WayVNC 0.7 Released As VNC Server For Wlroots-Based Wayland Compositors

[You-]

I'm really interested in solution like terminal server that allows users to have access to server and be able to run their own graphical sessions over the net.
Something like Citrix but open and supporting Wayland.
Any ideas?

Gnome with the WIP patches, or out of the boex by gnome 46.

Those include remote login. if you can live without that step, its ready now and has been for a few years - the preferred backend is RDP.

[avis]

long quote

Gnome being the only feature complete and relatively bug free Wayland implementation 15 years after the protocol was devised is the best confirmation that this issue is trash. You're right, oiaohm, you've got solid arguments, Everyone else who needs/wants to use Wayland may get wrecked, I hear you. Have a nice day.

[oiaohm]

Gnome being the only feature complete and relatively bug free Wayland implementation 15 years after the protocol was devised is the best confirmation that this issue is trash. You're right, oiaohm, you've got solid arguments, Everyone else who needs/wants to use Wayland may get wrecked, I hear you. Have a nice day.

KDE is the most complete implementation followed by Sway after Weston. Gnome is tail of the hunt.

Trash issues/bug reports do nobody any good. Quoting them over and over again is waste of everyone time avis.

https://gitlab.freedesktop.org/wayland/wayland-protocols/-/merge_requests/247

No please go and read 247. I am not against what everyone else needs/wants.

https://gitlab.freedesktop.org/wayland/wayland-protocols/-/issues/72

These are not trash.

Linking to 233 is pointless avis because like it or not

I wonder if Wayland developers could maybe consider writing a reference Wayland server which provided all these features and APIs and only required WMs and DEs to provide very simple window managers which drew window decorations.

The key-point of 233 was always implemented in reference Wayland server. So 233 is trash. Person successfully asked for nothing. If you are that foolish person avis you need be told that 233 is trash. The person who wrote 233 screwed up so badly its not even worth debating.

[avis]

long quote

KDE is trash (in your words) under Wayland: https://community.kde.org/Plasma/Wayland_Showstoppers

The rest of what you mentioned is not even a DE. Continue to be excited about Wayland only please don't make stuff up and don't make people who quote the best software industry practices look like idiots because it may backfire. Windows, MacOS, iOS and Android all have the only fucking display server. Wayland mandates multiple implementations which is just beyond asinine. With that, I've left this discussion because I'm 1000% sure you'll find excuses why it's OK. It is not and never will be.

This news piece alone is a perfect confirmation/proof why multiple display servers idea is completely broken and backwards yet you're arguing it's all fine. Yeah, just put out the fire.

[browseria]

For those who may be swayed into believing avis's arguments, I invite you to read Daniel Stone's last comment in 233, it explains it as succinctly as anything. If you didn't know, Daniel Stone is a MAJOR contributor to Wayland, Weston, Vulkan, OpenGL, OpenGL ES, EGL, DRM, KMS, graphics drivers, X11, X.Org, Linux kernel, GStreamer, Clutter, GTK+, WebKit for over a decade. Please consider the source.

Also, please read Window positions under Wayland, it is only 500 words, includes animated pictures(!) and will take you less than 5 minutes to digest. You will be more informed on the topic after. Note: this was written by the MIR developers, who have embraced Wayland and want to take it further in the future. Remember the Unix philosophy: do one thing and do it well. That is the essence of what Wayland and Pipewire and Portals adheres to. This is a different design than X11 on purpose - to not repeat the same mistakes of the past, and to provide a platform for growth and extensibility far into the future.

[oiaohm]

Wayland mandates multiple implementations which is just beyond asinine..

avis please show when in the protocol Wayland goes and mandates multiple implementations. In fact it does not.

X11 development early on goes the same way Wayland has.

Android has 2 different display servers the one they use for chromeos and the one they use in phones. "only fucking display server." so this is not right for all the OS you quoted.

Do remember MacOS and iOS use the same kernel but they use different "display servers". So andriod has 2. Apple really has 2. Windows is the only one with 1 out your list.

