually guarantees you can run any app, in any desktop environment, and it’ll work without mucking with gnome-tweaks-tool, gsettings, obscure Qt environment variables and magic incantations in ~/.config/colormanagement.d or whatever). But we’ve been here before – there was a time when anyone who needed accessibility features mostly used Gnome and old terminal apps, for example – if a compositor properly supports one critical feature, the people who need it will use it.

          If you mean the kind of stuff that content creators need to do their work, no, absolutely not, not unless the way Linux desktop (for want of a better word – they rarely run on desktop PCs specifically anymore) environments and applications are developed changes radically. Not so much in its technical aspects, for all its warts Wayland is a good enough vessel, but in its approach to non- or quasi-technical matters, like feature deprecation, release management and understanding of non-technical users.

          I don’t get their work but the kind of effort and focus content creators put in it is extraordinary. It’s a kind of work that requires a very wide array of things, from keeping up appointments and traveling to studios to coordinating shipments and travel.

          When your whole day, and all of your suppliers’ and partners’ days are divided in 30-minute timeslots that you need to sync up on weeks before a meeting and still possibly reschedule at the last minute, you’re not going to convince people to play Wayland roulette. You can’t give them an OS where they need to update their browser and password manager and whatever every time a security update hits so as not to get doxxed, but where there’s a 50/50 chance that calendar integration, or cam/screen sharing in browsers will stop working after every update, depending on what compositor they’re running, and whether they’re running on Nvidia or AMD or whatever. Or where their next update may feature a radical, bold new design for some feature, which has seen zero testing, and which is not going to get reverted unless hell freezes over because the designer who pushed it has a bold vision and commit access.

          People put up with that only if it’s essential for their work (like many us) or if they’re fans of that particular technology for whatever reason (also like many of us, I guess). For the vast majority of content creators (i.e. everyone whose audience isn’t specifically interested in it) Linux is neither, it’s just an OS whose two main desktop technologies have been in beta since the 1990s and undergo a ritual destruction every time they get to a point where all they need is bugfixes.

        5. 3

          I don’t have user numbers but it is just not the case that every “industry standard” only runs on windows and mac. In particular, when considering where funding for a lot of linux improvements come from, we should be quite careful in how much we rely on user counts. Some companies such as Pixar have with substantial user counts we won’t ever see in GitHub issues, etc. Yet Pixar has a huge impact on graphics card vendors even being interested in supporting linux at all.

    4. 7

      As a developper of small X11 tools, there are a few things bugging me with Wayland, even I would love to see it replace X11:

      Developing apps for Wayland is much more involved than with X11, even with wlroots, putting the barrier to entry way higher than it was with X11 ans libx11/libxcb.

      Wayland advertises itself as a “set of protocols” rather than a single monolith reference implementation like Xorg, but unless you are the KDE team, your only option is wlroots because there’s just too much stuff to implement to draw a rectangle on screen.

      In the end, Xorg is just a protocol as well, but it’s so complex that all the effort was put in a single implementation, and as I see it, that’s what is happening with Wayland and wlroots as well.

      I personally wrote a window manager for X11, and it’s dead simple. I tried doing the same with Wayland and had to give up, because wlroots was moving too fast for me to keep up with it and adapt to the API (because refractors to accommodate newer protocols) and there is no other viable alternative to it, because of how involved the development is.

    5. 7

      My biggest fear is that, before Wayland is “ready” and X11 is “deprecated”, there will be the third thing (or there already is?), which is “gonna be better than Wayland”, and “finally fix everything”.

      1. 9

        We’re past that. That was Canonical’s Mir. That’s Arcan. Yet, the future is still Wayland.

        1. 6

          Why is the future still Wayland? What does Wayland do fundamentally better than Arcan?

          1. 10

            Because all major desktop environments and distributions target Wayland, and Arcan isn’t anywhere near that level of adoption.

            As far as I see*, Arcan is available in AUR, Guix, Nixpkgs, FreeBSD Ports, Void Linux, and Slackware.

            Considering it’s been around since ~2013-ish, so its only a few years younger than Wayland, it does not show any kind of adoption levels comparable to Wayland, and I don’t see why that would change. Thus, it’s not the future.

            The future is not always the thing that’s technically the most sound, but the thing that people end up using, and that’s Wayland.

            1. 4

              but the thing that people end up using, and that’s Wayland

              It most likely will be, but isn’t yet.

              1. 4

                That is why it is the future, not the present.

