Arcan 0.6.2 Display Server Continues Working On Network Transparency

[oiaohm]

Everything is half baked on Linux desktop. Linux is a server backend and embedded device environment system. The last time it wasn't the case (excluding simple things like fluxbox, or xfce) was when everything was way too simple to be broken like in kde3, and the end of the gnome 2.32 cycle.

No take the rose colour glasses off. Gnome 2 and Kde 3 you there were still issues with gnome applications inside KDE and KDE application inside gnome having issues.

KDE 2 and Gnome3 is the time frame we move from X11 desktops having there own unique sound servers as in Esound for gnome and arts for kde to pulseaudio. Yes it KDE 4 and Gnome 3 where both of them agree that the solution forwards is pulseaudio and to kill off their own unique sound servers.

Dbus is 2006 when gnome and kde agree to kill off their own unique desktop IPC solutions. Remember KDE 3 first released in 2002 and last KDE release is 2008 and Gnome 2 is 2002 to 2010.

Gnome 2 and KDE 3 is when we see X11 desktop environments starting to do common API/ABI stuff. Lot of stuff that should have common API/ABI does not at this stage.

Please note gnome 1 and kde 2 time frame there is a survival of the fittest logic that neither had to work with each other at all. Yes that survival of the fittest logic is how we end up with XPrint as in the X11 server being a print server when we already had highly functional CUPS.

Lot of the issues with Wayland is items that should have always had a proper common API/ABI between desktop environments don't. Items like how to do global short cuts with single ABI need to be solved for X11, Wayland and Arcan. There is a long list of items that should have a single API/ABI for application developers to use that don't and this is not a new problem. The lack of single API/ABI for these features go back into the 1970s yes before Linux.

[MrCooper]

I wish it improves fast and challenge Wayland position as soon as possible. Wayland went for as low as possible "path of resistance", and the consequence of it is that, almost 14 years later, it is still unfinished, lacking [...]

Right, that's the main reason, not that it's a lot of work or pesky details like that.

If somebody had the power to make everybody else drop Wayland and work on Arcan instead, it would take another decade for Arcan to reach the maturity Wayland has today across the ecosystem.

I think we should set the bar a little bit higher and try to get it to challenge X11's position!

We need do a modern alternative to X11 and there simply isn't one in sight yet.

The trends seem clear, Wayland will succeed X as the most used primary display protocol on Linux sooner or later. It just takes time.

[timofonic]

Right, that's the main reason, not that it's a lot of work or pesky details like that.

If somebody had the power to make everybody else drop Wayland and work on Arcan instead, it would take another decade for Arcan to reach the maturity Wayland has today across the ecosystem.



The trends seem clear, Wayland will succeed X as the most used primary display protocol on Linux sooner or later. It just takes time.

I agree that X sucks and needs a proper replacement. That replacement needs to greatly surpass it in terms of security, efficiency, robustness, scalability, iterative Vulkan-like improvements and capabilities.

But Wayland is taking even a lot time to only have the outdated features of X. The things gets even worse when it's about innovative and powerful features. Something is wrong with Wayland development and design, in my humble opinion. A protocol with dozens of implementations and just some humble community done code sharing might be problematic too, wlroots to the rescue and I think mostly all projects should adopt it (including KDE and GNOME, of course).

Excuse me. Arcan is still experimental, but the results are amazing for their very limited manpower compared to Wayland and X.

I don't want a PulseAudio-like nightmare with Wayland, please not.