Why do environments using X feel "native" while environments using Wayland feel like a non-native Electron app or JavaScript web environment?

I know you're not supposed to notice a different, but when I switch between the two, that's what I notice.

>>105795341 (OP)
*difference

Not OP but also why would someone make a dedicated Wayland program when he could just make it to X11 and Xwayland with even more functions? Also people try to run entire DEs in XWayland now.

I'm guessing you're going to reply to yourself to talk about the cursor movement being vsynced like anyone gives a fuck

>>105795515
Wait, for real, wayland doesn't have a hardware cursor?

>>105795598
afaik every compositor syncs the cursor plane ot the primary plane, which is basically the same as a software cursor.

use case for selecting an output display for games?

use case for centered splash screen window?
https://gitlab.freedesktop.org/wayland/wayland-protocols/-/issues/67

Because of all the abstraction layers. I first noticed it with Java back in the nineties. It's like being stuck in a sandbox. Basically the polar opposite of a C64, templeos, or even MSDOS. But there are benefits, particularly in stability and security.

use case for protocol for testing use cases?
https://gitlab.freedesktop.org/wayland/wayland-protocols/-/issues/230

>>105795341 (OP)
idk i have never bothered to try gayland hehe
posted from my gnu/arch/x11/i3wm/no DE installed machine
btw i have never used an LLM to code and I refuse to even try

>>105795341 (OP)
what makes you think that "feel" is a metric?

it's the absence of tearing when you move a windows and absence of artifacts when you resize them. it's just feels less real and phony without them.

>>105795341 (OP)
Correct me if I'm wrong, but if I remember correctly, Wayland is drawing frames in some very fucked up way? That and other quirks implemented by devs just because they want to feel quirky and different would probably explain it.

>>105795691
Use case?

What's the use case for protocols? Who's enforcing it?
In OpenGL extensions are developed by vendors, and only when they are confirmed to be widely supported they're added to the core specification. But in wayland it's the other way around. Who makes DE vendors comply with this policy which has been confirmed as non-working countless times?

I remember frog protocols wer big news. But it looks like it didnt change anything at all.

>>105795341 (OP)
>>105795598
Yeah the cursor is one of the more notorious problems with Wayland. On top of it's usual input lag the cursor is also behind by a few frames.

Recently I've gone from being indifferent about Wayland to realizing it's doomed if it goes forward with Wayland. Which it looks like they already have. Not even because of Lunduke or any culture war bs. Wayland is fundamentally broken.
>>105796206
The FreeDesktop organization. In theory it's supposed to be a coalition of desktop developers to find out the best standards that work for everyone. But in practice it's more like Redhat tells GNOME what to do, GNOME impliments it, KDE is left to figure it out on their own and the rest of the desktops can go fuck themselves.
>>105796254
It's a shame the scope of Valve's interest is limited to gaming. They have done so much already but it's only a fraction of what needs to be done.

>the sluggish protocol that drops frames and just does not render some: hahaha this is good software i can feel it
>the correct implementation that was made so it renders every frame correctly: hahaha this is bad software i can feel it
Always do the opposite of what /g/ says

>>105796206
DEs regularly have their own protocols. Sometimes they get reworked to generic protocols and the DE-specific gets discontinued in favor of generic.
Freedesktop is pretty conservative with what they accept as generic protocol and it involves discussion from major DEs, so its unlikely that something insane non-enforacable gets merged.

>Tards itt pretend to notice the input lag of wayland
>At the same time they don't notice the jankiness of everything in X
lmao

>>105796911
I liked Linux PRECISELY it had that jankiness. I came to Linux to ESCAPE the "polished" corpo experience of Windows, where I couldn't really do anything I actually wanted to.
Now the corpo lockdown is encroaching on Linux. Wayland represents everything I came to Linux to get away from in the first place. Except it's even worse, Wayland is even more locked down and cucked than Windows ever was.

>>105795635
god this one pisses me off so bad

>>105796894
this is just not true, otherwise there won't be problems with wayland

does KDE have protocol to select an output display?

>>105797178

>>105797240
there isn't. most problems are just with Gnome.
>does KDE have protocol to select an output display?
yes, KDE has extension for that and very rich window/display management.
Selecting monitor for fullscreen its already in core protocol

Is there a way to create virtual display/monitor with wayland?

How do I add custom resolutions with wayland?

>>105797499
wp_viewporter, or an actual monitor settings

>>105797520
Ok so how do I use it? How do I add the video timings I want?

>>105796894
The difference is that X11 allows workarounds. It's an open platform where people can solve problems, even if jankily, and only then do they need to start caring about DE differences. On Wayland there are many problems that are fundamentally unsolveable, and even simple problems require multiple implementations due to DE differences

>>105797343
not true
https://wayland.app/protocols/
if you search for KDE, there is only "implementation details" protocols for KDE widgets. it's clear they don't attempt to create protocols to solve problems with wayland. instead, they wait for wayland protocols

>there isn't
"if i ignore it, maybe it will go away"

>>105795341 (OP)
because you touch yourself at night

>>105797533
depends on the compositor, gotta love fragmentation

wayland is designed by, implemented by, and advocated by tasteless spastics who don't use computers

>>105797520
don't bother, the only compositor I know of that allows you to set video modes directly like xorg is wayfire, the workaround is to use a custom edid.

>>105796827
>>105796911
you're used to the latency

>>105797499
override the edid entirely, you can use cru with wine to make one
https://drinkybird.net/2024/12/debian-edid/
might be different depending on your distro

>>105796911
In wayland you have to live with the jankiness. Sometimes you even have to change to a completely different wayland compositor which has a completely different way of working and its own tools. Often there are no ways around that. On x11 there is always a way around things. If you want the absolute lowest possible latency just disable compositor and vsync. On wayland tearing option doesn't even work yet (on kde plasma where it exists).

>>105798282
this is os the way. everybody should do this for X too btw.

>>105798282
>>105798774
Fuck no. I'm not rebooting every single time I need to make a new custom resolution. I use CRTs so I can just have whatever resolution I want, but also tuning the timings for optimal performance can take several attempts, which would be completely infeasible with this kind of bullshit.

>>105799843
im also on a crt altough i have most of my timings from when i used windows, but you can hotload the edid on linux https://wiki.archlinux.org/title/Kernel_mode_setting#Forcing_modes_and_EDID

>>105796480
>Wayland is fundamentally broken.
because the cursor is vsynced? you're retarded. name five actual problems with wayland

>>105795635
t. Sebastian "Just use a .desktop file for every unique window icon" Wick.

>>105800658
.desktop files are such a retarded paradigm. Oh, you want to assign an icon to this executable you downloaded that doesn't exist as a package anywhere? Well put some poopee doodoo metadata in this one random directory to map the icon to the executable.

>>105795341 (OP)
Wayland has the same (and can theoretically reach better) latency than X.
You're probably using a DE that does not yet support Wayland properly.

>>105800732
*composited X

>>105795341 (OP)
It's because of the laggy cursor. Wayland has a 1-2 frame delay on cursor updates compared to X11.

>>105800706
>Leaving the metadata in an easily accessed location is worse than just hardcoding it into an immutable binary

>>105800783
Yes, yes it fucking is

>>105800756
Isn't there hardware support in most GPUs for redrawing the cursor without updating everything else on screen? Does Wayland not use this for some reason?

