X.Org Server 1.20.11 Released Due To New Security Advisory

[reavertm]

The core release of Wayland as the team saw it ready for general consumption was in 2014, 7 years ago.
The first compositors had a usable state in 2017 with Gnome and 2019 with KDE.
For those 7 Years, Wayland already well established itself in the industrial usage, embedded systems and ARM devices.



No you can not, X11 is fundamentally broken by design.


WDM and WDF are driver frameworks for the hybrid kernel called NT.


Yes, its a driver framework.
You mean DWM. DWM would be the modern day equivalent to Wayland on Windows and it has a compatibility layer like Xwayland to display pre-DWM clients.


X11 is the protocol, you mean Xorg, one of the many X11 implementations throughout the years.
Xorg is a set of tools and a server implementing the X11 protocol.


X11 is a protocol, Xorg is that set of tools and a server you mean.
Imagine what, there are many X11 implementations out there, all making their own assumptions about the protocol. You just lived in that short time where about everyone was using Xorg. There were times in history where every GPU producer had their own X Server implementing the X11 protocol with various extensions for graphical acceleration where applications had to support those extensions.

To be honest, I have the strong feeling that you have not the slightest clue what you are talking about.

I think you are the one who is intellectually dishonest here.
Application developers use libX11, libXft, libXinerama and all other bunch of libraries.
For consumer - application, X11 is not a protocol but API.
From application pov all X11 network-transparent interworkings are entirely opaque.
When approaching X11 evolution you seem fixated on sharing the mindset of Wayland developers - "let's reuse ancient X11 client-server architecture, except focus on defining protocols better". Implementation is afterthought.
But it could have been just as well designed with focus on providing the best API and implementation, leaving internals implementation specific and possibly subject to change - a'ka irrelevant for your DE.
Now all 3rd party implementations are stuck with the protocol and if not all use cases were identified during its inception, everyone will have to be working it around or implement adaptation if protocol is updated.

[angrypie]

The real problem is nobody is using a common framework for MAKING their compositors. I tried Sway and it detected all my displays without issue... if KDE/GNOME don't use a common EDID parser, and device management, then Wayland is a failure not so much of the protocol but for creating even more fragmentation with no usability.

Cool, tell me more about how there was no fragmentation before Wayland and how the Linux ecosystem was a happy community of like-minded individuals.

[spstarr]

The real problem is nobody is using a common framework for MAKING their compositors. I tried Sway and it detected all my displays without issue... if KDE/GNOME don't use a common EDID parser, and device management, then Wayland is a failure not so much of the protocol but for creating even more fragmentation with no usability.

[Alexmitter]

Apparently not everyone cares because it look Wayland 12 years to take off and it's still relatively nowhere.

The core release of Wayland as the team saw it ready for general consumption was in 2014, 7 years ago.
The first compositors had a usable state in 2017 with Gnome and 2019 with KDE.
For those 7 Years, Wayland already well established itself in the industrial usage, embedded systems and ARM devices.

I was asking what others think about the idea that perhaps it would have been better to evolve X11 libraries in backward incompatible manner.

No you can not, X11 is fundamentally broken by design.

Microsoft did that between WDM -> WDF.

WDM and WDF are driver frameworks for the hybrid kernel called NT.

WDM and WDF is not a set of protocols everyone has to implement on their own.

Yes, its a driver framework.
You mean DWM. DWM would be the modern day equivalent to Wayland on Windows and it has a compatibility layer like Xwayland to display pre-DWM clients.

It's set of ready tools and libraries. That's what traditional X11 was.

X11 is the protocol, you mean Xorg, one of the many X11 implementations throughout the years.
Xorg is a set of tools and a server implementing the X11 protocol.

How is Wayland approach fundamentally better than "X11" approach? Releasing protocol is not always the best approach because anyone can abuse protocol, make own assumptions about it, implement it partially. You can say "then use good compositor". And I could say "it's broken by design if room for abuses was made in first place".

X11 is a protocol, Xorg is that set of tools and a server you mean.
Imagine what, there are many X11 implementations out there, all making their own assumptions about the protocol. You just lived in that short time where about everyone was using Xorg. There were times in history where every GPU producer had their own X Server implementing the X11 protocol with various extensions for graphical acceleration where applications had to support those extensions.

To be honest, I have the strong feeling that you have not the slightest clue what you are talking about.

[cynical]

Apparently not everyone cares because it look Wayland 12 years to take off and it's still relatively nowhere.

I’ve been using it for years. It works well.

Btw, I understood your question perfectly. The answer is that sometimes the architecture of the code is so bad that to make major improvements, you must start from a clean slate. Evolving it when it’s a spaghetti code mess makes no sense because you aren’t fixing the underlying problems.