          2. 9

            Not sure about fundamentally, but Wayland can run GNOME and KDE.

            Like it or not, that’s the bar for anything that’s going to replace X or Wayland. If regular users have to learn an entirely new DE, it’s not happening.

      2. 8

        That’s essentially inevitable. 20 years on, The CADT Model remains the best empirical description of the Linux end-user ecosystem in action

    6. 7

      I run dwm and dmenu over X.org. I tried switching and it didn’t go great. I’ve read over and over that dev momentum has shifted, but I still don’t have problems with X.org in its current state.

      I’m in a similar place with PipeWire: to-dos to check again a year from now.

    7. 5

      So there are two things blocking me right now from switching to Wayland on my Linux and FreeBSD devices:

      • StumpWM (although, I think Mahogany would be a workable replacement).
      • Truple, which we use for intermittent screenshots of our childrens’ machines (ThinkPad X250s running Linux Mint, and also Pixel 2s running Lineage OS).

      Word from the Truple folks is that Wayland screenshot support isn’t ready for them to support it, yet.

      I’d also like to be able to run xscreensaver. jwz has ruled out working on Wayland support himself so I don’t think this will ever happen ;(

      1. 2

        I am coming from a similar setup (currently EXWM but used StumpWM for many years before) and spent some time researching various Wayland options: Mahogany, River, Cagebreak . Unfortunately none of them really comes close to replacing the functionality of either EXWM or Stump. Even simple things like sending synthetic keys or even just passing through WM keys when needed seems to be a major headache and require ridiculous workarounds like defining a fake uinput device and hope that the input reaches the right window.

        I have spent many hours trying to figure out something workable but for now it seems like I am sticking with X11 as long as I can. From my perspective Wayland is a major headache with zero benefits so far.

        1. 1

          Yup, that’s my take too.

          It’s why I get so angry reading dismissive rants like the one posted here a while ago, about Wayland breaking bad tools.

          Wayland isn’t ready for prime time, yet. Agreed that it’s the future - as this article claims - but the future is still a ways away.

    8. 6

      I will never switch to Wayland and will never port my applications to use it. I’d sooner switch to Windows and just drop Linux support entirely. (Though odds are it won’t come to that, as X actually works really very well despite, or perhaps thanks to, its relative lack of git activity.)

      1. 13

        There’s a narrow line between pragmatism and dogmatism, this comment (especially without elaborating on why) seems to veer heavily into the latter. What reasons would you “rather switch to Windows and just drop Linux support entirely”, or do you just enjoy screwing over users who don’t use the technologies you like? Is there anything actionable devs can take from your stances, or are you just venting?

        Comments like this do little but spark more vitriol in what is already a contentious debate.

        1. 12

          (especially without elaborating on why)

          I’m trying not to beat the dead horse - I’ve gone into why in many comments on other threads, and this one is a specific call to action to port things.

          Porting things to Wayland is actually an enormous amount of work. Especially for me personally as a user, since it means doing something about my window manager, my beloved taskbar, every little detail of my workflow in addition to my application toolkit…. all for negative benefit; things work worse on Wayland than they already do today on X. And this situation is not likely to change for a long time; indeed, I expect X will continue to actually work adequately for ages.

          Open source works best when people act out of their own rational self interest then share the result with others because it costs them near nothing to do so - I made this for me, then shared it with the hopes that it might be useful but without warranty of any kind, etc.

          Switching to Wayland costs me a lot. What does it get me? And the constant streams of falsehoods out of Wayland proponents irks me to such a point where I don’t want to give them anything. Maybe if they started telling the truth instead just trying to shove this junk down my throat, I’d give them some slack. But zero respect has been earned.

          1. 3

            Your arguments about needing to change your window manager, taskbar, etc. are reasonable (desktop environment users can mostly be transparently migrated over, those of us on tilers had to change a lot), and yes, there’s development costs. But the same can be said about keeping up with updates to, say, core libraries (GTK updates, OpenSSL updates, whatever). Keeping software running as times and tools evolve is always going to be work.