>>105795598
This has always been a bit of a problem under linux, in general.
> be me working in linux shop circa 2008
> everyone (like, 50) worked on one big-ass server over ssh/remote X/samba
> typing in code. Server becomes sluggish. I’m like WTF?
> Go to the server room to investigate.
> retard admin is on the console desktop moving the cursor around annoying 50 other people because we all noticed it.
Yes, it was that bad. We tested it in-depth later.
I think it had something to do with interrupts and the global kernel lock.

As much as I think wayland is a giant turd-polishing exercise by man-children. I’m not sure you can blame wayland exclusively.

Windows servers didn’t have the problem at all. Different one. Admin kept turning on the glpipes screen saver on the VSS server. Had like a VGA card, no hardware graphics.
I nearly burst a vein in frustration before we fired that fuck.

>>105800791
god i wish you people actually got what you were asking for.

>>105800808
Yes video cards support a cursor plane. No Wayland doesn't use it. Kiosk display server not intended for actual desktop use. Flawed architecture. Enjoy your soupy "perfect" frames.

>>105800901
Works fine in Windows.

>>105795341 (OP)
The use-case for a compositor itself is pretty weak.
The hardware support is just not there as far as I know.

The hardware support to run X has been around since the early 90s and maybe even 80s around the time of 8514 graphics adapters (for the PC world, a lot of Xstations were 68000 based)

In the olden days, graphics hardware would often support “sprites” which is not something I think graphics hardware supports really well. Hardware cursors were basically like sprites.

Composition itself to support unusual window overlapping bitmap operations (like alpha, transparancy, and HDR on those) is pretty weak, and only for the “desktop environment” which isn’t even useful in and of itself. I mean generally, you’re working in program like kicad , or a browser or something, or full-screening a video game.

>>105801014
... no it fucking doesnt! you can change an exe's icon that you own and therefore it isnt immutable and therefore can be edited by any process you spawn (THE ENTIRE EXE BTW) or you need administrator privileges...

>>105801060
> you can change an exe's icon that you own
Sounds reasonable. You’re saying it’s *my* icon.

>>105801085
ok? this doesnt work on any computer that has more than one user... which is a default assumption every developer everywhere is obligated to make.

>>105795341 (OP)
feel doesn't matter but wayland is laggy as fuck and on some applications it's basically strangling your PC to keep its view of usability intact or disabling other core parts like the ability to screen record.

>>105795598
looooooooolll hahahahahahah
Wha'ts the use case for a hardware cursor, goy?

>>105795341 (OP)
I like when you change workspaces and everything has to place individually. It's charming.

>>105800998
> a cursor plane
Interesting.
That seems yo be a massive waste of memory if you had high expectations of what the cursor needed to look like. I.e.
- 24-bit color depth
- alpha gradient and transparency
- hdr
That’s at least 32-bits/pixel. For a whole screen sized plane? I doubt that.

Of course, I think the problem is they’re bitbltting the cursor and other elements at all! Should need to do it. But that’s composition.
It would work better with shared gpu/cpu multi-port memory like the playstation 4 had, but I think the idea might have something to do with Nvidia paying so everyone buys their new computer-like APUs.

>>105801115
In windows, it never really moved very well into a multi-user environment.

Although i think you can change the icon, and it makes a .lnk file or something for it. Theres’s a few things like that. And you *own* that extra file that points to the original.
You don’ have to use ‘resedit’ to change the icon.

I agree with the original poster, I installed “wine” and had to clean up a shitload of random icons on the gnome desktop manually and I came away with the feeling that it was insane.

I usually rin things from the app lanucher, snd most of them create a single icon, not 50.

>>105801155
> Should need to do it.
*Shouldn’t

And i mean to say, modern graphics hardware, probably doesn't really support sprites.

It would be cool for retro gaming though if we could get those things back. Like, the extra die space would be less than round-off error nowadays.
And all the patents are likely expired.

I remember playing an emulated xenogears on the PS3. Everything is fine until there was an in-game earthquake, and the PS3’s fans were going to jet the thing off the table re-creating that shaking effect.
Apparently, in the original, they just toggled a vsync bit in the crt display controller chip ir something to create the shaking effect effortlessly, but in the ps3 they had to move and bitblt the whole screen around rapidly.

>>105801212
that anon left out windows shortcuts intentionally and conveniently ignored the most important "immutable binary" part. he is arguing in bad faith and i feel no reason to be anything but belligerent towards him.
you are correct tho. the automatic systems for creating .desktop files are trash.

>>105801292
>that anon left out windows shortcuts intentionally and conveniently ignored the most important "immutable binary" part.
What?
In Windows, the application icon is directly baked in the exe.

>>105801323
You can have it choose and override it with any other icon, even one in any other .exe or .dll
Since windows 3.11 at least.

>>105801323
yes........ and you as a user cannot change it unless you own it or have administration privileges.
if the argument is that you as a user should be able to change any icon on any file that you own; i agree. if you want to change the icon for a system binary, you must use windows shortcuts (for windows) or .desktop files on most linux systems to change it locally...
"immutable binary" is still the most important part of what has been said so far.

>>105796254
It did though. fifo-v1 and the hdr protocol were merged a few months after that.

>>105801402
>have administration privileges.
which I have so where's the problem?

>>105801425
the multi user problem.
anyway. we arent talking about the same thing anymore. the argument has changed to something like "the .desktop (windows shortcuts) system works in all cases, but i want something else specific to my case." thats fine but im not here for that. i was only here for "Yes, yes it fucking is"

>>105801455
>the multi user problem.
I'm not gonna rice my icons on the corporate workstation so who cares?

>>105801461
nobody anon... nobody cares.

>>105795598
KDE and gnome both have hardware cursors

>>105795598
People misunderstand this one. Wayland does have hardware cursors but compositors usually only submit cursor updates with atomic commits meaning it is synced up with the video plane. The reason is because the cursor might tear (muh perfect frames). But if your game starts lagging, the cursor will too as if it was in software. Not sure if anyone lets you configure that behavior.

>>105801275
Nvidia GPUs for example still support it: https://github.com/NVIDIA/open-gpu-doc/blob/master/classes%2F3d%2Fclc997.h#L3385

>>105801573
Excellent explanation.
Atomic commits probably have a similar effect to what was going on in linux with the global kernel lock back in the day.
It’s a bit fascinating that a cursor “tear” … thanks for explaining that very well, too…. Is still a problem today.
We used to implement interrupt based software cursors on light-pen applications on 20 MHz 68000 machines at 60 Hz back in the 80s.
I estimate, very roughly, computers are 10,000 times faster nowadays, so that atomic call-gate through to the gpu over the bus must be pretty intensive.

I remember looking at the linux kernel hypervisor code, and some retard did a hit-and-run checking ages ago that exposed this low level ABI as an XML api. I imagine this kind of shit is still going on, we probably have to work harder to find and get rid of those people’s check-ins.

>>105801461
> rice my icons
Haha! Good point.
Remember the days of loading up command.com in debug and changing it to say “Yes, Master? >” and it was like having a real AI machine slave doing your bidding at your back and call?
Good times. I fell in love there, once.
Then they introduced “prompt” and it was way less fun.

works for me, noob

You're probably one of these people still crying that Linux doesn't barely work like it did back in 1998 where you had to spend 99% of your time configuring shit.

