To Wayland, or not to Wayland

I'd like to start a discussion and gather opinions on why people actually prefer Wayland over Xorg with well-configured compositors (providing vsync, fading, and other features).

As for myself, I use dwm with picom and have also tried dwl. However, given my workflow (terminal-based coding, browsing, etc.), I find it hard to notice any significant advantages of Wayland.
 
I never tried wayland. I don't intend to either. For a long time, I didn't see the need for any experiments, given that Xorg worked perfectly fine for me (and then, lots of experiences you're reading aren't exactly positive, at least that's my probably subjective observation over many years reading forums and similar). I'm also not convinced by the design, I think a "display server" should be capable of doing more than just compositing (fully rendered) windows...

Meanwhile, since I started working on my Xmoji tool, I'm "hooked" on X11 anyways, as I deliberately opted to base it on X11, utterly non-portable to anything else...
 
why people actually prefer Wayland over Xorg
People don't tend to prefer Wayland over Xorg.

It was possibly going to be the alternative Linux display system (in our grandkids lifespan?). But just like systemd, it doesn't really affect FreeBSD.

Have fun with it but don't be a mug forcing yourself to use immature, early adopter software if you have to struggle unnecessarily. Think of Wayland today being more similar to X10 compared to X11. A little incomplete.
 
Have fun with it but don't be a mug forcing yourself to use immature, early adopter software if you have to struggle unnecessarily. Think of Wayland today being more similar to X10 compared to X11. A little incomplete.
I've never understood why this particular topic resonates so much. Why not call it just the alternative? There is a lot of software out there, so why does Wayland deserve this attention from people who use Xorg and have no plans to change that?

I've been using tiling WMs exclusively for about 10 years, and mostly it was i3 under Xorg. About a year ago I switched to Sway just because there was some Wayland-native software I wanted to try. It took me little to no time to adjust, and like jb82, I still don't see much of a difference as a result. But is this a bad thing, or a good thing? I had no issues with Xorg, so for me, no difference is a positive.

Of course, if you need some missing features or don't really like the design, then there is no point to use Wayland. Fortunately, FreeBSD is not a mainstream Linux distro and comes without any graphical environment, so users are free to choose the one they prefer. I hope it stays that way.
 
I want Wayland to be just a short relief until firmwares (future UEFI or brand-new successors) virtualizes ALL hardwares and their runtime service become responsible for functionalities something between Wayland and X with strictly standardized way for ALL OS'es running on them, of course including Windows.
 
eseipi, I suspect that part of the problem is that RedHat, (the Linux equivalent of Microsoft to my mind), is planning to scrap X for Wayland completely, and it may be like systemd, where they force it upon most OSS software.
The only advantage I've seen is on a laptop connnected to a TV through HDMI, Wayland seems to be better for running videos on a couple of Linux installations. VERY subjectively, it seems a bit better for video even on my workstation.
But for me, there's no compelling reason to switch, and I lose one or two things, for example, I never did figure out how to get the date working in dwlb (the equivalent of dwm's status bar, where one can get the date working with xsetroot), nor was I able to get waybar,
sort of similar to tint2, to work as well as tint2, and I have some problems with Japanese input (which I use less and less frequently as time goes on).

So, for MY use case, and no one else's, there are reasons not to switch, save if I'm going to sit down and watch a movie. I mainly wanted to become sort of familiar with it, in case the sky actually falls and X disappears.
On the other hand, I'm pretty old, so it probably won't happen in my lifetime. :)
 
for example, I never did figure out how to get the date working in dwlb (the equivalent of dwm's status bar, where one can get the date working with xsetroot)
As for the dwlb, I use any_stdout_output | dwlb -status-stdin all. It connects automatically via BSD sockets. There's possibility of using more involved IPC between dwl and dwlb but this is good enough for me. I have a custom script running in the background sending there my custom status line once per 5 seconds.

dwl I start as:

Code:
dwl -s 'dwlb -font "jetbrains mono nerd font:size=13" -bottom -no-ipc' > dwl.log 2> dwl.err
 
I've never understood why this particular topic resonates so much. Why not call it just the alternative?
Exactly. It is not a standard. Just like systemd, it is just an alternative. The problem is you have some real strong "debaters" in the Linux community pushing it as a standard for every platform which of course is nonsense.
Of course, if you need some missing features or don't really like the design, then there is no point to use Wayland.
I wish people would do this more. I see so many beginners struggling with Wayland because they "think" it is the modern replacement for X11 and yet it is still missing some real basic stuff. Most compositors lack an xrandr or xbacklight equivalent for example.

I suspect that part of the problem is that RedHat, (the Linux equivalent of Microsoft to my mind), is planning to scrap X for Wayland completely
Everything with Red Hat these days is a half-truth. Their focus on Xorg has only slightly changed. The Gnome XWayland technology still pulls in 75%+ of Xorg. They know that their clients are still very reliant on X11; they just want to make sure everyone is now vendor locked into their Gnome compositor ecosystem for the desktop and systemd for the init system.
 
