Sure. That’s why I wrote “mostly pointless” but the runtime stack is self-contained in Steam and proprietary games are compiled by the developer anyway. There are limits to what CachyOS/Gentoo/… can do.

This is only relevant for people exclusively running open source games from the distro’s repository.

Steam games (incl. the Windows games on top of Proton) run inside Steam Linux Runtime, a container based on Debian 11 that is identical on every distribution. Use whatever distribution you like but the optimizations of CachyOS are mostly pointless.

Somehow with XWayland enabled, the app still specifically demanded an actual X11 session

I guess it’s because Horizon can probably act as a host to control the desktop and as client to control other desktop. The latter should work with XWayland, the former not. As I wrote: RustDesk works just fine. What RustDesk doesn’t currently offer with Wayland is unattended access. The desktop that’s about to be remote controlled gets a question to confirm remote access, at least under Gnome.

My somewhat educated guess is that it’s more likely that Gnome’s permission system gets a “always allow remote access” button before a X11 application gets a Wayland port when the decade until now a Wayland port was no priority.

Heck, I had trouble installing remote desktop for my work (they use Omnissa Horizon) on Fedora, because the app still exclusively supports X11, and Fedora removed it in version 42.

X11 applications still run under XWayland. The X11 session is gone, not all compatibility with X11 applications. Steam wouldn’t run if complete removal was the case.

What’s Omnissa’s stance there? Will they port their application? Will they hire a developer to maintain a X11 session?

ditching X11 will still be catastrophic for many users’ workflows.

Are these users hiring a developer to maintain the X11 session? If not, they need to adapt then and go with the times and migrate to other solutions. RustDesk supports Wayland just fine, for example.

I’m using Linux because I like þe control; if I wanted a nanny OS, I’d use a Mac.

I’m currently trying to read your comment on macOS and whatever your X11 system does somehow glitches some characters and swallow words? You like to be in control?

Say “screw it”, shift blame on Nvidia and not do anything to support Nvidia users (halving the userbase)

So keeping the X11 session around for a decade after Intel and Radeon had their drivers ready is “not do anything to support Nvidia users”?

Or do something about it and implement what is necessary to keep them supported.

Who is paying for this task? Have NVidia users set up a pledge drive? Did any PC manufacturer?

It’s not up to Linux “to support Nvidia”, it’s up to Nvidia to properly support Linux.

old hardware. Specially old nVidia GPUs.

“Fuck you, Nvidia” was in June 2012. People who bought Nvidia hardware after that really have nobody to blame but themselves.

https://arstechnica.com/information-technology/2012/06/linus-torvalds-says-f-k-you-to-nvidia/

In reality, every time you add another store, you are essentially giving those people root on your machine.

Shuttleworth made the exact claim like 10 or so years ago about Ubuntu not being a democracy and Canonical being root on all Ubuntu machines.

Is it a line in their internal 10 commandments?

Honestly no idea why the Framework Desktop isn’t being sold in a SteamOS configuration.

And a Steam console to go with it, please.

Current state of the rumor mill is that it’ll be a companion device for the stand-alone VR headset where traditional games are supposedly being projected on a virtual giant screen.

Rudra Saraswat, who’s been at the center of development since the remix’s early days, no longer has time to maintain the project due to university studies.

Reminder: The entire effort relied on a 12 year old kid only three years ago, so of course at some point he has actual responsibilities: https://www.theregister.com/2022/02/04/rudra_sarsawat_ubuntu_projects/

I try not to put much stock in black-and-white opinions because I think the answer is rarely that simple.

• Either Flatpak is supported out of the box or not.

• Either Ubuntu 25.10 launched with broken Flatpak support or not.

• Either Snap Store allows random crap or not.

• Either Ubuntu 25.10 launched with broken automatic security updates or not.

Sometimes facts are actually black-and-white and not shades of grey opinions. Also, brushing away simple truths by insisting they are extremely complex facets of a thing the common person does not need to get involved is a common disinformation tactic, used by the likes of climate change deniers etc. since decades.

How about you implement it?

No, absolutely not. Apple is anti Linux. Giving them money over manufacturers with actual Linux support is the completely wrong. AMD is the way to go.

coreutils is, well, important. It’s fine to bring new software in, but you have to test it. And they haven’t tested enough.

They abolished alpha and beta versions because Canonical’s QA is so good, they don’t need them any longer…

25.10 is an interim release specifically there for testing new things.

You’re confusing alpha and beta versions with an actual release. A release without long term support is still a release.

the only functionality I’ve ever found missing is USB-C video out

Pretty basic feature to miss.

I hope whoever is managing this does it better than Ubuntu’s rust debacle

Honestly: Why? Apple M processor platform support relies completely on reverse engineering. It’s not a production-ready platform despite what some people keep shouting in forums. Even old M1 and M2 Macs can only reboot running Linux since this summer: https://officialaptivi.wordpress.com/2025/07/26/linux-6-17-will-be-able-to-reboot-silicon-macs/

A Rust rewrite of the bootloader is the least of their construction sites.

I forgot that systemd had been allowed to take over /tmp and /run.

According to Debian everyone is allowed to take over /run

Why on earth would the permissions on /var/lock be something for systemd to decide?

Because – as LWN explains – there no longer is an overarching standards body who makes the decision, so anybody can make up their own.

Debian’s continued use of UUCP-style locking does seem to be more than a little bit dated. The FHS 3.0 is clearly reaching the end of its useful life, if not actually expired.

Seems like Debian is more the outlier here.