Long live Xorg, I mean Xlibre!

Updated: June 13, 2025

Imagine a software project that's been 15 years into making. A project that, after all this time, is still rather beta in quality. A project that can only do a portion of what its predecessor technology could and can do, and yet it is hailed as a "modern replacement". A project that no one really wants to use, as it's cumbersome, it breaks a lot of things, and doesn't do what it ought to. A project that is now being forced onto the users through arbitrary decisions, because it's the only way it could ever possibly be adopted. You would think this is something coming from a greedy big corpo like Apple or Google or Microsoft. Nope, it's the open-source "darling" Wayland.

Well, the FOSS community seems to have a reached a nice inflection point. Rather than embrace an inferior solution as the "way forward", there's a new contender in the display protocol space. It's called Xlibre, and it's a fork of the old and trusty Xorg (xserver). The goal of Xlibre is to modernize Xorg. I liked this news so much that I decided to write an article about it, even though there isn't a product for me to use, just yet. But sometimes, a story is all that is needed. Let's talk.

Dedo, why do you hate Wayland?

Let's answer this question, as it ought to come up. After all, in the Linux space, ad hominem is often a more powerful way of deflecting resistance than debating technological merits of software. And the answer is simple:

I do not hate Wayland. I have no personal beef with it. I don't really care. I'm not a developer. I am not bothered by software. It's a means to an end.

What I don't like is ANY, I repeat ANY software solution that champions mediocrity.

For example, Skype 8 over Skype 7. Microsoft removed the sane desktop-oriented Skype 7, which let you do multiple chat windows at the same time, and replaced it with a pseudo-touch "modern" turdling that could not do this thing, and many other things. New solution, inferior to old solution. Missing functionality. The end user suffers. Simple.

Another example. Microsoft introduced Windows Settings as a replacement for Control Panel. Many years later, Settings is still functionality inferior. It requires more mouse clicks to get to the right place, you get less information due to "visual minimalism" AKA "touch nonsense", it's ergonomically inferior (color, contrast, window borders, scrollbars), it responds more slowly - the UI latency. In other words, a modern turdling. Simple.

Yup, it is that simple. As an end user, do I get my functionality, yes or no? And the new thing must always at least match the old thing, and ideally surpass it. That's all that matters. Everything else is dev noise.

I don't care about programming languages, I don't care if Rust is more leet than C, or if Java has better garbage collection than whatever. Programming languages are entirely interchangeable, as there isn't a single unifying thing that makes this or that language superior. None of that.

Speaking of inferiority, what we have is the exact same thing in the Linux sphere:

• Wayland is supposedly a replacement for Xorg/X11, whatever. I don't care. Does it do its job? No. Or rather, more accurately, not as well as its predecessor.
• Wayland cannot do (or do well) tons of things: VNC server, remote desktop, SSH X forwarding, custom keyboard bindings, numerous accessibility options, legacy software, absolute desktop positioning, screen sharing and recording, CAD/EDA tools, performance, and then some. All of these are critical use cases, but they don't seem to matter in the shiny happy new future that Wayland brings.
• You think I'm ranting? Well, even if we ignore a gist that thoroughly explains all the gory bits, what about important software, like the aforementioned CAD/EDA stuff? Like what KiCad folks think of Wayland? This is just a wee example. You are welcome to peruse the annals of the Web, and I am certain you will find many more lovely stories of where Wayland falls short.
• Now, Wayland is heavily championed by Gnome (and vice versa), which sort of creates a conflict of interest. I remarked on this in my Fedora 42 review. With planned forced obsolescence of X11 components from future Gnome and Fedora editions (and, now it seems, Ubuntu, as well), the user is left with no choice. The supposed freedom of choice that Linux offers. Right. As an aside, Canonical doing this thing is a horrible, horrible choice, for Canonical that is, as it effectively makes it entirely at the mercy of whatever Gnome decides, and quite possibly, Red Hat, given the strong connection between the two entities. But hey, their choice, so whatever.
• On a functional level, this is a classic Microsoft Windows 11 TPM move. Just think about it. The only way to get distros to use Wayland seems to be by deliberately killing off old components. Forcing users to run beta-quality nonsense that still can't do tons of vital stuff that the old thing can. After 15 years!
• Even if you ignore everything else, accessibility. Of all things that Wayland doesn't do well, what about people who struggle with "modern" crap due to actual disabilities? I guess old and disabled people don't count, they aren't leet devs with nine monitors and a dark theme desktop, it seems.
• The whole forced Wayland story is sad. Amazing logic right there. The KDE team seems a bit more sensible when it comes to this, as the goal is to go Wayland only by Plasma 7, but it is still an arbitrary decision. It does not say: "WHEN THE PROJECT MEETS ITS FUNCTIONAL REQUIREMENTS". Nope, this is BY DATE XYZ, we will use it, be what may. Why? Because the only way to make failure into a success is by ignoring all the other options? The only way for Wayland to come first in a race is to disqualify all the other opponents? Mediocrity in its supreme form.

