Thread 105772989

So when will Wayland finally just "work?"

>>105772989 (OP)
Works on my machine

>>105772993
It really doesn't work on your machine anon.

Works on my machine (KDE Plasma 6.3.5 with AMD GPU)

>>105772989 (OP)
It works fine on non meme DEs

>>105773124
>ktard
pottery

>>105772989 (OP)
In 20 years.
Works on my machine better than X11 for most things (not gaming).

>>105772989 (OP)
Wayland does work, just not for your obscure use case.

>>105773610
literal lies

>>105773628
Not lying anon. I do actually run it and am not larping. Hyprland.

>>105772989 (OP)
What the fuck does that even mean? Why is the Indian call center now pushing this shit? Do they think they're subtle.

>>105773110
>works on my machine
>NO IT DOESN'T I BET YOU USE KDE LOL KRASHES
Kill yourself

>>105772989 (OP)
ironic, because wayland is one of the few things on linux that actually does "just work"

>>105772989 (OP)
2 more weeks

>>105773722
Did you remember to pay your subscription

>try wayland
>all my current utilities are broken
>clipboard barely works
>everything feels slow as fuck and the mouse feels like it's being dragged through mud
>look it up, apparently it's a feature of the protocol
>log back into DWM, buttery smooth once again
idk what the point of this joke protocol is, you'll have to pry x11 from my cold dead hands

>>105773163
>it works
NO NOT LIKE THAT

>>105772989 (OP)
Wayland is already perfect and feature complete.
inb4
>B-B-BUT MUH APP DOES NOT WOR--
Shut the fuck up. That is a problem with your app that YOU need to fix. Wayland is already perfect.

>>105774980
> I love lying
>>105774994
It's not perfect but the issues come from GNOME as always.

>>105773163
Why is Plasma hated again? Is it because you are expected on /g/ to rice your PC until it has nothing in common with Windows?

>>105774702
Global hotkeys?
>>105774994
And there goes Wayland's promised backwards compatibility...

>>105774996
GNOME is the champion of desktop linux. GNOME is opinionated and it is perfect that way.
>B-B-BUT I HAVE AN ISSUE ON GN-
Shut the fuck up. GNOME is opinionated and your opinions do not matter. Don't get cocky. Know your place.
>B-BUT I
No. Shut the fuck up.

>>105775032
YOU WILL HAVE USECASES
YOU WILL HAVE CUSTOMIZATION
YOU WILL HAVE QT
YOU WILL HAVE SERVER SIDE DECORATIONS

>>105775000
>And there goes Wayland's promised backwards compatibility...
all wayland protocols are versioned
If you desktop supports the protocols it works.
Just don't be GNOME.

>>105775077
No, thank you. Qt requires you to sign an evil CLA to contribute code to it. You're a consolewarring retard who doesn't know anything. RMS would be be ashamed of you.

>>105775000
>Global hotkeys?
Works on Hyprland™

>>105775104
Damn. Hyprland chads...I kneel. Truly the non-pozzed Wayland experience.

>>105774996
I don't know what to tell you retard. You want me to replace my ~10 xorg-specific programs, convert all of my configuration for these programs, into their wayland copycats for literally 0 gain. Not to mention that my window manager migration would give me a wm that literally just doesn't have the same stability or features.
>no screen tear
already fixed it years ago with the common x11 tweaks for it
>different monitor refresh rates
meme feature
>no snooping on other windows
xkcd 1200. Cookie stealing is the only scary hacking.

It's so fucking stupid how many man-hours has been wasted breaking the entire linux desktop for 20 years just for a marginally theoretically better replacement for x11. I am happy where I am. I wouldn't give a shit about you wayland faggots if you weren't so insufferable about x11 being dead. If wayland works for you that's great, but don't moralize to me about being a bad person because I want shit to work instead of having shit not work.

For some reason waylanders OK to specify on which display to show the window, but only if it's full screen window.
Apparently, there fullscreen windows have special treatment among waylanders.

here's your blazingly fast wayland wm bro. Is this your champion, trannies?

>>105775095
>Qt requires you to sign an evil CLA to contribute code to it.
Which says all contributions must be LGPL v2.1
And your access to the source code is protected by the KDE Free Qt Foundation
> You're a consolewarring retard who doesn't know anything
PC
>RMS would be be ashamed of you.
That's a badge of honor, i don't want to be associated with communist autists
>>105775126
>I don't know what to tell you retard. You want me to replace my ~10 xorg-specific programs, convert all of my configuration for these programs, into their wayland copycats for literally 0 gain. Not to mention that my window manager migration would give me a wm that literally just doesn't have the same stability or features.
Most of the functionality is now embedded into compositors. You don't need a million small programs
>already fixed it years ago with the common x11 tweaks for it
let me just say pico is shit
>different monitor refresh rates
and scaling without it being broken
Also how is it a meme feature
You have different monitors, they run at different refresh rates. One of them might be high refresh rate, the other one might be high res.
>xkcd 1200. Cookie stealing is the only scary hacking.
I literally don't care about that either, but at least due to that we can now have proper standardized ways of handling things like device permissions and file permissions now.


