I want to make my own dynamic tiler for wayland written in rust. Yeah yeah, meme meme meme. fast blah blah. It's the hardbooks I own and have interest in. I think it could be a fun challenging project.
I've been a developer for over 25 years, and unfortunately seeing a developer with this type of utter mental confusion is too common. Various Errors/misinformation/lies : 1) Term "window", does *NOT* come from microsoft. The terminology and concept of "window" was first used in Douglas Engelbarts NLS graphical system. Then later Xerox PARC also used the term "window" in their software. Later Digital Research's GEM desktop and Maconitosh also used the term "window" in it's software. MIT's X server used this term in 1984 before Micosoft Windows ever existed. 2) Wayland's "surface" is just a literal renaming of the common GUI term "window" that has been in use since the 1960's! , Again *NOT* from microsoft. Macintosh's Carbon API did this same renaming in the late 90's changing GUI object "windows" to "Vistas" 3) Programming a client or window manager in Wayland requires up to 4X the code than the same application in X11. And don't argue unless you are programmer and have done actual code. Because I have the code to prove it.
Are there only four layers? FVWM has more layers so I wonder how that would work out. Also it kinda sounds like they could have suggested additional atoms which would allow some better distinction if applications started using them.
From the spec: »_NET_WM_WINDOW_TYPE This SHOULD be set by the Client before mapping to a list of atoms indicating the functional type of the window. This property SHOULD be used by the window manager in determining the decoration, stacking position and other behavior of the window. The Client SHOULD specify window types in order of preference (the first being most preferable) but MUST include at least one of the basic window type atoms from the list below. This is to allow for extension of the list of types whilst providing default behavior for Window Managers that do not recognize the extensions.« Literally name it _NET_WM_WINDOW_TYPE_LAYERSURFACE
something that i really don't know why it was not implemented. Why making "window systems" if you can make terminals with graphics capabilities. Create a terminal, set the controller file descriptor to 3 or something different and then you can create programs that can control the graphics stuff of the terminal. And then you can easily create a window manager that can create subterminals also with graphics capabilities, and so no more you have terminal/gui duality. Plan 9 does something similar, but uses actual files instead of fds.
While there are protocols for smuggling images and videos over a terminal link, it's really not pleasant to have to do. Having a protocol designed from the ground up for graphics rendering is a lot cleaner and lowers the barrier to using graphical features.
Tons of things would have to be designed from the ground up. in my mind a perfect graphical system would be one reliant on commands from the ground up, you don't get an empty desktop for files, you get a terminal with a background image if you want. emacs kinda tries to replicate this behavior inside its own environment, so even if it's limited, its proof that you can have a reliant subsystem thanks to how many graphical capabilities the program has. On the other hand you have regular terminals who try to be minimalist, but people want to force its capabilities to do things they're not meant to do, like image display, so people end up building up their entire window manager around working out terminals, which is minimalist and customizable, but also a hassle at times when it doesn't have to be.
Wayland is unbelievably bad. Its mind boggling that it is recommended so much lately. OBS is practically unusable. Remote Desktop is ridiculous. Extreme lag spikes literally every minute. Its actually just so insane to me that so many people dunk on Xorg. Also its really funny to me that this guy said "Xorg is really bad. Its like Windows." Oh, Windows, the desktop OS that actually works? How terrible it must be to be like that. I really tried to give Wayland a chance for the past few weeks and it is just not there. Im not expecting development or bug fixes on Xorg, Im not "entitled," but if Wayland was my only option I would not be a Linux user at all, which sucks because there are so many things I love about Linux. Its unfortunate that such a major, important part of an Operating system has to be the trendy new debate. I hope they figure this stuff out, but from what Ive seen, they dont seem interested in making Wayland as robust as it needs to be, due to some philosphical reason or whatever it is.
I think he means the protocol schema was arbitrarily laid out according to what made sense at the time, resulting in technical debt attributed to people having no clear indication as to how to best implement features of their applications.
I want to make my own dynamic tiler for wayland written in rust. Yeah yeah, meme meme meme. fast blah blah. It's the hardbooks I own and have interest in. I think it could be a fun challenging project.
the waycooler project tried to do that but realized that rust doesn't work well with wayland design and remade it in C
@@notuxnobux Not true, they just messed up the bindings for libwayland so they scrapped it
@@notuxnobuxand then abandoned it
as for a rust written tiler wayland compositor, you can check Jay by mahkoh
@@socvirnylestela5878
Thank you, I'm checking this out for sure.
This explanation of Wayland layers, and how Wayland at its core works is so easy, accessible, and well explained. Thanks so much Vaxry!
I've been a developer for over 25 years, and unfortunately seeing a developer with this type of utter mental confusion is too common.
