When will Wayland be ready?

[Edward78]

https://linuxiac.com/xorg-x11-wayland-linux-display-servers-and-protocols-explained/, Wayland sound better. it is still unstable though.

[manikyath]

i'd have to go delve into the archives of bryan lunduke, but he was already joking about how wayland was taking forever like a decade ago.

[Needfuldoer]

Just like all Linux-exclusive desktop software, it will be ready "soon".

 

Flawed legacy software has inertia on its side; why roll the dice on something new when the devil you know is "good enough"?

[10leej]

1 hour ago, Edward78 said:

https://linuxiac.com/xorg-x11-wayland-linux-display-servers-and-protocols-explained/, Wayland sound better. it is still unstable though.

Wayland is ready once you determine it's ready or Xorg finally dies the death it's been due for 10 years.

 

I've personally been using wayland exclusively for two years now.

[Alexeygridnev1993]

If ain't broken, don't fix it. Yes, I know, X11 is older than me (and I'm 30), it's unmaintained and umnaintainable, it has too much code which is never used, it apparently has security issues... but, in reality, none of this matters. What matters is that Xorg is "good enough" for what it is meant to do, and that the benefits of Wayland over Xorg are not enough to force the developers to spend time and money to adapt their software to Wayland

[Herr.Hoefkens]

is there ever any project that is considered 'ready' these days?

i cant think of any, not even X11 wich saw its first release in 1984 is 'ready'

but technically wayland is at version 1.21.92, so and >1 is considered not in beta any more...

 

 

[maplepants]

I think the real answer is never. Wayland made the same mistake as btrfs. It went live when it was "good enough", except it wasn't really good enough and earned itself a pretty bad reputation. That makes it hard to attract both users and developers, making it hard to improve. Projects like this just kind of exist half complete until they're replaced. With zfs doing everything btrfs does, but better, the only thing keeping btrfs around at all is the zfs licensing drama (because oracle sucks), and while it's enough to keep btrfs from going away, it's not enough to attract the critical mass of devs and users you'd need to make it good.

 

Wayland has a similar problem. It can attract enough devs and users to run a basic desktop in GNOME (but not KDE). That's not enough to make it truly good. Absent some big changes in the Linux desktop world, this is unlikely to change.

[Herr.Hoefkens]

2 hours ago, maplepants said:

I think the real answer is never. Wayland made the same mistake as btrfs. It went live when it was "good enough", except it wasn't really good enough and earned itself a pretty bad reputation. That makes it hard to attract both users and developers, making it hard to improve. Projects like this just kind of exist half complete until they're replaced. With zfs doing everything btrfs does, but better, the only thing keeping btrfs around at all is the zfs licensing drama (because oracle sucks), and while it's enough to keep btrfs from going away, it's not enough to attract the critical mass of devs and users you'd need to make it good.

 

Wayland has a similar problem. It can attract enough devs and users to run a basic desktop in GNOME (but not KDE). That's not enough to make it truly good. Absent some big changes in the Linux desktop world, this is unlikely to change.

while one could say that wayland is 'favoring' gnome , certainly in certain departments , i do think there is allot of effort being done on the kde side of things , possibly even more then on the gnome side of things . https://community.kde.org/Plasma/Wayland_Showstoppers

https://community.kde.org/KWin/Wayland

 

and that is while this is possibly also valid : https://news.ycombinator.com/item?id=32295685

or

 

[maplepants]

11 minutes ago, Herr.Hoefkens said:

i do think there is allot of effort being done on the kde side of things , possibly even more then on the gnome side of things

I'm not saying that nobody is working on it, just that there's not enough being put into wayland for it to ever be good.

Look at wayland's contribution rate vs something really popular like systemd. 

 

It kind of doesn't matter how/why wayland ended up like this but this. But it's like btrfs, the ship has kind of sailed on it being good. Not many projects come back from this vicious cycle.