This all sounds like cope.

>>105775160
>there's an optional subscription to gain access to configuration files for hyprland and priority issue support
the horror
we can't be homosexual commies anymore that drive the Linux desktop to the ground because vaxry has an optional subscription.

RMS would call you faggots btw. Only complaining about the fact that the payment processor probably uses proprietary tech.

>>105772989 (OP)
10-15 years but then they will start replacing it with something else.

>>105772989 (OP)
zero issues with wayland on fedora + kde seethe

>>105772989 (OP)
Is that Vanilla Ice?

>>105775182
>future Premium-only services.
sorry this feature is only available for premium users.

>>105775163
>Most of the functionality is now embedded into compositors. You don't need a million small programs
https://arewewaylandyet.com/ you are literally retarded you can just read the list yourself. dmenu, polybar, slock, xrandr, scrot, all killed by the tranny cabal
>why is it a meme feature
getting to 60hz is a reasonable upgrade but going beyond 60hz is a hedonic treadmill. I have 165hz monitors I'm intentionally capping at 60 when I'm dual booted on windows.
>This all sounds like cope.
applying critical thinking to propaganda is not "cope" you dumb redditor go back

>>105775285
>dmenu, polybar, slock, xrandr, scrot, all killed by the tranny cabal
Oh no. Anyway I'll just use one of the 100 alternatives

>>105772989 (OP)
Wayland always works, and if something doesn't work it doesn't work BY DESIGN, fix your expectations chud.

>>105775285
>https://arewewaylandyet.com/
Arewewaylandyet has not been updated since 2022 anon
https://wearewaylandnow.com/
The rest of this is cope

>>105775529
The extra steps for nvidia aren't needed anymore

>>105775576
that is also true.
nvidia has relative first class support for wayland now. Ever since AI became big they invested majorly on improving drivers for linux.

>>105773163
>noooo you cant use the best wayland compositor

>>105773163
>troontard
ACK

>>105775032
>GNOME is the champion of deskt-
sis, gnome is a tablet user interface

>>105772989 (OP)
>fags(X11) vs troons (Wayland) ITT
I need the chuddiness of Xlibre and the just werks (on my machine) of wayland in a single project. I NEED CHUDLAND!

>>105774695
what anon meant (i assume) is that (You) dont actually work on your machine.

>>105774997
Literally everything is hated in /g/. This board is just a bunch of mixed raced American claiming whatever they don't use is self evidently jeet/jew/tranny.

>>105777409
Xlibre has yet to do anything as a project and motherfuckers are already pretending like it's relevant in anyway.

>>105777834
Dunno, I just want trannies to drop dead, that is all.

>>105777834
it's been out for a week and it already works better than xorg

>>105777834
Being a new release that contains all the xorg master changes is good enough on its own honestly.

>>105777844
Killing half the Linux user base. In fact I wouldn't be surprised if Xlibre has proportionally more trannies than x11

>>105777877
I'm trans and I'm unaffected by your hate speech.
X11 is the protocol, Xlibre is the implementation.
Learn the terms before you try to talk shit.

>>105777877
It's probably .01% women and 10% men larping as women, but they are so fucking obnoxious and annoying that I YEARN for total tranny death.
>>105777901
excellent bait

by the way nobody is asking you to use wayland
we're just saying it works for us
keep using x11 if you want

>>105774994
no it's not, go fix your shit wayland dev
I still get insane screen tearing on my intel igpu laptop

>>105775125
as long as you don't forget to pay your subscription

How long does it take to even make a display system? Apple did it with Quartz in just a few years of developing MacOS (X), and they transitioned it across 3 cpu architectures too.

>>105772989 (OP)
niggerland is troonware and they froze X11 development to push their own diseased software to everyone they can, it will never work correctly

>>105778245
>nobody is asking you to use wayland
lol

>>105780599
but unironically. security and performance are negatively correlated... causally

>>105772989 (OP)
Never.

Team #xlibre until #arcan ftw!

