Recently I've been seeing a particular website/group of people gain a lot of attention in the Linux/general open source operating system community
wayland.fyi
The site criticizes x11/xorg for it's unnecessary complexity (makes sense so far) and then recommends wayland as an alternative.
Additionally, it also recommends not to use the popular wlroots, but rather neuwld/neuswc. I agree with some of the points but not their solutions.
And as someone that has been using x11 on UNIX-like systems for quite some time, I also noticed some inconsistencies and hyprocrisy in their ideas and some misinformation.
So, this page acts as a sort of response to wayland.fyi, I will try to respond to every claim and line on the main page. (as of 3. June 2026)
First of all, I will go over every relevant sentence on the main page, then I will point out some additional things.
>wayland at its core is a very simple wire protocol
As much as this is sort of true, why are we measuring pure size as the factor of which is better/good?
>the default implementation sits at around 10k lines
Again, why are we measuring pure lines of code? Also, nobody runs just the protocol, you can't compare the full X ecosystem to just the wayland protocol.
>modern renderers... result in complex messy systems
Why are modern renderers now the problem? Sure, modern desktops are giant piles of weird messiness, but wayland wouldn't fix that.
>much of the insanity surrounding freedesktop can be ignored simply by not implementing certain protocols
I don't like freedesktop either, but this is stupid. Just not implementing things is not the solution, instead, you should implement them better and more concise. Simply leaving out an originally bloatly implemented feature is not automatically good design.
>wlroots kinda sucks
Yes. but so does wayland
>wlroots is bad because it brings in a lot of badly designed and overly large protocol extensions
Well why does it even need protocol extensions in the first place?
>wlroots is extremely linux-centric
Sure, I agree that this is a problem with many originally linux-developed software.
>our better compositor library neuswc makes wayland good because it works on BSDs
Same thing as above, this is good and all but you're ignoring that there are Operating Systems/Unices other than Linux & the BSDs.
>wlroots is fast moving and unstable
Sure
>wlroots is fairly low level
I can't verify if this is true as I've never worked with it directly before, but being low level does not make it bad.
>X11 is the worst offender of beign barbaric
X is definitely very unnecessarily complex, I agree.
>libXfixes and libXfont2 therefore X11 le bad
No argument here, I hope this is a joke.
>X11 client code often looks simpler because there is a lot that you do not see
Ironically, this mostly applies to wayland aswell. In wayland it's just hidden elsewhere instead of the protocol itself.
>Wayland makes what you do more explicit
Explicit for whom?
>>Wayland being composited by default is bad
>explains compositing
Strawman.
>Here is a quote from 1998, just to illustrate how not only is it not good, it was actually never good!
>...read it yourself on their site
Yes X11 is bad but by who exactly is this quote and what is this supposed to prove?
>X11 is a network-transparent unified server with window management and compositing handled by clients, wayland is not network-transparent, because it's not 1985.
Remember this, I'll get to this later.
>wayland and x11 are not similar in any way, they're fundamentally different
Nothing to comment on here, they're both UNIX(Linux & BSD for wayland) GUI software.
>the compositor handles all input, it could easily forword said input to all clients instead of doing it the X way
This isn't really a point for wayland, it just again proves that, in this example the security model, isn't handled by wayland itself, but by the compositor.
The page then goes on to list a bunch of alternatives.
My question here is though, why are there so many forks of forks of forks?
If wayland is actually that simple, why are there so many layers to come out with an actual desktop?
Now to address the network-transparency thing.
Network transparency wasn't necessarily invented because of good processors/resources being expensive and rare.
It was invented because distributed systems existed/exist.
This point is also where their hyprocrisy becomes especially clear, in the recommended software list, they frequently recommend Plan 9 inspired things, while also criticizing X for being network transparent.
Additionally, the (I think) creator of the group, shrub, personally often promotes Plan 9 and it's software, yet dismisses network-transparency in other things such as X11.
That's deeply ironic.
Another main inconsistency is that they literally use Linux and are developing a Linux distribution. (?????)
This makes absolutely no sense, they comment on how bad X11 is, yet use something that has pretty much the same issues!
I believe they think that only GNU has the issues of complexity and unnecessary abstraction, but this just isn't true.
^At one point they literally link harmful.cat-v.org as a citation for modern UNIX being "barbaric technology".
This is again heavily ironic as the harmful.cat-v.org site they link literally contains a page about Linux being a prime example of the worst in modern UNIX.
Also, UNIX was "barbaric technology" from the start, it isn't just modern UNIX-likes.
At the start of the page they also call wayland.fyi a "special interest group", refering to the autistic idea of a special interest. But I don't think any of these people actually have autism, as someone that is autistic myself.