I suspect that part of the problem is that RedHat, (the Linux equivalent of Microsoft to my mind), is planning to scrap X for Wayland completely
Well, in the Linux world Red Hat is basically the only remaining maintainer of Xorg and they decided that RHEL 10 will be Wayland only. It's their decision, no one is "planning to scrap" anything as Xorg is open source and anyone willing to adopt/maintain it is welcome to do it.

The real problem is that there's not many pople queuing for the job...
 
if you need some missing features or don't really like the design, then there is no point to use Wayland. Fortunately, FreeBSD is not a mainstream Linux distro and comes without any graphical environment, so users are free to choose the one they prefer. I hope it stays that way.
Right. XFCE is still Xorg only -> Wayland is a no-go for me. Simple as that.
 
I am still undecided which one I prefer. On my Linux laptop I prefer Wayland because the scaling looks better and I like the smooth touchpad gestures when switching virtual desktops.

On my FreeBSD Desktop I am trying out both a barebones Xorg and Wayland to see which one works better for my use case. Xorg works perfectly with "nvidia-driver", but Wayland needs "nvidia-drm".

If I had to choose between the two, I would side with Wayland just because it's newer software and might eventually replace Xorg, as people mentioned above corporate support dropping Xorg. Better to learn a new software in depth than start learning Xorg that could sadly phase away.
 
Well, in the Linux world Red Hat is basically the only remaining maintainer of Xorg and they decided that RHEL 10 will be Wayland only.
This isn't correct. Xorg is maintained by a number of individuals and companies looking at the recent commit logs.

RHEL 10 will be Wayland and XWayland only. There is a difference. Basically it is a glorified backend change for X11. For example, they will still ship X11-only libraries to keep their existing enterprise clients happy.
 
Exactly. It is not a standard. Just like systemd, it is just an alternative.
So why bother then? Another piece of software that is actively being worked on. Sounds like a good thing to me, whether you personally intend to use it or not.
The problem is you have some real strong "debaters" in the Linux community pushing it as a standard for every platform which of course is nonsense.
I'm a little confused: are you saying that we (FreeBSD community) won't be affected by this new shiny thing called Wayland and that users don't tend to prefer it over Xorg anyway, or that we need to advocate for Xorg because the Linux community is pushing Wayland (the protocol) as a standard and we could end up without a choice? Personally, I don't agree with either statement, and this is likely why I find this defensive bikeshedding hard to understand.
I see so many beginners struggling with Wayland because they "think" it is the modern replacement for X11 and yet it is still missing some real basic stuff.
All my stuff works perfectly fine, so I can't relate. I also didn't find Xorg any easier to use, even despite having much more experience with it.
 
jb82 many thanks. That actually enable me to get a dwm like bar. I start dwl with

exec dbus-launch --exit-with-session dwl -s 'dwlb -font "monospace:size=12"' 2>/dev/null


Then, after it's started I can run something like you mention, in my case

date "+%R %F" |dwlb -status-stdin all

which gives me a bar that works like dwm's. I've not yet figured out how to get the bar to open when I start dwl with both font and status commands working at the same time, but I just started playing with this a few minutes ago. And if worst came to worse, where I'd have to start it as I do, then type in the status part afterwards, I can live with that.
Doing it this way, the bar seems to do what dwm's does, mention each program when focused on the program and showing which tag I'm focused on. (Not going to bother mentioning each false step--just going to say that sometimes I could get one thing working but then the bar wouldn't show the opened programs or tags).
Anyway, thanks again.
As NapoleonWils0n and I have discussed this in the past, naming him here so he sees this post.
 
Another piece of software that is actively being worked on. Sounds like a good thing to me, whether you personally intend to use it or not.
Yep, this is all it is and it is a good thing. However if you state that in a i.e Linux forum, you will find that many of them are religiously adamant that Wayland is the future of all UNIX-like operating systems and that Xorg is "dead". This of course is not the case which we seem to agree on.

I'm a little confused: are you saying that we (FreeBSD community) won't be affected by this new shiny thing called Wayland
Yep. In the same way we aren't really affected by systemd. Sure, we have compat shims for it if we choose to use it but FreeBSD doesn't need to follow Linux. It never has.

the Linux community is pushing Wayland (the protocol) as a standard and we could end up without a choice
This is often pushed as tiresome FUD and is the main thing that beginners often need explaining (or reassuring). Wayland isn't a replacement for Xorg. And really it goes in circles to the first quote. Great that they both exist but it doesn't really affect Xorg.

All my stuff works perfectly fine, so I can't relate. I also didn't find Xorg any easier to use, even despite having much more experience with it.
It depends on the compositor. Unlike the design of Xorg, they all need to duplicate work and many don't keep up. It is a flawed design and I hope the Wayland ecosystem does better in future as it matures. I'm sure it will. X10 was crap too.
 
To add to the above. I've found this is how I can get a constant date in dwlb. As I said, I start dwlb specifying font size when I start dwl. Then, I make a little script.
#!/bin/sh (while true; do date +"%R %F";sleep 60;done &)|dwlb -status-stdin all
I then run the script using wemenu, and it works perfectly for me.
 
Back
Top