Will you switch to Wayland?
 

Has anyone tried building Wayland for Slackware? I'd love to try it, but setting it up seems like a lot of work. If you tried replacing X with Wayland, I'd love to hear about your mileage :)

I'm curious, I've watched a youtube video of the main developer talking about Wayland and X and I may give it a try.

I will switch if and when Pat decides to ship it in a new Slackware edition.

I'm missing the "I don't care" option ;)

so much of the stuff I use is X-only that I really don't even want to think about it.

I have a feeling that this is a great idea, and if it is implemented well, it's going to make Unix/Linux far more stable and better performing than it already is. No point in being a Luddite just because you can.

What he said. ^^^^^^^^^^ :)

I am not switching any time soon, or perhaps ever.

They have to first convince me that there is a benefit to switching, and that all programs I run are supported.

Whell, from what I have understood and I may be wrong there will be some kind of emulation that will extend the support to those programs that doesn't natively support Wayland but only X

Some of my friends are very, *very* enthusiastic about Wayland, and it sounds like the problems Wayland solves (mostly having to do with minor defects in high-performance video playback) are important to them.

They're not important to me. X11 works great for everything I need it to do, and giving up X11's network transparency for the RDP-like capability Wayland may or may not implement someday is not progress.

That having been said, aside from my baseline mistrust of any new code (new code is buggy code!) and reluctance to lose network transparency, I don't have anything *against* Wayland. If a time comes when it is robust, works well with fvwm, and allows me to use firefox, xfig, etc from a different computer without RDP-like loss of performance, I'll use it just as readily as X11.

To me, "Wayland" means "heroin-addicted singer of Stone Temple Pilots." That's "Weiland," really, but Wayland needs to be used by something that I use before it has more of a connotation than that. I'll wait for Pat to present it to me, and I'm in no hurry whatsoever. Either this solution or a pure X11 solution will find a way to stink. Heck, by the time it's in wide deployment, Wayland probably won't work on my video card for lack of 4D time-travel compositing.

this aint the most improving thing in wayland.
The big difference between wayland and X is the rendering of the GPU.
where X needs seperate software, like Mutter for gnome, probably something for KDE aswell.
Wayland is directly doing this, meaning more smoother rendering.

This is also the reason why KDE, Enlightment, MATE, gnome are starting to port to wayland.

I can see how "smoother rendering" could be important to people who demand flawless video playback, and I wish them luck.

I'm just not one of those people. Computers, for me, are for doing other things. So I'm likely to hold on to X11 for a while.

Thirded.

1. Its taken me decades to figure out X
2. X works.
3. X with Nvidia blob works great.
4. Wayland can suck my left testicle.

Well, maybe that last one is a bit over the top. I just don't see the supposed improvements from the hype. You are still going to need some sort of GPU hardware driver for optimal performance. You will still need all sorts of added software. You will still need X. Given the lots of Wayland devs are also X.org devs, I suspect any Wayland improvements will get ported to X.

what if wayland falls flat and mir shines, what then?

We'll wait until it's not "what if".

This presentation on Wayland and X is pretty good imo. I think it sounds like Wayland is a good idea and the core design of X is outdated and hard to work with.

I, personally don't understand the situation enough to really care. X11 works for me at the moment, but it sounds like Wayland is the future. If Wayland becomes stable enough to be included in Slackware, I'll start using it. I'll trust Pat's discretion here.

Or what if both fall flat as experimental software pushed by Canonical, and Xorg shines as it always has.

Ill second that.

If it's really better than Xorg. I voted for third variant.

Fourthed.
What he said, as well.

Not before it is included by the dev team

I'm certainly intrigued by wayland...

Wayland?
 

Seems to me we are going a sort of systemd route.

Xorg does for me and if it ain't broke etc.

I will definitely give it a try once there are other WMs/DEs than only Enlightenment and Weston that can run on it natively.