>>105801573
Still, they shouldn't need to bitblt the cursor at all.
That’s not what I think of the tem “hardware cursor”
Just upload the cursors, tell them where to move (or just open a shared region of memory and tell the gpu this is where the cursor is) and let the GPU draw it in hardware, more-or-less. That’s how sprites work. No tearing possible.
You got gpus with gigs of video ram for storing pre-uploaded textures and scenes, but they can’t find 1k to store a cursor set.
X windows, btw, works well assuming GPU hardware based cursor sets.

>>105802127
>let the GPU draw it in hardware, more-or-less
That's what happens. The thing is that they submit the update in a way that is synced to the primary plane so the GPU waits instead of drawing it immediately. For what it's worth, I have never seen the mouse cursor tearing on xorg. Maybe it can in certain circumstances and I'm just not autistic enough to notice but that's the "justification" anyway.

>>105802091
We bought slackware on CD.
I just can’t believe things like cursors are a problem in 2025 since it was solved in before 1998.
Sounds like there needs to be a configuration option to just forget about the atomic commit.

I picked up a new galaxy note not too long ago at best buy. I held the thing at angle and used the stylus to see the lag. It was worse than the galaxy note 2… the new one had, like, 1/3” lag behind the stylus.
Things are getting shittier.

>>105800744
Yes, which is the only useful form of X.
And no, games don't benefit from uncomposited X anymore, now that VRR is widespread on monitors and TVs and offers the same latency on Wayland.
And Wayland allows for screen-tearing if you really must have that.

>>105802244
Why would you use a compositor on xorg unless you're into ricer shit? It's pointless otherwise.

>>105802265
Because it looks nicer.
Why wouldn't I?

>>105802345
>nicer
So ricer shit.

>>105802362
If a vanilla KDE desktop is ricing for you... then yeah.

>>105802414
glad we understand

it's so fucking funny that literally the only argument that anyone can come up with is the way the cursor feels when it's a problem with the software they're using on top of wayland. plasma 6 feels great in a wayland session, so clearly they've figured something out that the garbage ass composers used with ricer WMs haven't

>>105802501
This.
I don't know about other DEs, but jumping back between X and Wayland sessions a few months ago on KDE I felt Wayland was more responsive than X.

Can someone explain why it only has 5 protocols marked as stable when it's nearly 20 years old?
https://wayland.app/protocols/

>>105802743
I was running ubuntu and didn’t know I was running wayland until this xlibre thing came up.
I use Xterm, and a bunch of other stuff like that, and i thought xeyes was broken somehow, it can’t track the mouse over things like the gnome settings windows, tracking only works over xterm and other X based apps.

>>105802841
When they’re all stable, it might have near feature parity with X.
> keyboard input grabber … unstable
One day it might approach X like security, too.

>>105802846
>it can’t track the mouse over things like the gnome settings windows
Yes that is expected. Apps should not know about mouse position unless they are the recipients.

>>105802841
And yet, they are enough to replace X for all but the most edge cases.
All without 40 years of code cruft.

>>105802925
what's the difference between waylandism and antisocial personality disorder?

>>105802925
>All without 40 years of code cruft.
Wayland compositors are already full of code cruft and legacy shit.

>>105802127
litterally the only reason they sync both planes is becuase they have some major autism about the cursor moving out of sync with windows.

I will not stop using wayland. Go take your culture war somewhere more receptive like plebbit.

>>105802925
>Apps should not know about mouse position
only those who are deranged as tinkertrannies could decide that software should not control the computer.

this again proves that wayland don't have technical problem, it has culture issue.

use case for software? GNOME already has everything you could possible need. GNOME browser, GNOME text, GNOME music (although it can't open music files). you don't need more.

>>105802501
>>105802743
>>105802846
you guys are actually retarded and cannot even feel 33ms of delay, because you're mentally slow. You clearly have never used Windows, MacOS or X11 without a shitty compositor. It feels WAY different. Especially on 60Hz displays.

16-33ms delay on cursor updates is entirely unacceptable and only coping apologists would say otherwise. Wayland is unfit for use, throw it in the trash.

>>105805066
you've fixated on something that literally isn't a problem because you're retarded. I've always been sensitive to latency, to the point where I imported one of the very first true 120Hz LCD monitors back before you probably even started using Linux. I've used Windows, MacOS, various DEs and WMs on X11 for decades, and I'm telling you Wayland on Plasma is absolutely comfy for desktop use on 240Hz or 60Hz. The problem is what YOU use, be it hardware or compositor.

I'm getting tired of fake voice man

https://www.youtube.com/watch?v=EVd-cgWoiMY

>>105795341 (OP)
You are mentally ill. Stay off /g/.

>>105805187
plasma/kwin is an exception because they do some black magic to minimize input lag, https://blog.vladzahorodnii.com/2020/12/10/compositing-scheduling-in-kwin-past-now-and-future/, they also don't follow the gospel of gayland closely.

>>105795487
No hdr on xorg nor per surface dpi. Enjoy being stuck in the 00's

https://devtalk.blender.org/t/vulkan-wayland-hdr-support/41214

https://blogs.kde.org/2025/06/02/revisiting-x11-vs-wayland-with-multiple-displays/

>>105805750
>Tracey Clark
i'm not reading tranny's flapdoodle

should be titled kwin vs xorg btw

>>105805815
They are the people deciding the future of linux desktop, not you. They have control and power over you, not the inverse.

>>105805734
>No hdr on xorg nor per surface dpi
and whose fault is it? who said that xorg is dead and no new features is allowed?

>>105805827
Kwin is a compositor for both xorg and wayland. It's xorg kwin vs wayland kwin; using mathematical reduction, you remove kwin on each side of the equation, you get xorg vs wayland. QED.

>>105805187
>I've always been sensitive to latency
evidently not

>>105805860
wayland to kwin is like html5 to firefox. one is spec, another is engine that implements the spec.
this is comparison of engines.

>>105806057
It doesn't matter from the user pov.

>>105795341 (OP)
What is the state of things? Haven't used linux for a bit and back then wayland was still in its infancy.
Heard X is going the way of the dodo due to corpo takeover of linux.
Can you even avoid wayland if it's supposed to be dogshit?

>>105806159
Both Wayland and X11 work fine.

>>105795341 (OP)
Disingenuous, retarded gorilla hours

>>105805840
>nobody works on thing
>ITS NOT DEAD ITS NOT DEAD REEEEEEEEEEE
Dumb ass. If it wasn't dead it would support HDR you fucking idiot.

>>105806159
The kde plasma Wayland session is OK, has a few niceties that aren't available on X like HDR and proper multi display VRR, but I still find issues with copy paste from time to time, if you hit a high gpu load the compositor can stutter , firefox's PIP mode doesn't stay on top like it should, games don't always launch on the correct monitor. So I stil end up using the X11 session for the time being, but I'd imagine It won't be long before KDE fixes it themselves, even if the gayland trannies controlling the protocol hold things up for another few years.

wlroots based compositors also function pretty decently,

can't speak for hyprland
haven't tried gnome for a while, and the last time it was not good.

>>105806177
>>105806236
So nothing major to really worry about, can still use X11 until wayland is truly ready.
Things don't sound as grim as some people make it out to be.

>>105805187
>I've always been sensitive to latency
then why can't you feel the atrocious latency of wayland? This isn't an opinion. It was an intentional design decision to make Wayland feel like sloppy shit. Wayland was never designed for desktop usage. It was designed for kiosks and consumer slop.

