Wayland's Wild 2024 With Better KDE Plasma Support, NVIDIA Maturity & More Desktops

[MrCooper]

The screen sharing that was implemented in early 2024 was greatly inferior and prone to disconnect and lock up the display.

"lock up the display" sounds like a GPU driver issue.

Screen recording with sound was a complete disaster, including in the programs such as OBS and the dedicated wayland screen recorders which were proclaimed to work.

You didn't describe what exactly the issue was, it's hard to imagine how the screen grabbing mechanism could directly affect sound recording though.

Global hotkey setting by programs such as VLC were not possible, and were excluded as a possibility by fundamental wayland design decisions. I don't know if that has been implemented somehow or will ever be implemented.

The global shortcuts portal is the solution for this, apps / toolkits have to actively make use of it though. Also, GNOME 47 doesn't support it yet, there's ongoing work which might land for 48 though.

I finally switched to Wayland a month ago with Ubuntu 24.10 and GNome. The input lag drives me crazy. I cannot figure out how I can fix this.

Input lag doing what exactly? And is this with the nvidia driver or another one?

With so many basic things such as colour management not fully implemented, I wouldn't call the Wayland KDE session at parity to the X11 one.

Indeed, e.g. the X11 one can't support HDR or displaying multiple apps using different colour states for their windows correctly at the same time. Not sure what you have in mind for colour management that isn't covered by the corresponding Wayland protocol already supported by kwin.

I am not asking why the feature should exist, I am asking why Wayland should take care of that, since Wayland has no concept of windows. The compositor handles windows.

To be pedantic, Wayland does have the concept of windows, it doesn't have a global coordinate system though.

[Weasel]

First, why would you ever want that? Window placement is the responsibility of the compositior not the display protocol (dp task is just sending window buffers to the right place).

Wayland is the compositor's protocol. There's no separation anymore.

And why would I want that? Listen dude, I don't give a fuck what your goals are. It's MY GOD DAMN COMPUTERS and I want to position my windows using MY script on all of them.

You are asking the police to respond to medical emergencies, maybe the sheriff in your town did work overtime as an ambulance driver but that is not how it works anywhere else.

Funny because you call the same number for emergencies and they route it themselves where appropriate. See how that works? The number is the "protocol" (for human beings). Scripts need one sane protocol, I don't care what the compositor does afterwards and where it routes it and who's in charge of positioning my fucking windows, as long as I get the result: they get positioned where I requested them to.

Period.

Second, wayland is not all of X, by design. You dont like it, start maintaining X. The rest of us are walking away from that garbage for a reason.

Ok but who asked?

Legit who asked?

I couldn't care less about your opinion. I'm not the one who called Wayland on "parity" with X11 when it's clearly not. That's not an opinion, that's a misinformation and wrong fact. Fucking deal with it. Mine is fact, yours is opinion, we are not equal. My argument is superior.

[finalzone]

With so many basic things such as colour management not fully implemented, I wouldn't call the Wayland KDE session at parity to the X11 one. Maybe it's suitable for the basic user/gamer, however, there are so many workloads that make it unusable.

Even Xserver has issues with colour management. The Wayland protocol aimed to reduce the overhead and learned from the mistake coming from the four decades messy X11 protocol. It takes time and patience to implement the incoming colour management given the variety of desktop environments. Logically, Wayland session has nothing to do about parity to X11 as majority of functions is deferred to the compositors i.e. KWin, Mutter and Weston. Let remind Wayland implementation is already in use worldwide notably Smart TV and appliances.

[uscracks94]

Funny that it's always other things at fault whenever something doesn't work with wayland.

[wertigon]

[QUOTE=Weasel;n1516089]Wayland is the compositor's protocol. There's no separation anymore.

And why would I want that? Listen dude, I don't give a fuck what your goals are. It's MY GOD DAMN COMPUTERS and I want to position my windows using MY script on all of them.

Nerd is upset a deep change in his favorite OS broke something, so now they have to learn a new way of doing the same thing. News at eleven.

Ever considered there might be a better way to do it? Like tiling, for instance? I've heard good things about sway and how they handle tiling. Although to be perfectly blunt, if I could get Gnome or KDE to support tiling the way sway does it, I would be perfectly content. Gnome, especially, already does a damn good job of getting out of my way so I can do work. Proper tiling support and one workspace per screen, and it's pretty much perfect for me.

Also, your request is an invasion of privacy at the very least, and would allow a script to push a window 100% off edge, essentially hiding it.

I'm not the one who called Wayland on "parity" with X11 when it's clearly not.

Funny that it's always other things at fault whenever something doesn't work with wayland.

This is because the scope of wayland is drastically different from X11. Things are meant to be done in a different order. It is no longer Wayland, but a combination of Wayland, Vulkan and extension protocols. There won't ever be a feature parity in the same way X11 did it. It could be argued that Wayland is feature complete but not bug complete, too.