Fragmenting at the display server is horrible common.

[avis]

For those who may be swayed into believing avis's arguments, I invite you to read Daniel Stone's last comment in 233, it explains it as succinctly as anything. If you didn't know, Daniel Stone is a MAJOR contributor to Wayland, Weston, Vulkan, OpenGL, OpenGL ES, EGL, DRM, KMS, graphics drivers, X11, X.Org, Linux kernel, GStreamer, Clutter, GTK+, WebKit for over a decade. Please consider the source.

He does not dismiss the issue. He says Wayland developers will not fix it because this is how Wayland is meant to be. Too bad not a single successful OS has been designed this way so far but let's wait ten more years, maybe then it's gonna be great and I'll be able to use native Wayland enabled XFCE or even IceWM or JWM. Too bad the last two have vocally rejected Wayland ports because of this very "non-issue".

Also, please read Window positions under Wayland, it is only 500 words, includes animated pictures(!) and will take you less than 5 minutes to digest. You will be more informed on the topic after. Note: this was written by the MIR developers, who have embraced Wayland and want to take it further in the future. Remember the Unix philosophy: do one thing and do it well. That is the essence of what Wayland and Pipewire and Portals adheres to. This is a different design than X11 on purpose - to not repeat the same mistakes of the past, and to provide a platform for growth and extensibility far into the future.

Wayland so far has provided a perfect platform for wasting thousands of hours of very scarce talented Linux developer resources as evidenced by Mutter, two KWin Wayland designs, XFCE (none yet), WayFire and many other implementations. Very few people out there can hack display servers with so many required features.

I won't read this Mir related document, not interested in at all. Instead please read this which is coming from a Wayland developer. Some people here believe I'm making an argument against Wayland out of thin air. You could read what an expert has in mind.

[fitzie]

avis please show when in the protocol Wayland goes and mandates multiple implementations. In fact it does not.

X11 development early on goes the same way Wayland has.

Android has 2 different display servers the one they use for chromeos and the one they use in phones. "only fucking display server." so this is not right for all the OS you quoted.

Do remember MacOS and iOS use the same kernel but they use different "display servers". So andriod has 2. Apple really has 2. Windows is the only one with 1 out your list.

Fragmenting at the display server is horrible common.

This is just cope. It's fine to argue that the linux desktop cannot be everything to everyone, limitations of opensource, etc etc, but what wayland, gtk+, gnome has done is nothing but stagnation for well over a decade. The entire standardization of wayland protocols was broken by the gnome cabal that wants to limit freedom for their purist vision. There is massive code duplication between the major compositors. And basic desktop features that are commonly used on macos, windows are just not available to linux users. The community has done some good work with wlroots, and that's starting to break some of the logjam, but it's really a toy for nitch tiling compositors at the moment.

The oddest part of all this cope is that it does GNOME the biggest disservice. If you want to build a simple thing like a dock for gnome, you _must_ use gjs, with an api that constantly changes, totally deficient documentation and capabilities (cannot even use GTK), and you cannot even isolate your dock in a separate process to protect the "compositor" from memory leaks, runaway processing, etc. All this so they don't have to create a stable protocol for the limited interactions a dock actually needs? It is a total lose lose, and it is a now approaching an entire lost generation of desktop developers. GNOME cannot even wake themselves up to all the lost talent, and architectural dead weight they have built up because they are determined to lock down customization. They've done a series of work to split out certain work to separate threads but it's not enough to keep GNOME alive. On the low end they cannot compete with phosh/phoc, and soon enough cosmic will eat their lunch on the high end. They cannot even do something basic and drop Xorg support (not to be confused with Xwayland/X11), which would be an obvious step to take if they want to stay relevant, that , and of course adding SSD, which we just accept that they will never admit they were wrong and still are wrong about.

It seems the only thing that will keep GNOME relevant is xdg-portals, which is a goofy little thing that may end up mattering just because of it's security properties but it will not keep users on GNOME at all.

[oiaohm]

The entire standardization of wayland protocols was broken by the gnome cabal that wants to limit freedom for their purist vision.