>>105806221
Nobody was permitted to work on it. Nvidia proposed a HDR extension years ago, but it was ignored because Xorg was purposely killing it to push Wayland.

>>105806159
It's great***************

*not actually

>>105806159
wayland has reached feature parity with x11 for a while, there's still a few niche features that you might want x11 for but those are quickly being implemented, and the downsides in x11 like terrible multi monitor are too big to overlook.
gnome is removing x11 this year, so yeah in a couple of years you won't have a choice anymore.

>>105806290
> feature parity
Its missing fundamental features that all other windowing systems have. These features will never be added for religious reasons.

are hyprland devs waylander also? they also believe that programs that have access to the cursor may cause the collapse of civilization?

>>105805187
Perhaps the HRT is affecting your perception of reality
Many such cases

>>105806290
>can't use different window managers with the same server
>can't draw anything without external libraries
>can't ssh forward
>can't run on other platforms
>feature parity

>>105806255
I'll give the doubters some slack, as the development process of wayland has been horrendous, took 5 years just to get a spec for color management merged when it should have taken a couple weeks, same for things like allowing screen tearing, the new hot garbage take on monitor selection, which could take years to resolve.

the people at the KDE project are the ones who deserve credit, they aren't afraid to work around issues that the gayland people refuse to solve in a timely manner.

>>105806550
https://wiki.hypr.land/Configuring/Using-hyprctl/#info
Hyprland exposes all the information

https://www.youtube.com/watch?v=OV-i3y61RAQ

>>105806774
Looks cool, but not something I can see myself using.
A lot of these rices I wonder if people actually daily drive them.

>>105806784
it might cough break

so none of you use these dwms?

windows-key enter, click 1, 2, 3

god I used it was hard to not
https://youtu.be/OV-i3y61RAQ

ree I still rn went windows key r

the problem with not just hyprland but well lit dwm


so you don't use it for a while. omg where was I?
it doesn't compile either that or it does werk

>>105795515
Yeah who cares about the future not being able to solve a problem that was solved decades ago on every major display system.

>>105806894
wayland not having a hardware cursor is a conscious choice. it's not about them not being able to solve the problem, it's about the fact that having a hardware cursor introduces a lot of issues when it's completely unnecessary with modern hardware.

>>105801573
>But if your game starts lagging, the cursor will too as if it was in software.
I always thought this is because compositor doesn't run with any special priority and so it gets no CPU time when someone wastes all cores and starts swapping memory on top of that.
On Gnome I've never noticed any cursor delay like you see in video games that hide OS cursor and draw their own in-game (those have visible latency even for insensitive dummy with 30 fps displays like me).

>>105802841
>goto wayland-protocols repo
>chech version 1.2 (9 years ago)
>pointer-constraints and relative-pointer have been unchanged in unstable since
>been supported in previous-LTS of all major distros
>recent KiCad wayland rant complains about lack of cursor warping, calling it optional unstable feature
it's so silly

use case for smooth cursor movement without stutters?

let's go
I'm not even polish but yutube seems to know me
https://www.youtube.com/watch?v=Ckom3gf57Yw

Do you have an NVIDIA card, OP?

There's a distinct laggy feel when using Wayland + NVIDIA.

>>105807214
X is laggy with my Nvidia card. Wayland is smooth as fuck

>>105807230
lol yes right I hate you nvidia driver pajets so fucking much

noo it sucks in wayland and it sucks in x, don't mind you paid a billion rupees for it

>>105807254
Your fault for buying hardware from companies that are hostile to FOSS.

>>105807254
Nvidia is basically perfect on Wayland now. X is still a laggy mess and has been like that for the last 10 years

>>105807293
Lies

>>105807317
Try it if you don't believe me

>>105807355
Unless something has changed in the past 20 or so days, you're lying.

>>105807399
Lies

>>105807404
It doesn't work on my Turing laptop, nor my Ampere workstation.
Way too many odd bugs and artefacts keep popping up that otherwise don't exist on my Radeon desktop.
I've tried Sway, Hyprland, Plasma, and even Gnome over the past month, yet all of them shat the bed from day one.

>>105807434
>many odd bugs and artefacts
Examples?

>>105807458
My monitor numbers being jumbled up after waking from sleep, only half of a window scrolling, random elements flickering, applet context menus not opening (or opening far away from their expected location - sometimes even on another monitor), windows not rescaling when moved to a different monitor with a differing scaling factor, etc.

>>105807293
sure if you use a compositor

>>105801573
Finally some reasonable answer in this retarded thread.

>>105807260
is alway's foss' fault isn't it? not share incompetence

>>105807260
anon is this foss in the room with you rn?
does it go read ze manual?

>>105805888
>>105806256
>>105806552
>>105806894
try plasma 6 on wayland instead of malding. it just werks

>>105807743
side question: what's a good distro to just "try plasma 6"? Is that KDE's distro fine?
else Kubuntu, Debian testing, Fedora or something else?
I have no idea who still ships plasma 5 and who already ships plasma 6, and who ships it modified vs upstream etc.

>>105807230
>>105807214
>>105807293
>>105807743
Benchmarks say X is faster, while Wayland introduces multiple frames of input lag.
https://gitgud.io/CrunkLord420/crunkbench
Waycucks will never run this simple benchmark because it proves them wrong.

>>105807800
Basically you don't want a debian based distro if you want kde

>>105807966
>Wayland likes to sync the numlock key state constantly, but that's fine since we just need it to blink once.
wtf

>>105807966
this kills the waytranny

>>105807966
>-8.33 ms latency
x11 can do time travel confirmed

>>105809634
the results contain some negative numbers, so probalbly yes

>>105809634
It means X11 can beat the speed of USB and however long it takes for the keyboard firmware to activate the LED. LED activation is not instant. The actual full trip latency is longer than the benchmark tests for.

>>105807966
Humans can't notice a delay of 30 ms. There's no usecase of making it faster

>>105807966
thank you CrunkLord420, any numbers for KWin?

>>105809604
hyprland problem, cope

>>105810337
Nope, this is universal. Wayland (as a protocol) is architecturally flawed.

Benchmark it yourself, you won't, because you know the truth already if you're not physically spastic.

>>105807966
Was that i3 without a compositor?

>>105810450
No compositor as that would re-introduce the exact same issue. Note that this is a Linux issue. WDDM and Quartz have no issue with cursor latency while being compositors. This was a design decision in wayland to lock the cursor to the framebuffer.

>>105810439
prove it's universal by benchmarking plasma 6, I dare you
hint: it's not

>>105810517
can you turn off the compositor for Plasma? I've tried Plasma before, both X11 and Wayland. I didn't benchmark them. But it's immediately obvious that Plasma X11 feels like soupy shit compared to X11/WDDM/Quartz.

You've come up with a flawed argument knowing the same issue affects Plasma 6 but does not affect other X11 window managers.

It's a flawed design, not fit for use. It feels like shit, it's immediately obvious. You really do not need to actually benchmark it. It's just club to beat obtuse faggots like you with.

>>105810536
>can you turn off the compositor for Plasma?
Yes

>>105810553
How? Seriously.

>>105810559
Well, on X11 you can turn it off. At least that was the case when I last used it a few years ago

>>105810536
until you come up with kwin benchmarks, the club with which you claim to be beating obtuse faggots is imaginary. in short: cope more

>>105810579
>>105810585
Benchmark it yourself if you love Plasma so much.