That's not an opinion, that's a misinformation and wrong fact. Fucking deal with it. Mine is fact, yours is opinion, we are not equal. My argument is superior.

The above statements are also fact. So is the statement that Xorg is an unmaintainable mess that is a poor fit to the way Linux rendering in particular and OS rendering in general has evolved. So, yes, I agree. Fucking deal with it.

[Gusar]

So we should have a way to discover where a window is in the core Wayland protocol because...?

I am not asking why the feature should exist, I am asking why Wayland should take care of that, since Wayland has no concept of windows. The compositor handles windows.

You asked "why would you ever want that", so yes, you did as why that feature should exist. And the answer is simple - people have made certain features part of their workflow. You won't move them to Wayland by denying them those features. Where exactly and how those features are to be implemented is for devs to figure out, the end-user doesn't care about implementation details. But that requires the devs to be willing to figure these things out, not go "why would you want that" or "we don't do that", like it happens far too often with Wayland.

[Weasel]

Nerd is upset a deep change in his favorite OS broke something, so now they have to learn a new way of doing the same thing. News at eleven.

Ever considered there might be a better way to do it? Like tiling, for instance? I've heard good things about sway and how they handle tiling. Although to be perfectly blunt, if I could get Gnome or KDE to support tiling the way sway does it, I would be perfectly content. Gnome, especially, already does a damn good job of getting out of my way so I can do work. Proper tiling support and one workspace per screen, and it's pretty much perfect for me.

Also, your request is an invasion of privacy at the very least, and would allow a script to push a window 100% off edge, essentially hiding it.

It's my script. If you don't like it, don't run other people's scripts.

There's no better way to do it, because Wayland can't do it. Period.

I take the tiling suggestion as a meme. The only reason tiling WMs or tiling idea in general exist is to make the population dumber, since they kill braincells at an accelerated rate. Wouldn't be surprised if the poor victims of it end up with below room temperature IQ over a couple of years of using them.

This is because the scope of wayland is drastically different from X11. Things are meant to be done in a different order. It is no longer Wayland, but a combination of Wayland, Vulkan and extension protocols. There won't ever be a feature parity in the same way X11 did it. It could be argued that Wayland is feature complete but not bug complete, too.

Then don't fucking claim it is?

Also, I can do it in Windows. It's not just X11.

So it's just Wayland here that's the problem child buddy.

The above statements are also fact. So is the statement that Xorg is an unmaintainable mess that is a poor fit to the way Linux rendering in particular and OS rendering in general has evolved. So, yes, I agree. Fucking deal with it.

Fact? Where's the unmaintainable-o-meter? Show me the measurement. Not claims or opinions.

You have no fucking clue what facts are. Get off my lawn.

[anda_skoa]

Where exactly and how those features are to be implemented is for devs to figure out, the end-user doesn't care about implementation details.

This is true but many people claim that they will only accept one specific implementation.
For example managing services with shell scripts even if the same workflow can be achieved with systemd.

They have established a certain custom solution so long ago that they can't even think a different solution might exist, even if it would be simpler.
Eventually they invest more and more effort into staying with an decreasingly supported ecosystem and get stuck even deeper in the sunken cost fallacity.

Or people who have invested into creating a custom solution that worked around missing features in their window manager of the time and insist that window manager developers need to add support for that instead of them using the much more capable scripting facilities modern window managers have built-in.

Unfortunately the overlap between people with ancient custom solutions and people with knowledge about the respective new facilities is often very small.
If there was a larger overlap this subset or persons would create and maintain compatibility interfaces.

[HEX0]

I take the tiling suggestion as a meme. The only reason tiling WMs or tiling idea in general exist is to make the population dumber, since they kill braincells at an accelerated rate. Wouldn't be surprised if the poor victims of it end up with below room temperature IQ over a couple of years of using them.

What's so wrong about people liking tiling? Touchpads cause me pain, so tiling feels a lot more comfortable. But according to you that makes us tiling users retards?

[wertigon]

It's my script. If you don't like it, don't run other people's scripts.

You really don't have any clue about security and what one could do with a "feature" like that.

Here's a hint:

#!/bin/sh
sh ./winamp heyeveryoneiamlookingatgayporno.wav &
setwindowcoordinates 1000000 0 winamp

Of course, this is the naive, obvious stuff. I don't have the brain the hackers do, and you can bet your ass some obfuscated version of this code exists somewhere.

There's no better way to do it, because Wayland can't do it. Period.

This is by design. See above.

Fact? Where's the unmaintainable-o-meter? Show me the measurement. Not claims or opinions.

All Xorg maintainers claim this. Not one, not three, but all agree the code is an unmaintainable hot garbage mess. By consensus.

Of course you are free to jump ship to something better. Like Windows. You seem to be very happy with that OS.