Some clarifications:
Due to the design of Wayland, running X applications (read: applications that either speak directly with X or use an older, not ported toolkit) in a Wayland compositor should be faster than running them on Xorg. First tests with 2D X applications on Weston support that.

Benefits from Wayland, at least the ones I know:
- Every rendered frame is perfect, no tearing in videos.
- Faster by design
- Security issues of X that can't be resolved due to design are fixed in Wayland.

Regarding your programs, any application that is directly dependent on a Xserver or uses a toolkit that does not support Wayland (for example fltk, Qt<5.2, GTK<3.10?) will be able to run using XWayland.

Of course you will. How should changing the display server change the need for hardware drivers?
Wayland is developed because the Xorg developers see problems in the X protocol and the Xorg implementation that they can't solve without breaking compatibility. Therefore they invented Wayland, backports to X won't happen.

Wayland is neither developed nor pushed by Canonical. Also, it is the Xorg developers themselves who don't think that Xorg shines, which is the reason for them to work on Wayland.
This has nothing to do with systemd at all, neither the way it is developed, nor is there any forcibly pushing into distros. Mentioning systemd in totally unrelated threads in the Slackware forum is, IMHO, only a way to derail the thread.
At least the Xorg security model is broken. According to the Xorg developers that is not the only thing that is.

Of course when or if ever it will be part of Slackware is up to PV and extended testing is necessary before even thinking about it, but to me it seems that the development model of Slackware and Wayland have at least some things in common. I would like to see it in Slackware in the future, if it works as intended.

As others mentioned above, I'll leave the decision to the people who know what they are talking about. I have way too little knowledge about Xorg/Wayland to make a truly informed choice. While I could do research and decide for myself, I don't really need to. I'm happy with the X server that Slackware ships with and if at some point Pat will make a decision to switch I'll trust his judgement. Long gone are the times when I 'had' to make any fundamental changes to linux to make it work for me (eg. audio/video/wifi, etc). Slackware works well for me as it ships.

Therein lies the key...

Well, it seems to benefit the Raspberry Pi at least
http://www.raspberrypi.org/archives/4053

If it means a smoother desktop on my PC as well and it doesn't crash all the time, sure I'll use it.

I forget where I read it but some place had a good article on wayland it might have been phoronix or toms hardware.
problem with Xorg is it has bugs some 20 years old and is dependent upon them when programmers go to do something it is not working out well. A number of the wayland programmers are former Xorg programmers.

I hope Wayland gets programmed well and catches up to what Xorg is doing though hope Wayland faster, more efficient and stable.

probably be forever for Pat to include in Slackware but thats not such a bad thing imo, Pat keeps Slackware one of the nicest distro's.

If you want to know what makes wayland different than X.
Please read this post on freedesktop.org: http://wayland.freedesktop.org/architecture.html

According to press releases of freedesktop
Wayland reached stable 8 months ago. with their version 1.0
From that point the DE's (KDE/GNOME/Enlightment) started to announce that they would port their DE to wayland.

Looking at the development cycle of the 2 bigger DE's (KDE and GNOME), the following possible assumptions can be made.

KDE (QT5 - Plasma framework 2) is ported to Wayland, with backwards compatibility of X.
so when KDE 5.0 comes out it should work on Wayland.

GNOME, 3.10 should have a working Wayland port, but X is still the primary Display Server.
with 3.12 Wayland should become the primary Display Server with fallback to XWayland for those applications that are not ported to Wayland.

Both KDE 5 and GNOME 3.12 are scheduled for 2nd quarter 2014.
So we can make a possible assumption that not before that time Slackware will have a stable release with Wayland (maybe in their -current)

The bug list is relative small compared to X. (the bugs that could not be solved in X are dealt with directly during development of Wayland)

I will consider a display server, if ports to at least two other operating systems have been demonstrated. I won't use a windowing system, which only works exclusively on Linux. What these "desktop environment" guys are doing on their playgrounds doesn't interest me, it's irrelevant for the IT industry anyway.