Various Errors/misinformation/lies :
1) Term "window", does *NOT* come from microsoft. The terminology and concept of "window" was first used in Douglas Engelbarts NLS graphical system. Then later Xerox PARC also used the term "window" in their software. Later Digital Research's GEM desktop and Maconitosh also used the term "window" in it's software. MIT's X server used this term in 1984 before Micosoft Windows ever existed.
2) Wayland's "surface" is just a literal renaming of the common GUI term "window" that has been in use since the 1960's! , Again *NOT* from microsoft. Macintosh's Carbon API did this same renaming in the late 90's changing GUI object "windows" to "Vistas"
3) Programming a client or window manager in Wayland requires up to 4X the code than the same application in X11. And don't argue unless you are programmer and have done actual code. Because I have the code to prove it.
😮😮😮😮 I too have faced confusion 🙀😔
I'm aware of the KWinFT project - drop-in replacement for KWin, but based on wlroots.
Not aware of any plans to shift KWin itself to wlroots.
mate …. are we all gonna be making this massive switch for the sake of standarization ..
Is this shill paid to talk shit about X11? The atoms are standardized by freedesktop.
He definitely has more experience then you to shit talk about xorg, as he actually built a functioning wm for it
Are there only four layers? FVWM has more layers so I wonder how that would work out.
Also it kinda sounds like they could have suggested additional atoms which would allow some better distinction if applications started using them.
From the spec:
»_NET_WM_WINDOW_TYPE
This SHOULD be set by the Client before mapping to a list of atoms indicating the functional type of the window. This property SHOULD be used by the window manager in determining the decoration, stacking position and other behavior of the window. The Client SHOULD specify window types in order of preference (the first being most preferable) but MUST include at least one of the basic window type atoms from the list below. This is to allow for extension of the list of types whilst providing default behavior for Window Managers that do not recognize the extensions.«
Literally name it _NET_WM_WINDOW_TYPE_LAYERSURFACE
Wine support yet? Gaming is what is driving desktop Linux. No wine/proton native support and not playing well with Nvidia are serious issues.
It's got actual structure!
something that i really don't know why it was not implemented.
Why making "window systems" if you can make terminals with graphics capabilities. Create a terminal, set the controller file descriptor to 3 or something different and then you can create programs that can control the graphics stuff of the terminal. And then you can easily create a window manager that can create subterminals also with graphics capabilities, and so no more you have terminal/gui duality.
Plan 9 does something similar, but uses actual files instead of fds.
While there are protocols for smuggling images and videos over a terminal link, it's really not pleasant to have to do. Having a protocol designed from the ground up for graphics rendering is a lot cleaner and lowers the barrier to using graphical features.
Tons of things would have to be designed from the ground up. in my mind a perfect graphical system would be one reliant on commands from the ground up, you don't get an empty desktop for files, you get a terminal with a background image if you want. emacs kinda tries to replicate this behavior inside its own environment, so even if it's limited, its proof that you can have a reliant subsystem thanks to how many graphical capabilities the program has. On the other hand you have regular terminals who try to be minimalist, but people want to force its capabilities to do things they're not meant to do, like image display, so people end up building up their entire window manager around working out terminals, which is minimalist and customizable, but also a hassle at times when it doesn't have to be.
I want to use Wayland. But it uses electron and wont pass through my main screen. So its xorg...
Wayland is unbelievably bad. Its mind boggling that it is recommended so much lately. OBS is practically unusable. Remote Desktop is ridiculous. Extreme lag spikes literally every minute. Its actually just so insane to me that so many people dunk on Xorg. Also its really funny to me that this guy said "Xorg is really bad. Its like Windows." Oh, Windows, the desktop OS that actually works? How terrible it must be to be like that. I really tried to give Wayland a chance for the past few weeks and it is just not there. Im not expecting development or bug fixes on Xorg, Im not "entitled," but if Wayland was my only option I would not be a Linux user at all, which sucks because there are so many things I love about Linux. Its unfortunate that such a major, important part of an Operating system has to be the trendy new debate. I hope they figure this stuff out, but from what Ive seen, they dont seem interested in making Wayland as robust as it needs to be, due to some philosphical reason or whatever it is.
Lmao just because some things doesn't work you doesn't mean it's a bad thing. Good rant anyways.
Works on my machine, so I call it user error.
@@VallThyo Yeah lol
Can’t say I’ve had those issues.
@@VallThyo Yep, pretty much
> Xorg is like windows
You mean “just works”? 😂
I think he means the protocol schema was arbitrarily laid out according to what made sense at the time, resulting in technical debt attributed to people having no clear indication as to how to best implement features of their applications.
Yeah, because windows is know to be stable, resiliant and safe! /s
this explains why waylaid is shit for NVIDIA and games.