alt Hacker NewsIn short, this reads like a mix of valid historical pain points and outdated assumptions.
The post frames Wayland security as “you can’t do anything,” but that’s a misunderstanding. Even under X11, any app can log keystrokes, read window contents, and inject input into other apps. Wayland flips this to isolation-by-default: explicit portals/APIs for screen capture, input, etc.
Moreover, the performance argument is weak and somewhat contradictory. The author claims there is no clear performance win, and that it's sometimes slower and hardware improvements make it irrelevant. But Wayland reduces copies and avoids X11 roundtrips (architectural win). Actual performance depends heavily on compositor + drivers, and I've found that modern hardware has HUGE performance improvements (especially Intel, AMD, and Apple Silicon via the Asahi driver).
The NVIDIA argument is also dated. Sure, support was historically bad due to EGLStreams vs GBM, but this has improved significantly in recent driver releases.
Many cited issues are outdated too. OBS, clipboard, and screen sharing issues are now mostly (if not entirely) solved in the latest GNOME/KDE.
I've been using Wayland exclusively on Fedora and Fedora Asahi Remix systems for many years alongside Sway (and occasionally GNOME and KDE). Adoption has accelerated in many distros, and XWayland for legacy apps is excellent (although I believe using the word "legacy" here would be a trigger word for the author ;-).
There's no stagnation here... what we're looking at is a slow migration of a foundational layer, which historically always takes a decade or more in the Linux world.
Replies
> Actual performance depends heavily on compositor + drivers, and I've found that modern hardware has HUGE performance improvements (especially Intel, AMD, and Apple Silicon via the Asahi driver)
Author’s argument is those hardware improvements could have been had for free with X11 upgrades. I’m not saying it’s a complete argument. But talking about architectural wins sounds like conceding the argument.
It reads like a user that tried Wayland again last week, found the same issues and wrote a piece that tried to summarize why they remain sad after 17 years of waiting for Wayland to address its issues.
In my experience I have found the xdg-desktop-portal for whatever reason to be completely non functional on Arch/Hyprland. It must be an issue with my config but on x11 I never had to think about this
This reads like AI/FSD-bro speak: "no, that's all old news, you clearly haven't tried the new cutting edge model/build bro! it's all fixed now!".
> Wayland security
Okay, that's great, but why would I care? If you can implement those security wins transparently in the background, cool. Otherwise, what I care about is being able to take a screenshot, not about some theoretical "security threat" from already vetted programs I run on my machine.
> OBS, clipboard, and screen sharing issues are now mostly (if not entirely) solved in the latest GNOME/KDE.
Oh, the clipboard works mostly correctly now, after some 17 years of development? Could not have come up with a more damning statement. Complete misalignment of priorities.
> Wayland flips this to isolation-by-default: explicit portals/APIs for screen capture, input, etc.
The problem is old (and even not so old) apps don't expose those APIs so interactions like UI automation on Wayland is limited, if not impossible. I'd love to grant a specific permission just for selected GUI apps, but I can't because they don't support it.
There's a reason why RPA software on Wayland is limited to web apps inside a browser. Or something extremely janky like taking screenshots of the entire desktop and doing OCR. But then you can't interact with unfocused apps.