Are We XLibre Yet? | TheaterFire

[-]

nightblackdragon@reddit

"Supportive" on Arch Linux where there is only AUR package. Typical probonopd.

[-]

luuuuuku@reddit

A half broken AUR package. You have to install three packages in the right order because of dependency issues

[-]

tamachine-dg@reddit

Please can we just let X11 die, it's been on life support for years

[-]

ObjectiveJelIyfish36@reddit

This is simply not an option for legacy NVIDIA cards that still work perfectly, such as Kepler. And nouveau is simply way too unstable to daily drive.

[-]

grem75@reddit

Will XLibre work with legacy Nvidia stuff? Original reports were it broke proprietary drivers.

[-]

ObjectiveJelIyfish36@reddit

I have no idea, but that was never my point. I was simply challenging the "Please can we just let X11 die" claim.

[-]

luuuuuku@reddit

But why? No one said that we should kill it today. But there is no point in trying to make Wayland features work in xserver when no one wants to do the work to maintain X compatibility.

After announcing to become Wayland only, both in GNOME and KDE development the vast majority reacted positive towards no having to make it work on X as well. There is no one willing to support it any longer

[-]

ObjectiveJelIyfish36@reddit

But there is no point in trying to make Wayland features work in xserver when no one wants to do the work to maintain X compatibility.

This post is literally mentioning XLibre and how its dev is literally doing the work.

And as always, it's people that don't even use Xorg anymore that are bitching about it.

[-]

luuuuuku@reddit

Weigelt is alone and that’s the issue. The last 17 years no one wanted to maintain Xorg. Now, Weigelt is working pretty much alone on it. But fine, that wasn’t the point. Let’s assume that Weigelt can manage that as one man army, what is next? The software must be packaged and distributed which given its broken compatibility with Xorg (his first release has breaking changes, breaking compatibility which is why input drivers must be rebuilt) is pretty challenging, so who will do that? I built rpms for Fedora to test it but I would not want to maintain a package in its current state. But even if that is not an issue, who will implement all the missing features? The main argument against Wayland is that not everything works yet but if you add features, desktop environments have to implement it, otherwise it’s mostly useless. But also on that side the vast majority of developers are happy not having to also support X11 in their code which is much easier. There is no roadmap either, what features will come and when is not known, it could take decades until xLibre has basic Wayland features.

Pretty much all people that actually have to work with Xorg as a developer/maintainer don’t want to do the work anymore (they just do it because they take the responsibility and do the annoying work for the community (or let Redhat pay developers because no one want to do it)). Who will package xLibre in the future? Given the political views of Weigelt, he won’t support Ubuntu, Fedora etc. Who will add the Features to KDE? Gnome? Who is supposed to do that? Reviving X would take hundreds of developers at many positions, after 17 years we found one, who does one thing.

But as always, it’s the X11 fans who just complain that the world doesn’t want to work for them for free. Why is it always the Wayland haters who never actually put in the work to improve X11? If X11 fans spent half the time they spend complaining about Wayland and the current X11 situation and instead actually put in the work, X11 might go somewhere. Why do you think like 90% of those who ever had to work with X think it’s broken beyond repair?

I’m not against X11 or xLibre (as I said I even tested it on the "hostile“ Fedora and packaged it for me. I’m still relying on Xorg for use cases and its obvious that Xorg won’t die any time soon (Redhat will maintain it until 2032 at least). The question should be: Is it better to continue implementing features in Wayland or is better to catch up after 17 years of stand still in Xorg with no roadmap and no developer support? Xorg will die, unless all that die hard fans of X start contributing as no one else wants to do that. And we’re talking about FOSS, so no developer should be forced to do something they don’t want.

[-]

ObjectiveJelIyfish36@reddit

Weigelt is alone and that’s the issue

That's simply wrong. 4 different people have commits in the main branch just from last week. There are also open pull requests from other people, too. The project is just at the beginning and it's already at a much better pace than what the original Xorg project was .

There is no roadmap either

Wrong again:

https://github.com/X11Libre/xserver/wiki/Are-We-XLibre-Yet%3F
https://github.com/X11Libre/xserver/wiki/Graphics-devices-&-drivers

Why is it always the Wayland haters who never actually put in the work to improve X11?

What are you talking about? The "Wayland haters", as you call them, have literally forked Xorg to revive the project and make it better???

I’m not against X11 or xLibre

Are you sure? You're attempting to criticize every single aspect of XLibre, whilst silently ignoring all the problems in Wayland (and OG Xorg), so it's really difficult to know what's your actual point is with this huge wall of text.

Is it better to continue implementing features in Wayland or is better to catch up after 17 years of stand still in Xorg with no roadmap and no developer support

In this scenario there's no such thing as "better". People will work on what they wanna work, you don't have a say in any of it. Why can't you just let the project be?

[-]

luuuuuku@reddit

That's simply wrong. 4 different people

I won't even comment on that, you clearly want to get it wrong. Even if there were 10 people, who maintains the rest? That was my point and that's not yet clear.

it's already at a much better pace than what the original Xorg project was.