>>105778746
> how hard is it
look at the arcan guys blog about that. the answer is its not at all difficult to make functional platform components, the hard part is convincing app devs and systems integrators to use them. so you allocate all your resources into getting systems integrators to use your components, make them barely usable for app devs, dont give a shit about end user experience, and secure the bag. thats how software works. when your company shrinks, the people laid off first are your most talented engineers. the javascript creator was fired

schizo wayland thread

We need poettering to make his own so everyone adopts it regardless of it is any good or not.
And probably will be.

>>105774997
historically kde has had issues with krashes, in particular starting with plasma 4, honestly it was pretty bad until plasma ~5.10? , things started getting alot better when valve decided to use it for steam os, now with plasma 6.3/6.4 its rare to encounter a bug, 6.4 really improved the wayland session quite a bit, still some small issues which are mostly protocol related, but overall very useable and stable.

>>105781128
>plasma 4
This is really it, Plasma 4 was a fucking shitshow, devs lying about shit and just completely changing everything and getting rid of old functionality for no purpose, and not listening to the users. It was released well before being ready and was a crashy bug riddled piece of crap. Devs also tried GNOME tier tyranny and not allowing users to have things the way they wanted (re that stupid fucking cashew, certain design decisions on the main theme, etc they literally even claimed that users changing things on their own computers was bad because then the graphic designer knowing that people made things different than his own preference would be sad )-: Also the cashew was non-removable ebcause what if some mythical totally new to computers users happens upon your computer?? How will they know what to do?? seriously.). Sorta like Wayland really.

Here’s an anecdote of the most recent time I tried wayland. It was on KDE plasma 6, tried using the X and wayland versions and the wayland version barely worked at all and looked all fucked up. I’ve also used wayland with gnome, sway and hyprland on a few different machines and have had heaps of issues that I just don’t have with X.

>>105780480
It's true
If you're going to cry about big corpos not wanting to develop X11 anymore, well, that's their choice
>>105780639
Trvke, except for seL4 which achieves both
>>105781223
What GPU and display? Very interesting anecdote

>>105772989 (OP)
two more decades
LMAOOOOOOOOOOOOOOOO

has any retard around here ever elaborated on why Wayland sucks ? All I know is that it seems to work for me.

>>105774702
SimpleScreenRecorder? I'm not gonna learn OBS to make some screencaps.

>>105781301
nta but wf-recorder works very well for me. i suppose you could bind it to some key and there's probably already some gui wrapper out there

>>105777834
>Xnest is nothing
>countless security and bug fixes are nothing
>being a release after 5 years is nothing

>>105781314
>i suppose you could bind it to some key
I won't financially doxx myself to use hyprland premium, I'll just use xlibre instead

>>105781337
What? lol
I run Hyprland just fine without a premium subscription or whatever

>>105781273
2020 Thinkpad E15, desktop computer with 1060ti + 1920x1080 shitty display and thinkpad t420. Also tried screen recording with obs using all the bullshit workarounds you’re apparently supposed to do and got like 5fps recording the desktop where x wouldn’t have that issue at all so I have no idea how people don’t seem to have issues with it

>>105781355
>1060ti
Rip, that's basically right before the transition period where NVIDIA started supporting Wayland well. For you I would honestly just keep using X11 with the proprietary drivers.

>>105781345
I still don't have a reason to use gayland, so I won't.

It just needs 5 more years of bikeshedding protocols and it will finally be good.

>>105781298
There's multiple aspects here that we could go into. But at its core, the failure of wayland is that it is not truly a replacement for xorg. The original developers built a bunch of highly opinionated decisions into the core design which has taken over a decade to slowly walk back on as they stubbornly admit that they were wrong. For a concrete example, compositors used to suspend the rendering thread of applications that were not visible. This might sound OK but no other platform does this and many applications expect to do work in the rendering thread regardless whether it is visible or not (games commonly do backpressure in the render thread for example). The correct approach would have been to not cripple applications but simply given them visibility events so they could make the correct decisions and optimize when it made sense. Anyways, there were many years of arguments over this topic and upstream wayland often landed somewhere along the lines of "your application sucks and is programmed badly". Well but they finally relented and walked this back earlier this year with the fifo and commit-timing protocols which implement vsync blocking in a way that doesn't stall the application. I think currently it only works with vulkan but at least it's a start. If they had just simply not chosen a stubborn, opinionated design like this in the first place, we wouldn't have wasted a decade arguing how to fix it. And even then, the fix is overly complicated and messy. That's how a lot of issues in wayland are like.

>>105782566
the success of wayland is that men have built careers out of it, and other men have spent their careers recommending it and getting social credit from recommending it. both were needed for wayland to succeed. building good software is a useless skill for grass eaters, only politics matters

API can't be wrong or bad. people who blame "wayland" are actually doing people behind it a favor. now they can hand wave all criticism as coming from nazis and chuds.

blaming "wayland", red hat etc is the software version of anti-semitism.