The main issue will be for the users that use one of the zillion available WMs (or, in a certain extent, the Gnome 2.x/KDE 3.x heirs). They won't work with Wayland, which requires to write a composer to talk to it (it goes without saying that it's far more complex than writing a WM). Plus, whereas X, Wayland does not draw, which also means you will have to bind to a toolkit to create taskbars, window decorations, menus and so.

To simplify things, it is question to implements them as plugins for the existing composers (Weston/Kwin). But all is still to do.

Wayland also does less things than X, which may mean the bugs will just be highter in the "graphic stack" (it's saner, but does not imply the global solution will be more stable). Plus, it does not currently run in the wide wild world, where crazy things that were not meant to happen tend to find imperious reasons to finally happen.

So, I don't claim here Wayland will fail nor that the raised problems won't find some elegant solutions, but it's IMHO too soon to discuss the opportunity of a switch.

For my first statement, I can only apologize and say that it was a knee-jerk reaction. I unreservedly withdraw said statement.

As for my second statement, I will do more research.

I assume applications built on XLib will not work in Wayland. That will be an obstacle for adoption. And people are very addicted to their window managers.

Still, Høgsberg sounds like a Norwegian or Dane, and it's well known that all things Scandinavian are good (except for surströmming and C++), so it's at least very likely that Wayland is vastly superior to X.

Vølund is also Norse for Wayland, and although I don't know how Wayland MA got its name, it's definitely etymologically related. From Völundskvädet:

Sounds promising. I see Debian Wheezy already has it. At the moment I don't see a reason to replace X though.

Not me. When I think of Wayland I remember getting lost buzzing around between Bedford and Framington searching in vain for a route that wasn't completely traffic clogged (Thoreau must be turning over in his grave) trying to visit an ex-gf. Bad associations all around.

Aside from that, by the time it's viable I'm hoping I've finally figured out how to get OpenBSD working on this laptop. Nothing against Slackware -- it's nice too -- but that's my home O/S, and I'll eventually return to it. I don't see OpenBSD ever supporting Wayland, particularly since one of their developers is a senior X person, on the board if I recall.

Actually, that's a good reason that OpenBSD WILL *eventually* support Wayland. Wayland is being devloped by many of the same x.org developers. As discussed earlier in the thread, the developers of x.org don't see a way to fix various security issues because of how x.org is designed, so they're designing wayland as a replacement for x.org.

Well, perhaps I should not speculate. I don't know what Matthieu Herrb's mind is, exactly, on the issue. There is this paraphrase of a small portion of his Feb 2013 fosdem talk that makes it sound like it may be a long time coming (or longer, as in infinite?) and that perhaps it's a headache for him and the very few others that do X server work in OpenBSD (anyone, anyone... Bueller?):
"Wayland will also be a mess for BSD and Solaris operating systems due to its dependence on kernel mode-setting, kernel input drivers, and Weston being designed with solely Linux in mind. " http://www.phoronix.com/scan.php?pag...tem&px=MTI5Njk

They will work with XWayland, but indeed, a better way would be to write a native plugin with the same functionality.
Again: XWayland.
It is planned as an option in Fedora 20 and as default in Fedora 21. Several phones are planned with it, using Sailfish and Tizen as OS, for example the Jolla phone. So having it running in the wild is coming in the near future.Kernel mode-setting is coming at least to the BSDs.
Sorry, don't know about that.Irrelevant. Weston is only the reference implementation of a Wayland compositor and not meant to be actually used (although it is possible).

Wayland needs to prove itself before it should be used.

I'd suggest offering it as a /testing package ONLY. After the code has been stabilized for at least a 6-12 months time frame following the initial release date for adequate testing. This way the developers have worked out the bugs, security issues, etc.

I don't see Wayland taking over for X.Org for some time until all the UNICES out there can get a full port of Wayland created and stabilized. With Weston being possibly Linux only, chances are X.Org could be still co-developed alongside Wayland until Weston gets a successful port, clone, or ability to be commented out in favor of a BSD friendly compositor that is Wayland friendly, which remains to be seen.