[angrypie]

The reference Wayland compositor already exists, and it's the GNOME compositor. I'm not saying this because I'm a GNOME fanboy (far from it) but because it's evident that most/many former Xorg key developers have moved on to work directly for RedHat. So it's actually even more like what you're describing: instead of having a reference Xorg server that is developed by a mostly independent team and can be used by everyone equally, we have a reference Wayland server that a single company (mostly) controls and tailors for use within its own product line (GNOME), and everybody else has to either conform to that single company's way of thinking (and its internal hierarchy) or go their own way and reinvent the wheel by developing their own server using the publicly released spec (which is practically mandatory with Wayland, because now the server is tightly integrated with the window manager).

If that's true they're doing a pretty bad job, because GNOME on Wayland still performs like a tech demo. Any CPU usage spike or I/O stalls are enough to make the GNOME Wayland experience unbearable, and worst of all, it loses all input events when it stutters.

The Wayland protocol is too bare bones to be subject to vendor lock-in. There's nothing being assimilated besides the session manager + window manager, which all Wayland compositors are required to do anyway.

[JackLilhammers]

The real issue here, as I see it, is that the #1 open source corporation are acting like a typical closed source corporation and instead of helping the open source community thrive on equal terms, they have decided to assimilate the #1 component of the Linux ecosystem (besides the kernel itself) into their own product, and then use that product in order to gain an edge over their competition and secure themselves a larger piece of the Linux pie, at the cost of everybody else who's now stuck reinventing the wheel.

This.

[Nocifer]

Apparently not everyone cares because it look Wayland 12 years to take off and it's still relatively nowhere.

You clearly did not get to the point of my original question in some previous post.

I was asking what others think about the idea that perhaps it would have been better to evolve X11 libraries in backward incompatible manner. Microsoft did that between WDM -> WDF. WDM and WDF is not a set of protocols everyone has to implement on their own.
It's set of ready tools and libraries. That's what traditional X11 was.

How is Wayland approach fundamentally better than "X11" approach? Releasing protocol is not always the best approach because anyone can abuse protocol, make own assumptions about it, implement it partially. You can say "then use good compositor". And I could say "it's broken by design if room for abuses was made in first place".

The reference Wayland compositor already exists, and it's the GNOME compositor. I'm not saying this because I'm a GNOME fanboy (far from it) but because it's evident that most/many former Xorg key developers have moved on to work directly for RedHat. So it's actually even more like what you're describing: instead of having a reference Xorg server that is developed by a mostly independent team and can be used by everyone equally, we have a reference Wayland server that a single company (mostly) controls and tailors for use within its own product line (GNOME), and everybody else has to either conform to that single company's way of thinking (and its internal hierarchy) or go their own way and reinvent the wheel by developing their own server using the publicly released spec (which is practically mandatory with Wayland, because now the server is tightly integrated with the window manager).

To be frank though, the situation was already a bit more complex: even though it's supposedly an independent tech, Xorg was already being funded by Red Hat and co since forever, so it makes absolute (financial at least, if not ethical) sense that when it came to its replacement they decided to officially hire the developers and officially assume control of the development.

The real issue here, as I see it, is that the #1 open source corporation are acting like a typical closed source corporation and instead of helping the open source community thrive on equal terms, they have decided to assimilate the #1 component of the Linux ecosystem (besides the kernel itself) into their own product, and then use that product in order to gain an edge over their competition and secure themselves a larger piece of the Linux pie, at the cost of everybody else who's now stuck reinventing the wheel.

[reavertm]

Apparently not everyone cares because it look Wayland 12 years to take off and it's still relatively nowhere.

You clearly did not get to the point of my original question in some previous post.

I was asking what others think about the idea that perhaps it would have been better to evolve X11 libraries in backward incompatible manner. Microsoft did that between WDM -> WDF. WDM and WDF is not a set of protocols everyone has to implement on their own.
It's set of ready tools and libraries. That's what traditional X11 was.

How is Wayland approach fundamentally better than "X11" approach? Releasing protocol is not always the best approach because anyone can abuse protocol, make own assumptions about it, implement it partially. You can say "then use good compositor". And I could say "it's broken by design if room for abuses was made in first place".

[cynical]

Except without full/best/reference/production implementation. And last time I checked it was primary problem of Wayland - everyone needs to implement its protocols. Reference implementation is just for demonstration purpose apparently, a proof of concept. You did not address that part so my question remains open.

Wayland is the protocol. It’s like HTTP. The compositors are like the browsers using the protocol to communicate. Does it make sense to complain about everyone implementing their own browser? They all have different needs and differing features. As long as they all adhere to the protocol and enable people to get stuff done, who cares?