>>105810241
can you tell the difference betweeen 30fps(32ms frames) and 60fps(16ms)?

if you cant you probably have boomer eyes, as most people can tellupt to 120(8ms)-240fps(4ms)

>>105806940
its litterally just autismic fits about the cursor not being in sync with window movement, thats it., thats the only issue it really causes, ands its only really a problem, if your compositor is a slow pos

>>105810618
the onus is on you to prove it sucks. for me, it just werks

>>105811236
You don't need to test anything. If you can't immediately feel it you're objectively and measurably sub-human.

>>105807966
Thats why they’re trying to kill X.
So they won’t have to break things just to compete with benchmarks.
Remember when Intel removed page permission checking to speed up their processors to compete with AMD? That didn’t work out well.

>>105806221
10 out of 10 wayland cultists can “see” the HDR support on Wayland vs. X even when they’re both running X without telling them.

>>105809785
Can confirm.
Keyboards are usb 1.1.
I just hooked up a USB ethernet adapter on a USB 1.1 4-port belkin hub yesterday.
The speed brings me back to the dialup days.
Makes sense again to enable hash mark printing in ftp.
This BBS is terrible.

>>105807029
KiCad and SDL hate wayland.

>>105811567
>SDL hates wayland
Weird statement given that it is often the only real world example of how to do some stuff in wayland.
I wouldn't know how to work around hidden window inhibiting without it. Or how to get system cursor theme for wayland-cursor, and various other little things.

>>105805066
Ok, so I am primarily a mac user with Xorg.
I don’t know if wayland exists at all for mac.
But, i wouldn’t consider ther performance of something like Inkscape used with a tablet that great.
The biggest disappointments are programs that use gtk widgets on Xorg. Those are absolutely abysmal, performance-wise. Unusable.
I’m thinking of things like OpenOffice.
Even individual gtk buttons feel the need to draw gradient shaded texturized buttons using rasterization.
Now run openoffice remotely and you can Watch It Draw line by line and worm it’s way down the screen.
Probably *most* of the problem is gtk.
And, I think, that’s a gnome product isn’t it?
Anway, it’s not there now, will never be, and is probably going in the wrong direction.

>>105811635
GTK is genuinely awful for it's own reasons. There's a reason people have consisting moving to Qt for over a decade despite it not being under the most free license.

>>105811613
The SDL is busting their ass to basically implement shit wayland should have done.
Wayland owes it’s existence to SDL.
So thank the SDL guys for implementing scaling, for example.
But, what they should have done, for the greater good, is let wayland drown in it’s own pool of vomit instead of cleaning it up themselves.

>>105811635
OpenOffice isn't GTK. It uses VCL or whatever it's called

I feel that the Wayland devs know it's garbage and that's why they're suddenly trying to make it mandatory in everything, because they think forcing it on everyone will lead to people coming in to fix it for them out of frustration.

>>105800642
1. it's gay
2. it's trans
3. it's still widely incompatible after several years of development
4. it's a very unnecessary rewrite of an already tested protocol
5. it's gay

>>105800642
>because the cursor is vsynced
yes, other reasons too, but this one is the top of the list right now.

>>105811666
Yes, shit Wayland should be doing. I've run into a few of those. Color profiles, keymapping... Wayland seems to push a lot things that are common to any WM onto each WM to implement.

Wayland shills will NEVER benchmark their shit because they KNOW it's laggy shit.

>>105807800
>>105808068
>>105808068
Kde neon, the official kde distribution

>>105813382
Referring to wayland like that is like mixing up all web browser together as "html5". Every engine is different software.
Can we stop doing this?

>does html5 support vertical tabs?
>ublock origin stopped working in html5

>>105813892
they can't stop because it's the only thing they have

>>105813941
This is just misinformation that shifts the responsibility for lack of features from DE devs to some "wayland protocols" repo where a bunch of fags (like 5-10 people) are discussing some bullshit for decades.

This is just wrong and a part of problem with current state of linux desktop. A pressure should be put on actual devs.

if programs dont work on gnome it's gnome fault
if programs dont work on KDE it's KDE fault
etc

if GNOME, KDE, hyprland (whatever) troons can't make functional desktop then they should stop development and ACK themselves. it can't be simpler than that.

im tired of this "wayland" boogeyman.

>>105806290
Wayland still has zero compatibility with Xfce, MATE, LXQt, Unity or any other usable DE. I'm forced to use a window manager like an autist and it sucks.

>No X11 forwarding over SSH

It's trash.

>>105814449
Lol, I discovered this an hour ago.
I couldn’t get it to work at all, and even by using xhost+ and setting display to the xwayland server IP at 0.0 didn’t work.
Do you have to turn on remote wayland / Xwayland somewhere?

>>105814643
I don't know, I don't use wayland trash. X just works. I'm looking forward to Xlibre namespaces, it's about time.

>>105814405
Both Xfce and Lxqt work with way land

>>105805066
>16-33ms delay on cursor updates
Latency has been measured between KDE on X VS Wayland and the results are exactly the same.

>>105804515
Wrong.
I and I alone should control the computer.
Wayland allows me to have a more fine grained control over my computer compared to X.

>>105803026
Now imagine what X looks like under the hood.
That's why it's slow and bloated as fuck.

>>105814643
You can try waypipe, but almost no apps worked with it when I tried. I used to remotely use X apps all the time, but I guess wayland devs don't do that kind of thing.

>>105816873
it's a very niche feature to put it mildly.

>>105816810
>I and I alone should control the computer
Wayland schizos believe that computer shouldn't be programmable

Programs control the computer. By taking away the control you reduce the programmability. Computer becomes less useful. People buy computers because they want them to be useful. Wayland loony ideology violates fundamental reason for computer to exist.

https://www.kicad.org/blog/2025/06/KiCad-and-Wayland-Support/

>>105816944

Almost like they made it to tick a box rather to get something good. Everything that uses portals with it break outright here.

>>105817249
Wasn't there similar problems with PCSX2?

>>105817249
That KiCad blogpost was a mixed bag. Some of it is valid criticism, some of it invalid, and plenty of it lacks details or motivation for why they need it.

>Window placement and restoration: Wayland does not currently allow controlling window position. This means that when you open KiCad, it can not remember where you last placed your windows.
>Can't restore docked windows
Wayland developers will tell you that the DE/compositor/shell should handle this and it should not be a part of protocol. While I agree, sadly the DE provides poor control over this, afaik packages can't install some rules, and it's probably very fragmented..
>Multi-window coordination: Problems with managing multiple KiCad windows simultaneously
details needed.do they want relative position between windows or something?
>Window dragging limitations: Dragging tabs and panels between areas is broken or unreliable
Definitely compositor bug. Gnome (or at least what Ubuntu shipped) used to be super bugy in some versions. E.g. I could not drop to right half of monitor. Not some protocol limitation.
>Cursor/pointer warping: ... it is conditionally available, depending on support for optional protocol extensions
relative-pointer and pointer-constrains combo allow plenty of pointer control, including warping in same window. they are unstable, but have been unchanged for 9 years and virtually everyone implements them
>Focus management: Unpredictable window focus behavior that can interrupt workflows
afaik only focus management is xdg-activation which allows you to spawn window with focus. what more do you want/need?
>Input device handling: Various issues with specialized input devices and hotkeys
need more details. but probably again left to compositor settings instead of protocol, yet DEs don't provide them for users

(cont.)

