The Linux Thread - The Autist's OS of Choice

[Gargamel]

Absolute nonsense. X11 was designed to display windows on screens.

And it works great what it was designed for, which is displaying un-accelerated 2D graphics, on one screen. Who does that in the modern day? Nobody, that's why they built all these hacks and extensions to make X11, a network transparent client-server model display system, do all this shit that it was never designed to do.

It sorta works, depending on your hardware and your expectations. Have you tried 2x scale on a two head setup using 4k monitors? How about hotplugging one of those 4k displays? How about 3D acceleration? Tear-free scrolling on Ivy Bridge, or really any Intel iGPU? Why the fuck are there three different mostly-broken drivers listed on the Archwiki for my Alder Lake machine and why do none of them just fucking work? What about the RX 5700XT in my Threadripper box? When will that work properly? And when the fuck will Chrome-based broswers be 3D accelerated on X11? Will it ever happen?

No, of course not. Nobody wants to keep patching up X11. Nobody wants to put effort into the legacy system that's already being migrated away from. It is time for something new, and something new started being worked on like 10 years ago, and that something new is called Wayland, and it is usable trash, but as I said, it has potential to rise up. It already does so many things better than X11. It just needs more time and development effort. The only people resisting the move away from X11 are people who don't want to let go of their favorite X11 window manager, people who need X11 features not present in Wayland, and contrarian niggers.

To be fair, X11 works perfectly on my old ThinkPad with it's might-as-well-be-2D Neomagic graphics chip driving a 1024x768 display. Last time I tried, which was a long time ago, it also worked pretty good on my 486 with a Hercules card driving an MDA display.

It was flexible enough to operate well across a network

It works great for that too, if the X11 client and server are pretty near in age. It is a real pain in the ass getting a modern Linux X11 server to play nice with my early 90s HP logic analyzer, which contains an X11 client.

EDIT: X11 servers have been being subjected to intense security testing for longer than any Wayland faggot has been alive. Some final minor bugs were detected with fuzzing back in 1995, FFS.

Security doesn't even enter into my problems with X11 and Wayland. But I can only imagine the clusterfuck that must be modern day X11 with all those post 1995 hacks and extensions...

 

[King Of The Kiwis]

Good news guys I used a different USB and the installation worked like a charm. I'm a Linux autist now just like you guys

 

[Susanna]

Good news guys I used a different USB and the installation worked like a charm. I'm a Linux autist now just like you guys

Only if you installed Gentoo or NixOS or something. Otherwise you're basically just a windows normie.

 

[AmpleApricots]

The thing people on both sides of the fence often don't truly understand about Wayland that that what you see now is pretty much all that protocol intends. It won't get "better" because that's fundamentally it. It's the linux equivalent to that meme picture "how to draw an owl" basically leaving everything to whoever actually tries to implement anything. It came to be by a bunch of nerds bikeshedding about standards and maybe huffing their own farts a bit. Follow the progress of the wayland protocol over the last decade with an open mind to what the people involved with it are actually doing and saying and you'll understand it'll never go anywhere. It's the perfect example of linux userland made by people who don't actually use linux.

 

[Gargamel]

The thing people on both sides of the fence often don't truly understand about Wayland that that what you see now is pretty much all that protocol intends. It won't get "better" because that's fundamentally it.

Gnome, KDE, and couple big distros (or perhaps Red Hat) will push hard to make their extensions to the Wayland core the most widely used ones. The most popular combination(s) will become defacto.

It came to be by a bunch of nerds bikeshedding about standards and maybe huffing their own farts a bit. Follow the progress of the wayland protocol over the last decade with an open mind to what the people involved with it are actually doing and saying and you'll understand it'll never go anywhere. It's the perfect example of linux userland made by people who don't actually use linux.

This is well and truly unfortunate, and it's a shame we couldn't have done better.

 

[Jeff Q. Anime]

x11 without support for mixed refresh rates simply makes it unusable for serious gaming in 2024

 

[Betonhaus]

Is there a way to cap uploading to wan without capping upload to lan?

 

[Sperg Coalition]

I don't really buy into his Wayland salesperson lingo but it has the net-positive effect of exposing the behind it, more and more randoms are picking up on how batshit the situation is with display protocols/servers and its surrounding discourse.

I get the feeling that this lunacy is also making people switch back to Windows due to having less bullshit to deal with ironically... The state of Desktop Linux is pretty sad ngl

-snip-

I'll have to agree with @AmpleApricots on the Wayland situation and I'll say that both protocols simply are too outdated for a "modern" desktop experience: I think that having a monolithic kernel does not help since it is difficult to, say, move the graphics/windowing system in the kernel like Windows did from NT4 onwards.

 

[AmpleApricots]

Eh, I'm personally happy with X. I'm used to it and it works for me and the workflows I have and the hardware I am using. The bloat and complexity argument might apply if we still had early 00s machines but for anything not-vintage X is not really complex software. There are a lot worse offenders in daily use nowadays.

Microsoft Windows is NOT an OS to look at as an example of something not burdened with tons of legacy bloat.