>>105782566
so why don't KDE for example send these visibility events? is the "wayland" holding the entire development team hostage?

>>105782660
That's true. In a sense, I don't even blame some of them. It's a lot more fun to start from scratch on something that doesn't work at all and make it functional e.g. like what a lot of compositor developers did. Plus it looks better on your resume. I'm in a similar boat myself where I added wayland support to a client which obviously required way more work than the xorg backend which was mature and needed nothing more than occasional fixes or tweaks for refactoring. But if we're talking about well-designed windowing platforms, wayland/wayland compositors are absolutely not that. It is a shift away from "traditional" windowing systems and I see zero evidence that the direction that wayland chose is any better than the old way.

>>105772989 (OP)
doesn't work on my machine (because it uses macOS)

>>105772989 (OP)
Wayland works perfectly if you're not an nvidiot and you use real DEs like KDE or GNOME.

>>105782757
It's a problem with the core protocol and goes beyond merely KDE. Graphics APIs (opengl and vulkan in the case of linux) all have someway to render in sync with your monitors vblank. Games often will turn this off, but in many cases this behavior is quite desirable (like watching videos). It differs a little but opengl and vulkan both have abstractions to windowing backends to handle this since the display server obviously needs to be involved (x11, win32, wayland, etc.) So this means there's inherently a layer somewhere which is a communication with the display server and the graphics API. For linux users, this is mesa most of time unless you use NVIDIA (it's their blob in that case and who knows what it does). All that is to say that the implementation of such graphics APIs in userspace lies there in mesa. So mesa is forced to use wayland APIs to communicate in a way to accomplish what developers normally expect if you call eglSwapBuffers or what have you on wayland.

So back to the specific thing I'm talking about, a very basic task for any opengl/vulkan implementation is to be able to render while synced to vblank. Before this year, the only way to do this in wayland was for mesa to for frame callbacks from the compositor so it would know that it it is safe to flip the buffer. Obviously, there were presentation timestamps and other such things but fundamentally there was always a hard block in mesa on waiting for frame callbacks. The extra bad thing about this is that it an indefinite block that stalls the thread because compositors didn't send frame callbacks unless the surface was visible. You could say "well why don't they just always send them". They could but it would break applications that do actually expect this behavior but also realistically compositor developers weren't going to do this because they never wanted to. Again, it took over a decade to even admit this was a flaw and I think it only works on vulkan right now.

>>105782796
right. the rule is if you have skills, you do something that hasnt been done before, if you dont, you redo something that has been done before badly

>>105782940
And just to reiterate. Wayland is full of self-inflicted flaws like this. The whole mess with scaling is another good example. It's fixed now with a couple of protocols that essentially bypass the core behavior. As the usual pattern with wayland goes, it took many years of arguing before acknowledging it was an issue and then some more years to fix it. Well designed systems simply wouldn't have such problems in the first place.

>wayland
>well-designed
even x11 did a better job of that

>>105782940
no, it's a problem with linux developers, because most of them endorse this attitude, as the path that should be taken to develop software.

visibility events could be implemented in KDE yesterday. they could say "fuck this protocols, we want to make cool working software". and they could go to the SDL repo and open issue: "listen guys, we have this cool feature in KDE which is beneficial for all games, merge ASAP". but none of that is happened.

>>105783043
Seems like a pointless semantics argument. Either way, it is working around deficiencies in the core design which is the fundamental point. btw, they would need to open an MR in mesa.

>>105773163
I also choose to use systemd and systemd-boot on Gentoo with Plasma :^)

>>105772989 (OP)
Just werks for me on Ubuntu.

>>105782940
>mesa is forced to use wayland APIs
forced by whom? linus torvalds himself? why can't they just expose that API in a non-retarded way? this has nothing to do with "wayland"

>>105783222
how do you expect to communicate with the display server without using its api

apparently, MESA and KDE refusing to implement much needed features is a "wayland" problem. interesting...

>>105783252
What are you trying to argue? Do you even understand the problem? The client API of wayland is crippled. mesa/KDE/whoever didn't come up custom protocols to say "fuck you" to upstream (a valve guy did though) so I guess if you want you can blame them for that. But none of that changes the fact that wayland upstream, which includes official protocols/extensions, was crippled and deliberately so.

>>105783231
are MESA need a seal of approval from Sebastian Wick personally to implement anything?

Wayland cursor input is very high latency and feels awful. It's just a design flaw. It's honestly unusable.