They won't. X applications like xterm will work, because each will run in its own Xserver where it will fit the root window. It's explained there (§VI).

A WM is an application which filters the events between the X server and the root window's children (which means all the windows the different applications create in the same X session). So, the only thing you can expect with XWayland is an Xserver running a WM with nothing to manage. The window management in Wayland can't be anything but a component of a compositor.

There are currently 2 major components in active development that is changing the whole DE's.
Wayland in combination with systemd is very powerfull for the desktop experience, and would bring linux desktops closer to apple and microsoft desktops in the way of user experience.

Both these components are still in their child shoes (wayland being more mature imo)
Both these components are developed with different parts of the system.
Kernel additions, glibc changes, dbus changes, probably alot more.

When it comes to BSD, they are just saying it doesn't work on BSD.
Instead they should be working with wayland to make this work.
Also systemd is here an option, as they have their own kernel the can work with the devs of systemd to make this work.

PS. I see systemd having only benefits in a desktop enviremont at the moment.
wayland can be used also as a remote desktop server. (No more seperate VT's for each user)

Hey, I didn't choose Slackware to have the worse Apple and Microsoft "user experience".

Does wayland need to be everywhere? At least on OpenBSD mailing lists, I've never seen any sign of anyone with plans to work on wayland support. The people doing real work there already have long TODO lists, I suspect. If the toolkits support both Wayland and X11, is there a need for BSD application porters to care about what's going on in the GNU/Linux display server world?

Well said. Whenever someone says something about wanting the M$ or OS-X gui experience, I have no idea what they are wanting, unless it is (useless to my needs) eye candy.

I second this sentiment, and then some. Every time I've had to touch a windows machine at work, or my wife's Macbook at home, it really impresses upon me that X11-based UI is more sophisticated, and more powerful.

Compare and contrast copy-and-pasting text segments from an almost-screen-sized window to another window:

Under Windows or MacOSX, the source window's text is not always selectable (also true under X11, but less frequently), the destination window is not always capable of taking the paste, the destination window has to be brought to the foreground for the paste to happen, and sent back to the background to make the source window entirely visible again, and highlighting text is insufficient to actually copy the text (a control-key sequence is necessary, or clicking on a menu).

Under X11, the destination window remains behind the source window, and it's just left-button to copy + middle-button to paste. zip zip zip, all done.

I know there are WM's which emulate the click-to-focus behavior of Windows and MacOSX, but these are broken. Do not use them.

let me explain it a little different.

you should see that wayland combines the power of both X11 and uses the DRM more.
this way Linux gaming natively will be even more powerfull.
drawings of windows/pixels will go smoother and faster.

thats what I ment with the user experience.

for systemd, its basically the session manager in systemd in combination with wayland that makes it more powerfull/user friendly.

I stand corrected.

Other than that, once again, systemd has nothing to do with Wayland (Rebecca Black OS, a testing distribution for Wayland, is based on Kubuntu and not using systemd at all), neither do Windows or OS X. Most of the Wayland developers are seasoned Xorg developers that develop Wayland to overcome the flaws and problems with X on modern computers. This has nothing to do with the "Apple and Windows user experience".

I'll probably be playing around with it and seeing if it's possible to make it work on Slack. However, for non-testing boxen, I'll be waiting until it's either included in Slackware or maintained on SlackBuilds.org.

I don't need another Microsoft Windows-Like Operating system either.

I use Linux to get as far away from Microsoft as possible. I guess I might end up using BSD to get as far away from Linux as possible eventually at this rate.

Do proper drivers even exist at this time for Wayland?

In what way is moving to Wayland moving towards a Windows like operating system? New things are not necessarily bad things.

The feeling around these parts is usually 'if it isn't broke, don't fix it". Well, many of the developers of X seem to think X is so broken it is beyond repair, leading them to create Wayland.

Support for wayland in 'stable' distros is still far away, but I think it'll turn out to be a good thing in the end if it creates a more sane, better performing display server.