>>105817471

>OpenGL call throttling
easily solvable, but poorly documented. thanks SDL >>105811613 mix of two workarounds
The workaround is to register frame-done callback on surface and just don't swap until it signals ready.
And second it to implement a wl_display_dispatch_with_timeout() that actually reads from socket (unlike wl_display_dispatch_pending()) but does not block. This way you don't stop receiving messages when you don't call eglSwapBuffers().
Also solves managing multiple windows on same thread.
>High CPU/GPU usage
>Graphical glitches
my bet is on buggy Gnome in early versions
>Clipboard functionality
Hehe this one is spicy. My main suspect is the poorly solved inhibiting, because when process is stuck in eglSwapBuffers() it can't process clipboard requests.
On there is also a deadlock when you don't use async event loop because when you copy-paste from app to same app because you are expected to write and read on same pipe at the same time.
But it's all solvable. Again thanks SDL for reference implementation. Love you.
>Modal dialog behavior
>External tool integration
Legit wonder if Wayland devs think this is useless usecase - opening dialog window as centered to main window. On Gnome they just appear at random positions unrelated to origin, sometimes even on different workspace or monitor whereever you have focus right now.

>>105795341 (OP)
My Laptop has a 4200M and is using that iGPU. I've felt like I needed a new laptop after an update because everything felt slow, sluggish and not good to use, imagine an XP era Eee netbook running Vista, that slow, so bad even my brother complained that I need a new device.
After a while while fucking around with the login-screen I found out that for some reason that update changed the selected DE to use gayland instead of X11. I changed back to X11 and that 12 fucking years old laptop is flying again as if was just 2 or 3 years old. I do not know why, but gayland literally leeches compute and makes old hardware impossible to use. It is not a drop-in replacement and it will, same as Windows, cause a lot of e-Waste.

gayland needs to be stopped, people need to come back to their senses and just clean up X11.

>>105814278
>a bunch of fags (like 5-10 people) are discussing some bullshit for decades
really makes you wonder what's taking them so long

>>105818500
They would be killing their golden goose if they did their job.

>>105795341 (OP)
>it just feels different bro, it gives me the wrong vibes and stuff. What's the color of your energy, bro?
These discussions are lovely.

>>105817671
This is my experience too. Pretty sure it was corpo idea to make Linux undesirable for people moving from older Windows systems.

>>105816873
> waypipe
Thanks. I’ll play around with. We use remote X sessions all the time.
We have these massive, beefy servers with 1TB of ram, and running dozens of VMs with Windows going back to XP to test backward compatibility.
We typically remote-X into several of them, do whatever we need, and log out.
Some of them have snapshots going back a decade. The latest one is using pop!os just for variety I guess.

Plan 9 was supposed to be the future, I think remote display and display servers was “baked into” as well as compute servers.

Wayland seems to be a step backward by a bunch of guys that lived their lives using their parents home PC playing games on a single television, and can’t see out of that experiential box.

I think I’m starting to see how that “use-case?” meme started.

>>105817513
Sounds like SDL should be running the show due to technical competence.

>>105818778
Plan 9 did a lot of things right. It's really too bad it didn't take off. Remote CPU and disk. I always clocked the wayland people as humorless laptop users.

>>105818928
Well they took some time to get things right. Remember that screenshot when someone complains about games not spawning on primary monitor? That was early Wayland support in SDL that just took first listed monitor as primary despite no flag. There are still pretty significant fixes from what previous-LTS Ubuntu ships and git head. But afaik nothing in some time and wl support in SDL3 is stable.

>>105818633
I came to this conclusion too. It’s the only one that makes sense when you look at it critically from a big picture point of view.

>>105816792
Show your benchmarks, make sure you disable to your compositor or your results are invalid.

>>105818585
>33ms input latency
>vibes

>>105813904
it's not my fault that all wayland compositors have the same problems because wayland is architecturally flawed.

Honestly a reference implementation would not have solved anything, or changed this discussion in any way.

>>105813904
>>105813892
That is the most retarded, forced meme ever. No one uses the implementations when they're singing gaylands praises, so why start pointing fingers when people point out it's flaws?

Most of you faggots use a compositor on X and don't notice the added delay anyway

>>105819489
>disable to your compositor
You guys need to drill something onto your skulls: no one uses desktops with tearing.
If you REALLY must have that, then keep using X. No one cares.

>>105819863
>no one uses desktops with tearing
I do

Windows and MacOS doesn't have this weirdo problem where their inputs feel like shit when using a compositor. It's a Linux thing. Fix it, or I'm just staying with uncomposited i3wm like an adult who can actually feel the shitty input latency.

>>105816810
The code bloat meme is dumb and stupid. Xorg is not anymore complex than a typical desktop wayland compositor. Both have the same performance because they both use DRI3.
>but it still has code for legacy garbage X no one uses
yeah and wayland compositors all support legacy garbage too. what a shocker, turns out they didn't design correctly the first time and it's been over a decade of implementing shit that was already in X.

>>105819863
you can just use tearfree bro, you don't need the composite extension

Nvidia heard our call, waylanders. they started a new driver written in Rust.

>>105820099
>Nvidia heard our call, waylanders
Nuh uh. The biden government literally put every american corporations balls into a vice and said "rewrite in rust, or ELSE!"

The biggest lobbying action in the last millennium and no one is talking about it.

All I want is to be able to use OBS hotkeys so I can use push to talk. Even if the window is active, this shit never works and I'm forced to use voice activation. I know you can kind of make it work with a some hacky workaround with websockets, but I don't want to do that. Also X11 fucks with my screen for some reason so I can't use that.

>>105820751
>push to talk
usecase for keyloggers?

>>105800835
>Admin kept turning on the glpipes screen saver on the VSS server.
sovl

>>105819662
no one cares tr00n, xlibre will pwn your faggot ass in the future anyways.

>>105819863
Xlibre doesn't tear without compositor

I've been using KDE Plasma 6 Wayland on AMD hardware and I've long accepted that
i) I will never get cursor latency better than on Windows, be it in Desktop or 3D apps, even if I force tearing through window rules.
ii) I will never get proper HDR support and color-accurate work will always be fucked, partly due to apps not adding support for either.
iii) I will never get working screen+audio sharing outside of OBS.
iv) I will always use KDE's System Settings GUI app because configs files are undocumented and there are no feature complete CLI tools.
v) I will never get something like Coolbits on NVIDIA GPUs and X so I can fuck every other standard to hammer in my own overclocked EDID for cheap ass monitors and tell them to shut the fuck up and display an image.

As for the latency, Just pretend the latency doesn't exist. Yes, you can live with it. If your monitor doesn't support anything but 60 hz out of the box, you can live with that too. Just don't pay it much attention and you'll be fine. And honestly color-accurate work, HDR and all that stuff is just unnecessary bloat. Your computer can work without all that crap and display an image. For free. And hey, at least I can assemble a jigsaw in Palapeli or play Spider in my KPatience app.

>>105800835
if you are running anything with the vaguest sense of a modern graphics card using dri that wont happen, guessing you were using some ancient chipset that provides basic framebuffer acceleration and nothing else.

>>105820920
what are you talking about? kwin_wayland has 1 frame of lag max, its the only wayland compositor that doesnt have significant input lag.

>>105807743
>plasma
Anon..