Bold claim, any evidence to back that up?

Wrong again:

That is not a roadmap. An arbitary list, used to frame certain parties as "hostile" whilst ignoring xLibres hostility towards many projects and a GPU support list.

But where will xLibre go in the future? Why should anyone care? As said, even if xlibre succeeds it's need support from many developers to make it compatible. Xorg has a proven track record and has maintained stable releases, backed by Redhat for next decade, why would anyone choose a non compatible fork that does the same?

So, if they want to succeed, they'll need to announce what direction they're going to.

have literally forked Xorg into XLibre to revive the project and make it better

Weigelt did, after 17 years of standstill. If there are so many working on it, why is it dying?

Yes, there is an active repo now, but most contributions aren't significant for the code. There is a large quantity of contributors trying to put their political views, personal dislike and conspiracies into the project.
The future of xlibre will depend on whether they can get rid of those, otherwise it won't do anything but splitting the community.

People will work on what they wanna work, you don't have a say in any of it

It's funny only one side is allowed to do that (which was the primary reason for the fork)

[-]

luuuuuku@reddit

If I understand it correctly, it might work, you just need additional steps (all input drivers are "broken" right now and need to be rebuilt to support xLibre). Issue is that there is no usuable documentation, so in reality we don’t really know.

But compatibility is likely not something that xLibre will bring

[-]

aliendude5300@reddit

Nvidia Kepler cards are more than 13 years old. How long should they be supported? That's ANCIENT in hardware terms.

[-]

ObjectiveJelIyfish36@reddit

So people should just throw away perfect hardware "just because"? Fuck the environment, am I right?

And so what it's a 13yo architecture? The official driver for it supports Vulkan 1.2, DirectX 12 11_0, OpenGL 4.6, and NVENC/NVDEC, which are features more than enough to run AAA games from its era, let alone render a freaking desktop.

[-]

aliendude5300@reddit

No. They can run it until it breaks on the last LTS release that supports their hardware, long after driver support is dropped. They just won't benefit from running the latest software.

[-]

ObjectiveJelIyfish36@reddit

They just won't benefit from running the latest software.

That's unacceptable because there's nothing TECHNICAL preventing the Kepler hardware from "running the latest software"; It just that the official driver doesn't support Wayland through GBM, hence owners of that card are stuck to Xorg.

And the proof of that is that mesa developers are working on adding Kepler support for NVK as we speak, which will hopefully allow people run on Wayland sessions in the future.

Until then, though, projects like XLibre are more than welcome — assuming it intends to maintain compatibility with older NVIDIA cards.

[-]

CrazyBranzy@reddit

Bro is fighting for vintage tech

Continue!!

I want old stuff to live and show it's powers :3

[-]

ReservoirPenguin@reddit

That , and not everyone has money for the latest Pentium 4.

[-]

ObjectiveJelIyfish36@reddit

You can't be older than 12 years old.

[-]

ReservoirPenguin@reddit

Man, so clueless. I run the original Titan, it's still a very capable card AND it provides 1 TFLOPS of fp64 for my fluid simulation. Just in case you wondered GTX 4090 from 2023 also provides 1 TFLOPS of fp64.

[-]

grem75@reddit

Are you under the impression that unmaintained proprietary drivers will work well with this?

[-]

GregTheMadMonk@reddit

If X11 needs to be "let die" in order for Walyand to flourish, I have serious doubts about Wayland's readiness. And if it's ready, then it should be indifferent to whether or not people still choose to support/use X11.

[-]

tamachine-dg@reddit

I have serious doubts about whether it's a good idea to try and keep X11 alive through revival projects like X11Libre, despite the fact that it's outdated and has insane security problems like any program being able to read any keyboard input even if it's not focused

[-]

barfightbob@reddit

insane security problems like any program being able to read any keyboard input even if it's not focused

This is the most crazy talking point around the whole X11/Wayland thing. Every operating system, every graphical environment has been doing this for more than 30 years. It basically boils down to "don't install malware." Don't install untrusted software or use sandboxing or isolation to protect your most sensitive stuff.

Regardless XLibre has added additional security and although I don't understand X Security and X Name spaces my assumption is that it handles most concerns.

[-]

shroddy@reddit

use sandboxing or isolation to protect your most sensitive stuff.

But that sandboxing does not do shit, if the sandboxed program can not only read but even send keyboard input, that's one reason Xorg does no longer cut it. XLibre will have support for namespaces that prevent it, we have to see how well that works.

[-]

barfightbob@reddit

That's when you would use isolation. Run it in a VM.

But honestly, if you don't trust a program and are afraid it will act maliciously, don't run it in the first place. There are plenty of other more sensitive attack surfaces other than the desktop environment.

[-]

tamachine-dg@reddit

What are you suggesting? That devs shouldn't bother implementing security measures, because anyone can just install malware?

[-]