If this were to have been decided by a big corpo name, the FOSS folks would be seething and foaming with fury and indignation. I know I hate every little Microsoft and Google shenanigan with passion. But apparently, doing arbitrary choices like this in Linux seems "fine". The same applies to many other similar and intertwined components in the "modern" Linux desktop stack, like the entirely unnecessary systemd and PulseAudio, also over-complicated, quality-lacking solutions that don't really live up to their promise or offer any significant advantages to the common desktop user.

The best, or the worst part is, Xorg ain't dead. It's actively developed and maintained. But there's forced obsolescence, and so, we now have a fork. There's a new project called Xlibre, which should focus on modernizing the Xorg stack (and undeniably, it needs that). Seems like a better, more elegant and far less resource-intensive approach than Wayland. And it also promises to be a complete solution, unlike Wayland, which isn't, and the way it's designed, cannot be a complete solution.

The question then becomes: do you go for Wayland, pretend it's "all fine" and ignore the 20-30% important use cases it will never satisfy, or you go with Xlibre, which should satisfy all the legacy cases (already does), and whatever the future brings?

The answer is obvious.

So what should we expect, then?

I guess, the same old desktop, as always, plus modern perks. Indeed, I can attest to Xorg doing its job superbly well. I've got a bunch of physical (and virtual) Linux systems, and they all use X11 sessions. HD/UHD displays and fractional scaling, check. Nvidia drivers, check. Intel, AMD drivers, check. Hybrid graphics with Nvidia PRIME, check. Display color calibration, check. Nested X sessions for old games on modern displays, check. Manually remapped keyboards, checks. Steam and Proton, check. WINE software, check. It's all there.

This means, if the Xlibre team does its job well, we should get even more quality in the future. The only open question is, when and how will Xlibre get adopted across the distro-sphere? That will definitely an interesting development of its own, as it surely will reflect the "political" tug of war in the FOSS sphere. It is entirely possible this project will be shunned due to pure ego games.

Now, I personally don't care about politics at all, on any level. I don't care about developer drama, their flame works on social media, who said what, or all that nonsense. All I'm interested is high-quality software solutions that focus on providing superior functionality to the end user, without regressions, without taking away critical tools. Alas, this doesn't seem to be the chief consideration in the software world, big corpo and Linux included. But now, we might actually get a sane display server protocol whatever.

A challenger had appeared. Great. Wonderful.

Conclusion

Typically, I am opposed to the constant forking and reforking in the FOSS and Linux world. Someone doesn't like something tiny, boom, fork. This is usually how it works, and why we have 300+ distros, most of them derivatives of a basic set of four or five, with only 5% variation among them. But in this case, it is necessary. Wayland is simply the wrong solution. If somehow, magically, it fixes all its problems tomorrow, then great, fantastic, thumbs up, I'm all for it. Only it won't, and it can't. And thus, as a threat to legitimate end user needs and important desktop functionality, it shouldn't be promoted or adopted. Not until it at least reaches functional parity with X11 (which it can't). But even then, it ought to surpass it, otherwise, what's the point of the last fifteen years?

Xlibre might be the answer. Now, it might also not be the answer. For now, there's great hope. The proof is in the pudding. Xlibre will need to show it can deliver, that it's stable, robust and mature, and that it can meet the requirements, current and future ones. At the moment, Xlibre seems like it's the best potential solution. Well, I guess I said everything I had to say. Bon voyage, and party on!

Cheers.