>>105820948
Now boot up into Windows, open up the same app and enjoy truly latency-free experience.
>input lag
This is almost never a problem with Wayland, actually. It should be called output lag.

Also, this is very important but I kinda forgot about it.
vi) I will always get extra latency if I enable ICC or god forbid Night Light mode.

>>105810559
Easy, turn off that checkbox or even better, use the other one to turn it off automatically when full-screen.

>>105821008
time for inputs to update the screen is input lag, and once again if the compositor isnt causing the lag, then your configuration is fucked up some how, guessing you havent setup your dxvk config.

>>105821079
>time for inputs to update the screen is input lag
This should be called latency. Input lag is the time it takes for the inputs to get processed by the application but NOT displayed, nowadays every other retard thinks latency = input lag presumably due to tech incompetence of news outlets since like 2010s.
>dxvk config
What is there to setup with dxvk, anon? Either way I can see additional latency in literally any app, even if it's built natively for Wayland (via SDL or shit). XWayland apps also suffer. Granted overall the latency is bearable.

>>105821262
for dx8-11 games a translation layer called dxvk is used, valve's proton uses this , the source project can be found here https://github.com/doitsujin/dxvk
the dxvk.conf allows you to control some of the behaviour , and should be put in you /home/<user>/.config folder , with it you can override the renderahead liimit , many games default to 3 frames, but you can force it to 1.

additionally you can change the vulkan present mode, by default I believe it is FIFO, which is triple buffering essentially,
The environment variable MESA_VK_WSI_PRESENT_MODE and can be set to fifo, relaxed, mailbox or immediate. immediate is the lowest latency, but is not vsynced, mailbox would be the next best, then relaxed, then fifo. if you are using a vrr capable monitor , then this environment variable is less useful.

>>105821399
>with it you can override the renderahead liimit , many games default to 3 frames, but you can force it to 1.
I'll look into it but I've long since lost hope. Furthermore, as I said, there's additional latency in native Vulkan/OpenGL apps which don't use dxvk.
>vulkan present mode
Yeah anon I know all about it already. I've already tried every shit imaginable, I just can't get display latency better than on Windows. You can also show which mode the application is currently using with MangoHUD and also change it there.

>>105821484
You must have some kind of issue, because I've literally measured it before, and its lower on linux vs windows on average

>>105820015
>feel like shit when using a compositor. It's a Linux thing
Clearly you have some problem with your graphics stack or DE.
Are you perhaps using Nvidia?
Windows and MacOS do not have snappier desktops than, for example, KDE Plasma.

>>105820089
>>105820901
>no tearing
Then you end up with the exact same latency as Wayland. Go ahead and measure it. It's easy with any relatively modern phone camera in 480 fps mode.

>>105819863
me too. unironically even.

>>105821671
>and its lower on linux vs windows on average
I'd bend if you manage to run osu!lazer https://github.com/ppy/osu/releases/latest/download/osu.AppImage with less latency on Linux than on Windows.
By default it runs on SDL2 over X(Wayland), but you can force SDL3 with OSU_SDL3=1 and Wayland with SDL_VIDEODRIVER=wayland
I didn't exactly measure it but from what I can feel it's
Windows < Plasma X11 < standalone gamescope session < Plasma Wayland

vii) I will always get fucked up mouse sensitivity on Linux due to fractional scaling sensitivity multiplier.

>>105821772
>Windows and MacOS do not have snappier desktops than, for example, KDE Plasma.
lmao, Waytrannies believe this shit.

>>105818633 Checked
>>105819329
Nothing else makes sense. No one who wants to improve something would have done as gayland devs have done.

I genuinely have to believe anybody who doesn't feel the input latency in Wayland is just underage, otherwise it makes no sense.

>>105821968
question, are you using an nvidia card? I only ask because I have an amd card, nvidia's proprietary driver may be doing something different than the opensource driver.

>>105822274
you're just slow.

>>105822291
it'd be the opposite, genius

>>105821984
You've been living under a rock, buddy.

>>105822321
Why do you weirdos double down on shit people can test and disprove in like 5 minutes?

>>105822290
I'm currently using a laptop Ryzen 9 365 APU with Radeon 880M and RADV driver. However the experience on a 5700G with RTX 3090 desktop about a year ago was quite similar (Windows < X11, I didn't test Wayland at that time)

>>105822382
hmmmmmmmmmmmmmmmmmmmmm
https://github.com/ValveSoftware/Proton/issues/408
seems like valve isn't using raw inputs, kek,

maybe try using https://github.com/GloriousEggroll/proton-ge-custom, they have a patch to enable it

>>105822435
I've been using kron4ek wine-tkg builds for windows applications, I think tkg also has patches for raw inputs.

>>105822435
i tried to play quake on proton but the lag input was atrocious. i looked up the code and discovered that wine gets the mouse input by a message from the wine server, wine server gets the input by a message from xorg. the three message loops (xorg, wine server, wine client) may have a artificial delay to not eat all the cpu power. instead, proton could just read the input from /dev/mouse0 but no, all this instead.

>>105822522
>kron4ek wine-tkg
doesn't look like they do, looks like they only apply winestaging patches and/or the base proton patches., worth giving GE a try at least

>>105822274
it makes a lot of sense when you realize they use window animations, fade outs, and whatnot.

>>105822664
nvm wine staging has them, i would still give it a try since they have some other stuff .

>>105796911
what jankiness?

>>105822671
Windows and macOS do as well, yet you don't see anyone complaining about input latency on those.

>>105822364
if it's that easy then why don't you test plasma 6 on x11 vs wayland and prove wayland is laggy

>>105823013
Not that guy but I was curious about this issue , and while the plasma wayland desktop is not laggy itself, I tried playing Doom eternal (notoriously bad input lag), and bro there is something fucked up about xwayland because with vsync enabled Legit felt like I was dragging the cursor through molasses, and with vsync disabled it was still worse than tearfree under X. I could've sworn it wasn't like that 3 years ago when I last tried using a wayland compositor, clearly something is broken.

>>105823013
I'm not pulling out a slow motion camera for you if that's what you want. Numbers were already shown in this very thread >>105807966 >>105809604
>b-but that's Hyprl-
Cope.

>>105823087
>suggesting hyprland and nvidia are representative
cope hard, lil bruh, cope hard

>>105823076
are you using proton-ge for raw mouse input?

>>105823152
>n-no u!!
Called it.

>>105823183
just because you call it doesn't mean it's immune from criticism. you huffed pre-cope and you're coming back for more. try providing plasma 6 benchmarks if you want anyone to care

>>105821772
>exact same latency as Wayland.
For the mouse? No. Xorg uses the legacy API for the mouse. It doesn't matter what the primary plane is doing.

>>105823165
yes, i dunno what they are doing, but something isn't working correctly, I'm gonna try a different compositor and see if its the same.

>>105823013
Because KDE is keeping plasma x11 in maintenance mode getting ready to be removed for plasma 7.0? They hate x11 so much i wouldn't be surprised if they sabotaged it somehow to force users to switch, i mean after all they forced SDDM to Wayland and lied about showstoppers. Fuck them to be honest.

>>105823408
they don't hate x11, it's just that their wayland support is finally good enough. and it is good enough. in fact, it's great. and sddm still supports x11 so what the fuck are you talking about?

>>105823408
I don't think they have said they will remove it, just that its not a focus anymore, kde still has liquidshell as an option for instance, since kwin_x11 and kwin_wayland are now split , There basically is no issue with just leaving it mostly as is, so long as the plasma desktop still ships with the X11 backend enabled for qt, it will continue to work.