I will stop you there. You need to look at the protocols that have got into Wayland. Gnome party fairly much never vetoed wayland protocol change. Including when KDE wanted to add server side decorations. Biggest disruption to wayland protocol development is Nvidia wanting to go the eglstreams route so making it basically impossible to make prototypes of many things. This in fact dead end many wayland debates before 2020.

There is massive code duplication between the major compositors.

There was massive duplication before Wayland existed. So this is not a new thing and has existed as long as DE have been using compositors. This is a surprising thing if you count lines of active code(as in code in fact built into binaries) kwin wayland has less code than kwin x11. There is a lot of code that only exists in both KDE and Gnome and every other X11 compositors that only there to deal with x.org X11 server quirks.

And basic desktop features that are commonly used on macos, windows are just not available to linux users.

This has been true under X11.

There is also a problem what you call basic desktop features does not mean they are right in the Linux environments.

https://gitlab.freedesktop.org/wayland/wayland-protocols/-/merge_requests/247

This work here for absolute positioning under wayland is worthwhile read. Learn very quickly that it takes little more work than saying Windows and macos and X11 has this feature so should wayland. There are a stack of outstanding cases with absolute position and X11 screwing up this is why the promised absolute window positioning for wayland in fact requires to have the feature that Wayland compositor just straight up ignore absolute position request if it so pleases. Linux world has such things as tiling windows managers.

Of course Wayland developers are not wanting to add features to the Wayland protocol if there is not a valid use case. This is a once bitten insanely shy problem with wayland as well.

https://en.wikipedia.org/wiki/Hyper_Text_Coffee_Pot_Control_Protocol

There are things that were added to the X11 protocol without a valid use case that make the hyper text coffee pot control protocol look sane..

Remember macos and windows have limited windows management designs and different security systems to Linux. The difference means where X feature may correctly own may be complete different. Yes x feature like absolute positioning may require unique Linux world limitations so that everything works close to right.

We do know at this point 90%+ of applications don't need absolute position to function.

xdg-portals is not a gnome project. xdg-portals is a flatpak project. It common for every other party to have implemented a new xdg-portals feature before gnome has as well.

Wayland protocol its a newish protocol and its a long process to go though looking at real world example applications working out what they really need and what limitations should be there. Microsoft and Apple have not gone back to scratch on there protocol completely ever.

One hard reality is like it or not the dispute between Gnome and KDE started in the 1990s. Getting those two on the same page to share the same compositor is not going to happen any time soon even if Wayland had never happened. Wayland developers had to pick there battles. Getting in the middle of a Gnome vs KDE battle not worth the headache or the delay that would cause. Just let them code their own compositors as long as they like.

Why does wayland have so many compositors is basically the same question as why did Linux have so many windows managers, distributions , X11 compositors... with the same answer because people could.

[browseria]

He does not dismiss the issue. He says Wayland developers will not fix it because this is how Wayland is meant to be...

You are correct, this is the way Wayland was meant to be, but I think you missed this part:

It's not that we've never thought about doing what you're demanding, but that we have thought about it and reached a different conclusion to you.

You are also incorrect, in that he _does_ dismiss the issue; he even explains _why_ he arrived at a different conclusion than you and leaves breadcrumbs to the source. However just like the Tanenbaum v Torvalds debate eventually time will tell which one is correct pretty definitively, that time is just not now.

The same as the human body is constantly refreshing itself, software that is part of a dynamic system requires constant maintenance or "refreshment". If no one is willing to maintain a particular technology/workflow, what choice is there in the matter? One could argue with the rate the human body replaces its cells too, but it is unlikely to have much effect.

...I won't read this Mir related document, not interested in at all....

This may be the source of the misapprehension...never reading anything that challenges one's assumptions will make it very hard to recognize what necessitates change in the first place!

...You could read what an expert has in mind.

I have read what "this expert" wrote, but I have also come to a different conclusion than him. So did this guy (developer of Budgie Desktop). Vive la dif·fé·rence!