Tried it on PopOS and wondered how anyone could use it at all. Installed fedora on a different machine and it’s flawless. Probably just the age of PopOS at this point.
I don’t know what kind of program it is, but if it works on X11, you could try forcing it to run inside Xwayland by unsetting the WAYLAND_DISPLAY variable. I’ve had to do this with Qt Creator because dockable windows didn’t work at the time on Wayland.
I think it’s more of a consensus than a debate for those who had an nvidia card. a bunch of things don’t work as they should, because nvidia drivers are bad.
Even UEFI has its haters, and they’re calling for a return to BIOS in the aftermath of the Gigabyte UEFI vulnerability: https://www.youtube.com/watch?v=BTjj1ILCwRs (scroll through the comments)
You bet. Not many, but they are extremely passionate emotional about it. They mostly grace the anti-intellectual cesspools with their presence (twitter and such).
I used to be like that, about two years ago, mainly because of some bad experiences with compiling Rust programs from source. Then I realized that I’m literally never going to be affected by it since I never compile the kernel myself. Now I’m learning Rust myself.
It’s not just about memory safety, honestly. Rust is a modern programming language. C is a fucking fossil buried in a tar pit of unmaintained libraries and workarounds. A newly minted programmer won’t likely choose C as their primary language when things like C#, Python, and Rust are available (I work in a university, I see what our students do), and the current generation of C-only maintainers will eventually retire, with nobody to replace them. Being openly hostile and abusive against Rust will hurt the kernel in the long term.
This is the Dreadnought effect once again. A new thing appears that thoroughly changes the landscape, and the people who can’t adapt push back with fear and violence. Crossbows and arbalests have been effective for centuries, who needs this ridiculous new “gun” thing?
I mean, take a look at it from their POV: You’ve spent decades building this and suddenly new people come in and want to change very fundamental things in the project. Replacing them with something you don’t know and don’t want to learn.
Change must be slow and steady and shouldn’t make the original developers feel like they’re being pushed out.
Both sides are severely lacking in considering the people on the other side.
That’s not what’s happening. Right now, modules written in Rust are completely independent from modules written in C. If a Rust module wants to interact with a C module, it has to use that module’s API, as it is intended. If that API changes, the Rust module has to adapt to it. Nobody is forcing C developers to change anything they don’t want to. In the latest tantrum, a kernel developer called Christoph Hellwig was exposed sabotaging Rust projects: he was NACKing Rust PRs that interacted with the DRM API even though they touched none of the code he was maintaining, nor required any changes be made by a C developer. In the fallout, two Asahi/ARM developers left the mainline kernel, and one of them retired completely because of burnout.
Wrong opinions can exist if they are formed using incorrect assumptions. Hellwig had no right to sabotage the Rust projects, and certainly no right to call Rust-for-Linux developers a cancer on the LKML.
People make it sound like its some extreme time consuming task to learn rust. Rust actually gives helpful compiler errors tho and there’s a lot of resources online.
I was able to start making some basic things in rust (like an ascii-rendered brute force n-body simulation) with the help of a few google searches after just like 2 days of messing around in my free time. I’m sure reading kernel stuff requires much more advanced knowledge than what I have but it’s really not a large barrier.
They are using a shitty language that does almost nothing for them and doesn’t guarantee the final binary will do anything close to what the source implies. There’s no sides to this. It’s just reality.
If they aren’t trying to migrate away from it, it’s not good. If they actively insist on harassing the people trying to help them migrate away from it, they are the problem.
Linux users are peaceful* and level-headed*.
* barring discussions about Wayland, X11’s obsolescence, Systemd, Pipewire, Rust in the kernel, or even UEFI at times
Don’t forget arch btw
It’s NixOS and CachyOS these days.
Do you even nix, bro? I have Nix on my company issued Windows laptop inside WSL, btw.
Do you even NixOS bro? I have NixOS on my personal issued Thinkpad inside
/, btw.I tried to use Wayland. My windows flickered to black. I switched to X11. No issue. I’ll try Wayland again next year. -casual Linux user
Narrator voice: “Six years later, they still haven’t tried it again.”
I try every year and every year I get a different result. Currently on Wayland, next year’s update might force me back to X.
Love the Wayland stability.
Tried it on PopOS and wondered how anyone could use it at all. Installed fedora on a different machine and it’s flawless. Probably just the age of PopOS at this point.
Slic3r doesn’t work on it.
No idea of why. (But I suspect it’s about the several monitors thing.) Will probably try again in a year or 5.