>>105783329
The crux of the issue is that it was impossible to make a good graphics API implementation in wayland because the wayland client API is limited. This fact alone makes it a wayland problem because the display server is not capable enough to do things other display servers can do (xorg and win32 don't have these problems). Theoretically, mesa could have come up with some custom protocol, implement and hope other compositor developers adopt it. Obviously such a protocol would be pointless if no one used it so for example since Sebastian Wick is a GNOME developer, it would have needed his seal of approval to work in GNOME at least. Regardless of your opinion on that, it doesn't change the fundamentals of upstream wayland being limited. In fairness, there is some overlap between mesa members and wayland members (notably Daniel Stone) so feel free to give them shit. I'm not out to defend their integrity or something.

>>105781128
>>105781216
I'm a Debian user and used to hate plasma for this reason, my only experiences with it were probably in debian 8 and 9, which might've still been on plasma 4. I stayed away from it for a while since then.
Recently had issues on my nvidia laptop with guhnome, and didn't want xfce4 on my gayming PC because I wanted something that looked and felt premium (I use xfce4 all the time on my thinkpad so don't accuse me of having poor taste, it's my favorite desktop, just didn't work in this instance)
Ended up with Plasma+Wayland and it actually works. Extremely stable, In fact I didn't even know I was using Wayland until I started tinkering a few weeks into playing games. Using debian 13

>>105783252
>KDE
it's usually GNOME, to be honest. KDE seems to have implemented shit even before they got merged into Wayland. it's usually GNOME and specifically Sebastian Wick that veto PRs

>>105783394
unironically an issue with your setup
name your wm or de, gpu vendor, kernel version, and distribution

Ok so here’s what is going on with any of these complaints of certain systems or software here spammed

In each complaint one choice is inherently insecure and the other secure and certain people want everyone to use insecure software so they can infiltrate systems

That’s all this is and ever will be, trying to trick people
Since most people here are low ranking or unemployed it seems to be a waste but they have to do it everywhere otherwise a community could foster against this tactic

Every topic on 4chan is like this
If it’s good for people it’s derided and if it’s bad for people it’s praised

We all know the main players and their countries and simps

All they needed was simp staff here and they have that

>>105783495
then, i'm afraid, it's a (GNOME) problem

>>105783676
False, Wayland has forced vsync for the cursor resulting up to 33ms of additional latency over X11 outside of full screen tearing mode. Desktop mode is always vsync'd.

>>105783832
33ms would be the delay at 30Hz you retarded faggot, stop being dumb (1/30)*1000

X11 is more complex so there’s probably more holes
Wayland is not so there’s less
That’s why everyone is complaining
All this other shit is a distraction for autismos

>>105783863
That's correct. Wayland can cause up to 2 frame delay on cursor updates. On average it's 1 over X11.

>>105783891
stop using nvidia

>>105783832
this is a compositor problem
https://lobste.rs/s/oxtwre/hard_numbers_wayland_vs_x11_input_latency#c_edq7tn
KWin doesn't have this issue

>>105783744
>inherently secure
there's only contextual security
all security is made out of non-secure elements

>>105784022
>oh no muh cursor might theoretically tear
who the fuck even cares? I don't think I've ever seen the cursor tear on x11.

>>105784036
You will refuse to counter my main arguments which are sound and based on keen observation and will attempt to nitpick on things that are meaningless to the central point

All of this shit is just a full court press to make people choose bad decisions on any topic, go to any thread here in any board. It’s so obvious that pointing it out is almost unnecessary.

I am done talking about it. I don’t need to prove truths.

>>105784102
your foundational premises are wrong so your derivations don't apply to reality.

The problem is that linux baseddevs and troons are afraid to take responsibility to carve out new features and API. They need a daddy figure that is red hat, freedesktop, wayland, whatever, that tells them what they need to implement.

Linux is no longer a hacker culture. It's a culture of pathetic corporate bitches who want to be given orders.

>I need PROOOTOCOLS, please daddy Sebastian Wick give some PROOOTOCOLS to implement!!!
pathetic

>>105785040
linux has been commercial since the late 90s

>>105783990
Not relevant, it's like this for everything. It was a design choice. It's a flawed design. It doesn't fit my use-case. My use-case of having a mouse cursor that doesn't feel like soupy shit.

There's, like, 10 working linux graphical programs (if being generous), and 20 desktop environments that want PROOOOOOTOCOLS for non existing software to conform.
This is seriously pathetic. This is worse than rust game engines situation.

>>105785379
>10
Other than blender, what else?

>>105783891
>WAYLAND IS BROKEN
>Argument comes down to 1 frame difference in cursor updates

lmao