People always want these "pure, lean" implementations of some software, but in reality for some things you just gonna have a heap of ugly edge case code if you want it to be useful in real situations. Just the way it is. A lot of software starts out with that noble goal, ends up having to implement all the workarounds and edge-cases the software it was meant to replace also needed in order to be as useful and not rarely ends up with having no tangible benefits over that older software in the end.

Some people also seem to be violently opposed to the "fracturedness" of Linux userspace (the existence of both Wayland and X as choices, the many kinds of different init systems that exist etc., you can start flamewars in some places just by mentioning you *gasp* don't use systemd) which I always thought was one of Linux biggest strengths actually, because if you didn't like the decisions made in one part of the collection of software your OS and workflow is, you can just switch to another one. If there is only one implementation of a needed software because of it's complexity making it basically impossible for smaller groups to offer a different solution, it will not lead to a good solution that will benefit all, but to (nowadays most likely corporate-backed) tyranny by a group of a select few who can hold the entire userspace as hostage via chokepoint and enforce their design decisions and rules if they choose to do so. Tons of examples of projects attempting to do just that. I will always oppose the homogenization of Linux userspace software on principle because of that. The more ways to do a thing, the better. I honestly don't care if the linux desktop experience is "newbie/normie/casual user friendly" because I don't think that'd add any value for me, quite on the contrary. I don't even feel like pretending I care about that like many other linux users seem to do. I'm perfectly fine with telling people that if they find using a Linux system too complicated, maybe Linux just isn't for them and they should stick to another OS. This "Let's make linux usable for casuals" is a net-negative IMO. If the price of that is less corporate fingers in the cake and somewhat more esoteric setups of software which doesn't work "out of the box" like MacOS/Windows equivalents, honestly, so be it. Having everything dumbed down to the lowest common denominator as decided by a "central linux userspace comitee" (most likely by corpos and career-ideologists with their own agendas, sticking their CoCs everywhere - because such hierachies just attract the worst kinds of people) is in my opinion not really a good thing.

 

[cipheranarchist]

I use a custom build of Openbox which runs on X11, as I understand it is unable to run on Wayland. I'm not sure how to move forward with it, with the lack of support for Wayland and X11 increasingly looking like it has no future beyond its current design.

 

[Rusty Metal Skull Gun]

I use a custom build of Openbox which runs on X11, as I understand it is unable to run on Wayland. I'm not sure how to move forward with it, with the lack of support for Wayland and X11 increasingly looking like it has no future beyond its current design.

Don't believe the BS about X having no future. It's been around for decades, and there's no reason it won't be around for a good while yet.
At the very very least, it's going nowhere until Nvidia cards work flawlessly on Wayland. Nvidia is big enough that it can't be ignored.

 

[cipheranarchist]

Don't believe the BS about X having no future. It's been around for decades, and there's no reason it won't be around for a good while yet.
At the very very least, it's going nowhere until Nvidia cards work flawlessly on Wayland. Nvidia is big enough that it can't be ignored.

Yeah I must admit I have a certain amount of skepticism about it all, still. I guess it's why I've been sitting on the fence and been in no real hurry to ditch X11.

 

[Trans Fat 41g]

Open Source has had a knack for keeping decades old software and even abandonware (not saying X11 is abandonware) alive, even if Wayland railroaded Xorg out of the mainstream support, someone else would take up the role of maintenance. It isn't a matter of if, but a matter of when new work gets implemented, even if future changes are "just because we can, so what?".

Never underestimate persistence.

 

[Gatdam Animal Person]

Killing X11 would be like Microsoft killing the Win32 API: total retard destuction.

 

[GNU Abyss]

I don't think anybody is abandoning X11 completely, backwards compatibility exists in the form of XWayland and will likely be receiving maintenance going forward. Still don't trust anything touted as New Thing Good, especially when it has a strong smell of IBM/Red Hat. Xenocara port from BSD to Linux was discussed earlier in the thread, that would be good news. Hopefully the FVWM dev doesn't develop any brain worms in the meantime.

 

[jeff7989]

Wayland is a barebones trash full of artificial limitations imposed by trannies who develop it (Look up "female" developers who work on it). Those limitations are excused by "muh security", which in 2024 is a big red flag to me.

 

[Gatdam Animal Person]

Still don't trust anything touted as New Thing Good, especially when it has a strong smell of IBM/Red Hat.

Sounds like the classical "embrace, extend and extinguish" technique, where you have an army of paid Red Hat developers making rapid progress, but at the same thing swaying the project's direction to their benefit.

IBM is playing their long game since the OS/2 era.

 

[Kosher Dill]

There is a statement from Prof. Dr. Stallman on this. It isn't a real, direct quote.
https://www.gnu.org/gnu/incorrect-quotation.en.html

My favorite part of this is that the first line isn't italicized. Because he would like to interject.

 

[Miller]

X11 is old, cobbled together, and never designed to do what it currently does. Wayland is trash. They're two kinds of garbage.

Wayland might rise above trash status if enough people put in the effort. A lot of progress has been made in the past couple years, if Brodie's videos are anything to go by. It has a long, long way to go.

X11 does not have a future.

Why not fix things and call it X12?

 

[Trans Fat 41g]

Why not fix things and call it X12?

OpenBSD's Xenocara is a patch set for X, we just need to factor it in for Linux usage provided whatever other process is necessary to appease distro jannies, I suppose (fuck that)...