I'm not specifically happy with Wayland, I don't care much actually. And I don't have any animosity against X neither. X is an iconic project, and inspiring too. I also spent almost 20 years using it.

But it is dying. Its design doesn't correspond to the way things work anymore. It is a nightmare to maintain because its (old) code is incredibly complex, because it is full of features, and these features need to be kept for backward compatibility but aren't used much anymore. Modern toolkits actually bypass its features when they can: instead of drawing using X directives, they stream huge pictures they draw themselves. X wasn't designed like this, but it was getting in the way so this was shoehorned into it. It's now common to have two screens with different dpis, which X doesn't handle. With X, drivers are provided with X, but it is now normal to have the drivers in the kernel instead.

The X developers went away to design and develop Wayland to start from a clean slate, with all the insight gathered all these years. Drivers and apps are now developed against Wayland and are less and less tested on X. Things are going to startup breaking, I suspect librairies will startup removing X support at dôme point, etc.

X is not an alternative to Wayland. It is its past. I don't wish X to die at all. I have no feelings here. But that's what's happening.

I don't mind someone using X knowing all this. But we are commenting on a guide directed towards newcomers. Because all these reasons, making them use X is bad advice.

If you have a working install using X, there's no urge to switch. Don't fix it if it ain't broken. It might even be clever to follow the upgrade path of your distro if you didn't customize anything.

Apart from that, I don't see any reason today to use X instead of Wayland unless specifically something doesn't work. And if it is the case, I hope you reported the issue, because one day you might be forced to switch and if the bug is still there, you will have to live with it.

We are not in a situation where we have two competing alternatives between which you can freely chose. Both involve the same developers, and one of them is considered obsolete and isn't developed anymore.

It's not that we are Wayland fans (at least I'm not). It's just that Wayland is all we have going forward.

It's not because it's just a different alternative. The problem is that X11 is not pushing the graphics stack forward while at the same time dragging down on projects that feel they need to support it because some users don't want to let go of X11.

How long until running software designed in 1987 is reserved for retro-computing? The graphics stack has changed way too much since then.

(damn, the X protocol v11 is even older than I thought, but I guess Wikipedia's entry is not considering minor revisions to X11)

---

https://en.wikipedia.org/wiki/X_Window_System

The graphics stack doesn't need to be pushed forward for, what?, 99% of computing use cases? 99.9%?

What graphics stack pushing forward is needed to write documents, spreadsheets, view charts, write code, edit video/images, do data analysis, etc?

I wouldn't says "push forward the graphic stack" as much as fitting current hardware and architectures that are already there, and being able to apply the insight gathered on X all these years.

Writing code and doing spreadsheet was already possible on the Atari ST but you wouldn't use its software stack today.

> The graphics stack doesn't need to be pushed forward for

Hardware evolves and software should either grab its magic ball and plan for that, or adapt to changes to avoid dragging hardware evolution back.

What does X offer for new users that Wayland doesn't? Alternatives are great but they have to be different in some way. I moved to Wayland more than 6 years ago (hardly an early adopter) and the only difference is less pain. Why would a new user choose that pain and learn archaic X tools? As an existing user it's different and you would only move it Wayland offered you something, but for new users Wayland should be the default now.

>What does X offer for new users that Wayland doesn't?

(Besides stability) it offers more options. The PineNote is an e-ink stylus tablet that only supports Wayland, and I've been looking for a dead-simple mouse-driven compositor (because tablet). X had bucketloads of WMs, most of them long-since stabilized and featured complete. Wayland has nothing. Nothing in the Debian-stable (which comes shipped on the PineNote and is stable) repo, at least.

Is this somewhat niche? Sure! Is it a huge fucking pain in the ass? Also sure!

I'm not against Wayland, in premise - it supports process isolation (potentially useful for KDE Activities in some amazing ways) in a way that X just can't and never will. X is not the future. But frankly, neither is Wayland. It's barely even the present (ask wacom/Krita users, or anyone with accessibility needs, what they think of Wayland devs), and actually loses some features (multi-window stuff is impossible-by-design on Wayland, which means entire swathes of software will just literally never switch to Wayland), and while Wayland does gain features and possibilities, they're niche features just like the X features that Wayland broke.

The "X is a clusterfuck" argument becomes increasingly weak, since Wayland is more and more clearly a clusterfuck itself.

At this point, I think Arcan is the future (www.arcan-fe.com), although please note the main dev is specifically avoiding the limelight before v1.0 and thus writes in a deliberately dense and unapproachable fashion. It's super elegant and a genuinely groundbreaking project.