Everything.

I appreciate the information, and I’m willing to give it a shot again when I next need to do a distro switch or a new installation, but until now my experiences with Wayland have basically been a stream of broken things over several days as I try to reestablish my workflow in a new desktop environment. The time it all goes successfully is the time I’ll be sold.

Like I said, I use Linux in my classroom, and I heavily use global shortcut keys set via script for individual lessons, with fullscreen opening of applications that don’t have automatic support and shortcut key based window switching all without mouse input to create a seamless presentation for my students.

Global shortcuts and wmctrl, which form the critical backbones of this system, simply don’t work in Wayland.

And to suggest it’s just a perfect transition is wrong. I don’t use Steam Link, but if I did? Doesn’t work in Wayland. Everyone constantly bemoans that applications should be rewritten for Wayland, but one of Linux’s advantages is eternal backwards compatibility so software can actually be FINISHED.

Wayland isn’t the kernel and it shouldn’t be held to the standard of the Linux kernel, but do you remember when Linus Torvalds publicly screamed at and berated a developer for a change to the kernel that broke a userspace application and then having the sheer GALL to suggest the application developer was at fault? Wayland evangelists could stand to be a little more understanding that people don’t like it when you break functional userspace applications, force developers to work on stuff that is FINISHED to get it working again, and then blame them for not getting on board with your changes. You know who does that? Google.

Look, Wayland works for you and that’s fantastic. Use whatever you like. Linux is Linux and one of the most beautiful points of Linux is freedom of choice. What I take exception to is the people in this thread who are acting like anybody who isn’t on Wayland is crazy and insisting there’s no good reason to still be on X11 just because they personally don’t understand why someone would need features they need. Anyone expounding that “Wayland is a 1 to 1 replacement for X11 and superior in every way!” is either being intentionally disingenous or a cultist. You know who insists users are wrong for having their own use cases and workflow and wants them to change to their preferred system because THEY don’t think the other use cases matter? Microsoft.

I’ll be happy to make the switch to Wayland… when I do a system install or update and it happens invisibly and I don’t suddenly have to wonder why all of my custom scripts no longer work.

It’s not that I have issues - it works just fine in the domain it’s designed for. It’s that the Wayland system does not provide feature parity with X11. I make extensive use of window manipulation using xdotool and wmctrl for my daily use case, and those are both unsupported on Wayland. It’s a fine system for users whose use case fit with its design. It is not a feature complete replacement for X11.

I’ll never make the claim that X11 is perfect, but my use case requires features that are either not built into Wayland yet or simply won’t be built into it in the future.

I’m sure it’s a fine product, but asking me to change my workflow to use it is a non-starter. When it reaches feature complete support of X11 functionality, I’ll consider changing.

Just off the top of my head, Linux Mint, which I know because Waydroid is incompatible with the machines I use in my classrooms. Even if it were compatible, unless the lack of global hotkeys has been addressed changing is a non-starter.

That’s a fair point, and it’s the Waydroid team’s unquestioned right to use whatever technologies they want to build their software on.

But just throwing it out as a solution to a general Linux question when there’s a VERY good chance it’s incompatible with major distros is omitting critical information.

It saddens me to see you being downvoted by the Wayland evangelists when it is CLEARLY not a (EDIT: feature complete) replacement for X11 yet. If I could upvote you twice, I would.

Major, MAJOR caveat here.

No, the OSI model is fine.

I’m talking more about sandboxing an interpreted app that runs a container that runs another sandboxed interpreted app, both running their own instances of their interpreter with their own dependencies and accessible through a web interface that is accessible through yet another container running a web server that is running in Python with a virtual environment despite being the only Python app on the container, which is then connected to from another sandboxed tab on a sandboxed browser on your machine.

But hey, at least it isn’t, god forbid, a MONOLITH. That would require someone to take the time to understand how the application works.

Their UX and UI are their bread and butter, but as someone who has done extensive web app development for use on Safari browsers, if I had a nickel for every time their browser just IGNORED a standard, broke one that previously worked, or added new “features” that broke a standard, passing the responsibility of building a workaround down to individual developers… I’d have a few dollars anyway. I don’t have much faith their code is all that good compared to average under the hood and the UI, and I think their reputation unjustly leads users to turn a blind eye or give them a pass when their stuff DOESN’T work or works BADLY. “They’re Apple… everyone else seems happy. I must be doing something wrong.”

It doesn’t help that every new generation adds a new blackbox abstraction layer with little to no end-user benefit, the possibility of duplicated functionality and poor implementation, security concerns, poor support, and requiring a flashy new CPU with system crashing speed tricks to maintain a responsive environment through 12 levels of interpreters.

To be fair, I think people looked at the folks who were into mermaids as outlandish too. “Been around for a long time” doesn’t mean the same thing as “totally accepted socially”.

This works. I ran a linux distro in off hours on my work laptop for years this way.

My Nvidia card won’t properly resume the display after suspend with the default suspend script, but if I correct the script file, every time aptitude updates the nvidia drivers, it restores the bad version of the configuration file. If you set the file immutable with chattr, aptitude throws a fit and goes into a broken state when it can’t overwrite the file on a driver update.

So I keep a good copy of the script file in the directory, and in my pre-suspend script file I overwrite the main suspend script with the good version. Every single time.

Red Star OS!

Possibly. But it’s also pretty common in many instances of technology adoption that as more users come, the quality gets worse, and while open source doesn’t have to worry about a shareholder-driven profit motive driving it, it’s still easy to wind up with a muddled focus. I wouldn’t expect that Linux and all of the associated software projects that make the functional desktop are going to be an exception overall. If you’re an open source developer working on a project now, basically any user is some form of power user, and it’s easier to find consensus of what to prioritize on a project not only because Linux users tend to be better about understanding how their software works and are actually helpful in further development, they’re also likely to direct development towards features that make software more open, compatible, and useful.

Now fast forward to a future where Linux is the majority desktop OS, those power users are maybe 5% of the software’s user base, and every major project’s forum is inundated with thousands of users screaming about how hard the software is to use and, when bug reports and feature requests are actually coherent, they mostly boil down to demands for simpler, easier to understand UIs. I can easily imagine the noise alone could lead to an exodus of frustrated developers.

Some things are better for NOT trying to be the answer for everyone.