            (tangent below)

            I will say, “things work worse on Wayland than they already do today on X” is flat-out false for my usecases and something I hear parroted constantly and hasn’t, since 2018 when I switched to Wayland full-time (back when it was beta-ware at best), been true for my usecases at all. I run (sometimes, at least) mixed DPI monitors (i.e. a normal DPI external monitor at 100% scale, but a laptop panel at 125-200% scale), which is something Xorg notoriously can’t handle in a reasonable way (there’s toolkit-level hacks, if I’m willing to only use QT apps, which while I wish I could, Firefox and all Electron apps are GTK). Every time I play with Xorg again to see what folks are on about, I have so much screen tearing and flickering and general glitching that it distracts me from watching videos or playing games. Last I checked (this may have changed), Firefox didn’t support VA-API hardware acceleration for YouTube videos on Xorg, only on Wayland, which directly costs me CPU cycles and thus battery life on portable devices. Xorg (and all window managers I’ve ever used on it) allows a window to fully claim control of the screen in a way the window manager can’t override, so if a game selects the wrong fullscreen resolution, I’m stuck - potentially so stuck that I need to run pkill on the process from a VT after cracking out Ctrl-Alt-F2.

            So… I mean, look, I’m glad Xorg works for you, and works for enough other folks that some projects like OpenBSD are devoutly sticking to it. That’s the beauty of open-source, we can use what works for us. But if we’re just sharing anecdotes and throwing words around, there’s my set: Xorg is unbelievably broken to me as an end-user (nevermind as a developer, I’m looking at this solely from the lens of my “normie” usecases). I will accept Wayland’s lack of global key shortcuts and patchy-at-best screensharing abilities (which are slowly improving) over literally any experience Xorg offers these days.

            And so your last paragraph about “constant streams of falsehoods” and “shove this junk down my throat” confuse me. What falsehoods? The Wayland folks say it works, and wow, does it ever, as long as I don’t need a few specific workflows. And if I need those, then Xorg is right there, at least until it runs out of maintainers. Why the virtiol and hate? I see tons of truth in the space; maybe read better/less inflammatory articles? Check out the work emersion and a few others are doing while maintaining wlroots. They seem like fairly straight-shooters to me from anything I’ve read.

            1. 5


              When somebody tells you something about their own experience, I don’t think that counts as “parroting”.

              Why the virtiol and hate?

              Why indeed. I think you’ll miss a trick if you dismiss it. Wayland feels like systemd in this respect.

              They’re also similar in that they both seem to be instantiations of the old “something must be done; this is something; therefore this must be done.”

              1. 2

                For what it’s worth, I pick and choose the technologies that work for me. Wayland yes, systemd no, Pipewire yes, Flatpak absolutely the hell not. I understand where your sentiment in that last sentence comes from, but I’m not sure I ascribe it to all of the modern Linux renovation projects. Some (eg. Pipewire) seem quite well thought-out, some (Wayland) seem… uh, maybe under-specified to start and rushed out, but mostly recoverable, some (systemd) I think had far, far too much scope in a single project, and some I just outright disagree with the UX considerations of (Flatpak). Again: I’m glad we get to pick and choose in this niche, to some degree (you can argue this only really exists on certain distribuitions like Gentoo and Void and not be horribly incorrect…)

            2. 1

              Xorg (and all window managers I’ve ever used on it) allows a window to fully claim control of the screen in a way the window manager can’t override, so if a game selects the wrong fullscreen resolution, I’m stuck - potentially so stuck that I need to run pkill on the process from a VT after cracking out Ctrl-Alt-F2.

              I don’t believe that’s true. For example twm keeps full screen clients in their own window which you can resize and move around.

              1. 1

                This would be a pleasant surprise of a thing to learn that I’m wrong about. It might also vary game to game? I seem to recall some games taking exclusive control of the rendering plane, but if a WM is capable of fixing that, then that solves that point to be equal with my Wayland experience (where “fullscreen” is a lie the compositor makes to the client, and I can always Super-F my way back down to a tiled window)

                1. 1

                  I believe WMs which allow this functionality are not EWMH compliant, but there is nothing forcing a WM to do it one way or another. Try it with twm and see if you experience the pleasant surprise that you hypothesize.

                  1. 1

                    I’m afraid your interlocutor is correct on this point. Under X11, screen locking apps have no special permissions, so it follows that if you screen saver can take over your screen so can any other app. Compliance with EWMH is completely optional for your window manager or any other X11 client.

                    That being said I’ll still be using X11 until I can’t run a modern web browser or Steam with it and or Debian drops it from stable. X11 is a crazy mess, but it’s the crazy mess I know and love.

                    Or maybe I’ll switch to Arcan.

                    1. 1

                      You’re right, clients can bypass the window manager using the override-redirect flag. I wonder if games use that flag in practice – full screen programs generally don’t, but games might use it for better performance.