>>105823465
Way to ignore all my points with cultist babble. Google it if you dont understand.

>>105823560
No they will definitely remove it. They already announced that. Qt will also drop support for x11 around that time.

>>105823653
>Qt will also drop support for x11 around that time
yeah, I'm guessing you don't have a source for that.

>>105823653
>Way to ignore all my points with cultist babble. Google it if you dont understand.
you didn't provide any points

>>105823802
x11 is redundant, you don't keep redundancies around because they add maintenance cost.
once they think wayland is good enough and they're done removing all the glue code x11 is getting the snip.

Where is my:
Window placement and restoration
Docked panel positioning
Multi-window coordination
Window dragging
Cursor/pointer warping
Focus management
Input device handling

you smelly nerds.

I at hyprand protocols and it seems that vaxry troon only bothered to implement shortcuts. Where is window/cursor position etc, where is all that stuff that software developers wants, huh?

>>105823894
>Window placement and restoration
https://wayland.app/protocols/xdg-shell#xdg_surface:request:set_window_geometry
>Docked panel positioning
https://wayland.app/protocols/wlr-layer-shell-unstable-v1
>Multi-window coordination
XDG shell as above, but maybe also https://wayland.app/protocols/xdg-foreign-unstable-v2
>Window dragging
Ditto, usually implemented by compositor though
>Cursor/pointer warping
https://wayland.app/protocols/pointer-warp-v1
>Focus management
https://wayland.app/protocols/pointer-constraints-unstable-v1
>Input device handling
https://wayland.app/protocols/text-input-unstable-v3
https://wayland.app/protocols/input-method-unstable-v1
https://wayland.app/protocols/tablet-v2

>use linux, have problems
>install windows, never have problems..

>>105824250
>install windows, never have problems..
if only. malware, terrible ui, bloat, ads, one fucking drive...
people who use linux have used windows in the past mate, we know what it's like, you're not gonna convert anyone here.

>>105823653
>No they will definitely remove it. They already announced that. Qt will also drop support for x11 around that time.
there has been no announcement for qt7, you are literally making shit up, qt is a toolkit, there is no reason for them to remove it since X is still default for many diistros and will continue to be, just because gnome is retarded doesn't mean everyone else will do the same thing.

>>105824284
>qt is a toolkit
gtk is a toolkit, and they're removing x11 from gtk5 just fine.
if you want to keep using x11 you're free to stick to an old version for the handful of years qt6 and gtk4 will be supported and start preparing to move.

>>105824197
what a weird thing to do.

>>105824272
enjoy the OS equiv of skyrim modding for 30 mins of gameplay

>>105824294
gtk only cares about supporting gnome, qt doesn't

>>105824357
yeah no.
gtk is also used in windows and there's a solution in the works for android, plus plenty of large applications like firefox that use gtk to this day.
qt will drop x11 support, maybe not in qt7 but most definitely by qt8, for the same reasons that gtk will, it's a huge burden to carry around unnecessary dependencies.

>>105824374
>gtk is also used in windows
GTK on Windows sucks.

Qt actually works well on many different platforms.
And removing X11 support would also mean dropping support for several platforms like BSD

>>105824393
>BSD
also in the process of moving to wayland
the deprecation of x11 is a decading in the making, and it's only going to ramp up as more applications and distros drop support.

>>105807966
>https://gitgud.io/CrunkLord420/crunkbench
>i3wm and hyprland
>nvidia
Can someone make these using stuff people actually give a shit about?
Like AMD and a real DE like Plasma (Kwin)?

>>105824412
>also in the process of moving to wayland
only FreeBSD is

>>105824393
Gtk sucks in general.
It’s good they’re dropping X11 support and I can go back to enjoying faster, smaller binaries on the Athena widgets.
… it’s funny back in the Sun pizza box days, we thought Athena was bloatware. How times change.

the shift is coming. (2 weeks)

>>105824422
Everyone copies FreeBSD anyways.
Once they finish moving everyone will follow.

>>105824427
>Gtk sucks in general.
and yet it's the most common toolkit on linux
>>105824422
remind me why we - or qt should care about that?

>>105824435
>why should a cross-platform toolkit care about being cross-platform

>>105824442
cross platform has never meant being available on every possible device
they go where the people and the money are, BSD is not that.

>>105824431
> coming to BSD
Yaay! Wayland is in worse shape on BSD than it is on Linux.
I can’t wait to experience the crashes, the apps I can’t upgrade that stop working. The monitor juggling, the lag and latency, and the 60% drop in FPS, etc. mainly because I have nothing to do, and all day to do it in, so it will be like restoring an old jalopy with the amount of tie I'll have to spend tinkering with it, reporting bugs, getting banned from their wooly mammoth server.
It’s like heaven.
On and something about HDR and possibly security… I can’t wait for those.

>>105824449
Why not drop Wayland support while we're at it too?
Barely anyone anyone uses Linux and it's a way more work than keeping the old X11 support around.

>>105824374
Qt is very modular and can support different backend with minimal extra work.

It has official support for obscure things like Haiku and bsdfb. All these are used way less than X11 on BSD.
There's no reason for them to remove X11 support, a display server that runs anywhere (including OS X and Windows).

>>105824575
>can support different backend with minimal extra work.
doubt. time will tell.

>>105824589
They have been doing it for decades.

The Linux kernel does the same, listen to Linus and Alan Cox:
https://www.youtube.com/watch?v=Nn-SGblUhi4

>>105824589
you're litterally retarded,
qt officially supports RHEL 8 and 9 and SLES 15, of which will be supported until 2031, 2034, 2037 respectively, not to mention a ton of embedded targets , so you can expect the X11 backend to be around for a least a couple more versions,

qt8 or 9 is the earliest I would expect it to be dropped.

>>105824642
That retard asking that speech-length question is an insufferable prick.
I didn’t watch until the end, but I assume that Linus, without saying a word, responded by drop-kicking that weaselly little guy right into the wall while the audience capped and cheered with unbridled joy.

>>105824491
> security?
Let’s put it thus way, for the first 10 years of wayland’s life, it was subject to DOSing CPU infinite loops if anything sent it a message bigger than 4096 bytes causing a (random, hard-coded) core protocol change last year.
This is despite the fact it handles dynamic buffer resizing. Well, maybe no longer… code is there though. Dead code.
This is in the fundamental message loop.
They’re literally nowhere.

>>105824756
Solution 2: kill remote/network support.

touch pad mouse scrolling in google chromium is absolute pain in gay land

it's either too fast or too slow

>>105823087
>I'm not pulling out a slow motion camera
I'm pretty sure you have a smartphone with at least 240 fps slow-motion mode. Get out of your ass and do it.

>b-but that's Hyprl-
I cannot guarantee to you that every single compositor under the sun has implemented Wayland correctly.
I care about the one I use, KDE, and I can tell you that at least on my AMD system on Arch, Wayland has at least, if not better, latency than X.

There is absolutely no limitation on the Wayland protocol that stops it from having the same latency as X. Any case of higher latency is the fault of the compositor or graphics drivers.

>>105825839
You're objectively and measurably subhuman who cannot detect 33ms latency with your hands and eyes.

You will never benchmark because your know the results already.

how to make touchpad in gayland scroll like in xorg?