I don’t know what kind of program it is, but if it works on X11, you could try forcing it to run inside Xwayland by unsetting the
WAYLAND_DISPLAYvariable. I’ve had to do this with Qt Creator because dockable windows didn’t work at the time on Wayland.env --unset=WAYLAND_DISPLAY DISPLAY=:0 /path/to/progSlic3r? Original Slic3r? Why are you still using that?
It takes longer to configure a newer slicer than to test if the old one works in wayland :)
When I need something more advanced, like customized supports, I’ll migrate. Didn’t happen yet.
variable width extrusions are great, you can get a whole lot more detail in thin areas just automatically now
There’s also tree supports (usually less material and much easier to design for and remove), lightning infill (weak infill for non-structural models)
maybe some other features that i’m forgetting
Forget features, just the basic path generator is so much better in the newer variants
you forgot nvidia
there’s a debate about NVidia ?
I think it’s more of a consensus than a debate for those who had an nvidia card. a bunch of things don’t work as they should, because nvidia drivers are bad.
UEFI is OK. Try mentioning Secure Boot.
Even UEFI has its haters, and they’re calling for a return to BIOS in the aftermath of the Gigabyte UEFI vulnerability: https://www.youtube.com/watch?v=BTjj1ILCwRs (scroll through the comments)
fuck the bios stage, just load grub from firmware
UEFI is a standard. The implementation of it depends on the manufacturer.
Do users care about Rust in the kernel? The others all make sense.
You bet. Not many, but they are extremely
passionateemotional about it. They mostly grace the anti-intellectual cesspools with their presence (twitter and such).I used to be like that, about two years ago, mainly because of some bad experiences with compiling Rust programs from source. Then I realized that I’m literally never going to be affected by it since I never compile the kernel myself. Now I’m learning Rust myself.
If anything, the memory safe Rust should be encouraged…
It’s not just about memory safety, honestly. Rust is a modern programming language. C is a fucking fossil buried in a tar pit of unmaintained libraries and workarounds. A newly minted programmer won’t likely choose C as their primary language when things like C#, Python, and Rust are available (I work in a university, I see what our students do), and the current generation of C-only maintainers will eventually retire, with nobody to replace them. Being openly hostile and abusive against Rust will hurt the kernel in the long term.
This is the Dreadnought effect once again. A new thing appears that thoroughly changes the landscape, and the people who can’t adapt push back with fear and violence. Crossbows and arbalests have been effective for centuries, who needs this ridiculous new “gun” thing?
I mean, take a look at it from their POV: You’ve spent decades building this and suddenly new people come in and want to change very fundamental things in the project. Replacing them with something you don’t know and don’t want to learn.
Change must be slow and steady and shouldn’t make the original developers feel like they’re being pushed out.
Both sides are severely lacking in considering the people on the other side.
That’s not what’s happening. Right now, modules written in Rust are completely independent from modules written in C. If a Rust module wants to interact with a C module, it has to use that module’s API, as it is intended. If that API changes, the Rust module has to adapt to it. Nobody is forcing C developers to change anything they don’t want to. In the latest tantrum, a kernel developer called Christoph Hellwig was exposed sabotaging Rust projects: he was NACKing Rust PRs that interacted with the DRM API even though they touched none of the code he was maintaining, nor required any changes be made by a C developer. In the fallout, two Asahi/ARM developers left the mainline kernel, and one of them retired completely because of burnout.
Wrong opinions can exist if they are formed using incorrect assumptions. Hellwig had no right to sabotage the Rust projects, and certainly no right to call Rust-for-Linux developers a cancer on the LKML.
People make it sound like its some extreme time consuming task to learn rust. Rust actually gives helpful compiler errors tho and there’s a lot of resources online.
I was able to start making some basic things in rust (like an ascii-rendered brute force n-body simulation) with the help of a few google searches after just like 2 days of messing around in my free time. I’m sure reading kernel stuff requires much more advanced knowledge than what I have but it’s really not a large barrier.
They are using a shitty language that does almost nothing for them and doesn’t guarantee the final binary will do anything close to what the source implies. There’s no sides to this. It’s just reality.
If they aren’t trying to migrate away from it, it’s not good. If they actively insist on harassing the people trying to help them migrate away from it, they are the problem.
Well, you should probably read the last sentence of my previous comment, it does apply here as well.
Oh no, I’ve read your last sentence. The people in one side of that issue are the problem. The issue itself has no sides.