[Herr.Hoefkens]

6 hours ago, maplepants said:

something really popular like systemd.

this made me laugh a little , because if the wayland team would have to deal with the same amount of sh*t , that went  Lennard Poettring's direction , they might wanna seek out some shelter for a while , because its comming in cubic tonnes pretty soon then.

one big difference between wayland and systemd , is that systemd is the init system, (every system needs one), that takes care of what used to be a gazillion of startup scripts, one or more  for each service , and  didnt even take care of restarting them when they crashed and so forth , so in serverland while it took a long time before accepted , is kind of verry usefull, wayland in contrast is a Display Manager/Compositor protocol, and not even a 'server' in the way X11 is a server , so it really is only usefull for local use and on top of that limited to systems that require a Gui Interface (no embedded, no iot , no servers,.. leaving just desktops and workstations.

 

ps , on the kde side of things there is : kwayland , kwin, and kwinft . wich is more then what can be said about gnome , who apparently have no need to extend wayland what so ever , as if it was taylor made for them  

 

also not defending wayland here , while i can switch to it (have everything compiled with both X and Wayland support ), when using wayland most electron apps simply refuse do display anything, and mouseclicks + where menu's appear is a bit at random... if they open at all at times. and besides i require some x trickery for my dual monitor setup (both screens are equal in size physically, but one is 1920x1080 max and the other is 1650*1050 max , so  have to fake the resolution on the second monitor to be also 1920x1080.

wich as far as i know is still not possible in wayland

[maplepants]

9 hours ago, Herr.Hoefkens said:

one big difference between wayland and systemd , is that systemd is the init system, (every system needs one), that takes care of what used to be a gazillion of startup scripts, one or more  for each service , and  didnt even take care of restarting them when they crashed and so forth , so in serverland while it took a long time before accepted , is kind of verry usefull, wayland in contrast is a Display Manager/Compositor protocol, and not even a 'server' in the way X11 is a server , so it really is only usefull for local use and on top of that limited to systems that require a Gui Interface (no embedded, no iot , no servers,.. leaving just desktops and workstations

That's a fair point. Desktop linux is a very small part of the Linux world so comparing desktop to non-desktop software puts desktop stuff at a disadvantage.

 

If you prefer you can compare wayland to KDE Connect an unambiguous desktop LInux hit. The result is the same, wayland has no hope of ever being good. 

 

I think the other questions about wayland don't really matter. Does it favour GNOME? Are KDE devs working hard enough on wayland? Is a compositor model even a good idea? The answer for all of them is the same; who cares. These questions matter for X12 or whatever succeeds both it and X11, but because wayland itself will never really get much better the wayland answers don't really matter.

[Herr.Hoefkens]

19 hours ago, maplepants said:

That's a fair point. Desktop linux is a very small part of the Linux world so comparing desktop to non-desktop software puts desktop stuff at a disadvantage.

 

If you prefer you can compare wayland to KDE Connect an unambiguous desktop LInux hit. The result is the same, wayland has no hope of ever being good. 

 

I think the other questions about wayland don't really matter. Does it favour GNOME? Are KDE devs working hard enough on wayland? Is a compositor model even a good idea? The answer for all of them is the same; who cares. These questions matter for X12 or whatever succeeds both it and X11, but because wayland itself will never really get much better the wayland answers don't really matter.

i can see your point.

on the other hand , linux desktop has been stuck with X11 for a verry long time , and while configuring that piece of acient tech , has become a breeze (compared to what it was , i remember spending hours trying to get the nvidia driver to work back when Xorg was still XFree86, and the driver would work fine , just X would not use it , and if it did it had no screens , and if not that the keyboard was messed up, and if not that a 3button mouse with scroll suddenly was a two button mouse ,.... and so on.

 

it must be said that for the specific purpose of a desktop graphical shell, X11 is allot of unnessesary detouring. if im not mistaken its pretty well documented by this (even more senior then i am by a mile or two ) 'youtuber' (youtuber fees kind of wrong to categorize him as) in this video :

 

[Herr.Hoefkens]

COINCIDENTALLY : JUST RELEASED