Wayland 1.5 Released

An anonymous reader writes "Wayland 1.5 has been released, along with Weston Compositor 1.5. Wayland/Weston 1.5 carry many new user features, with a new libinput back-end, XWayland support, a full-screen shell, and many other changes. This release is particularly important as Fedora 21 will run on GNOME Wayland and X.Org Server 1.16 will be released this summer with integrated XWayland support."

Comments

Will it really go the pulseaudio way?

[Jorl17]

Wayland and its never-ending stream of stories seems to be going the way pulseaudio did. It's heavily criticized, cuts down some features or is somehow buggy, but seems to give many users what they want, or at least that's what all these crazy stories point to.

As long as I can still run X atop Wayland, I don't really care. I loved pulseaudio when it was being bashed already. Maybe I'll love Wayland too? Has anyone here actually seriously tried this thing before bashing it?

Re:Will it really go the pulseaudio way?

[morgauxo]

>> As long as I can still run X atop Wayland, I don't really care.

If I can't get the applications I want as X apps anymore because everything has moved to Wayland but Wayland still doesn't support remote display then I will care deeply.

Re:Will it really go the pulseaudio way?

[Microlith]

Wayland is critically important, which is why (unlike Pulseaudio) it hasn't already been rolled out yet. Qt has integrated it, Gnome has, KDE is porting KWin to implement it. There have been fairly few technical criticisms, the only one I've seen made with any muster has been network transparency - but even that could be solved rather easily given the way Wayland works with framebuffers.

On the flip side, Xorg has you dragging around unused cruft and the way it interfaces with the kernel forces some possible security holes be left open, holes that Wayland will fix.

Re:Will it really go the pulseaudio way?

[Microlith]

It's highly likely that Wayland's remote display will beat X. Virtually none of the features (remote drawing) that X provided over the network are used today (line/polygon drawing) and tool kits like Qt/GTK+ have you shipping framebuffers across the network, something built around manipulating frame buffers should be able to stream them over the network, individually, to a compositor on your system.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Wayland will never support remote display because that's not not it works. Someone could write a compositor that does the job, but the best anyone has come up with is VNC... ...which, IMHO, makes X11 look like a snappy protocol.

With any luck someone will write something of value, but I'm not holding my breath. I'm guessing we'll end up with Wayland -> VNC -> X11 Display showing VNC window -> NX. Fuck me. Client server will die with Wayland. :( No, it's really not for the better.

Too bad it's throwing the baby out with the bathwater. Can't seem to get any middle ground these days, what with this and systemd.

Re:Will it really go the pulseaudio way?

[Microlith]

Wayland will never support remote display because that's not not it works.

Wayland does not work over a network inherently, but there's no reason you couldn't forward the buffers over the network and have them composited remotely.

Someone could write a compositor that does the job, but the best anyone has come up with is VNC... ...which, IMHO, makes X11 look like a snappy protocol.

Except that X11 over the network with any modern toolkit is already effectively forcing X11 to do what Wayland will do - only X11 does it badly and without compression. And VNC sucks because it has to poll the whole desktop - Wayland could forward individual applications.

Re:Will it really go the pulseaudio way?

[tlhIngan]

Except that X11 over the network with any modern toolkit is already effectively forcing X11 to do what Wayland will do - only X11 does it badly and without compression

Well, you could run it over SSH with compression enabled.

Though, run something like firefox over remote X, and that's a real dog. You could pretty much download and install firefox locally before the remote X version of firefox would even give a hint that it's starting.

(I had to remote firefox because we were diagnosing an issue where some machines could access a website and others couldn't so I needed some more data).

Oh, and this was over SSH with X forwarding, no compression, GigE network all the way. The only thing was the target system was Linux, the X server was Cygwin-X on Windows.

Starting anything more complex than an xterm remotely is an exercise in patience.

Re:Will it really go the pulseaudio way?

[dbc]

And Wayland remote display is going to happen when, exactly? Is it on the roadmap? I'm asking seriously -- if there is a roadmap, point me to it, I don't follow Wayland devopment outside of the occasional rant-fests on Slashdot like we are having now.

There are certain environments where remote display is the *only* display, so if Wayland doesn't have it, Wayland doesn't go into those environments.

Re:Will it really go the pulseaudio way?

[Microlith]

Basically all that. Even over GigE simple things like gvim are a dog.

If I can stream a game from my desktop to a tablet and play it with virtually no latency, on Windows it should be possible for something implementing Wayland to stream individual application frame buffers across a network effortlessly - hell, it could do it with applications that are live on a remote screen and keep them alive if the remote server disconnects, something that always annoyed me.

Re:Will it really go the pulseaudio way?

[Microlith]

No timetable. Wayland is a protocol between applications and the compositor, so remote support depends on the compositor. It's already being tackled, so I would be surprised if it didn't happen shortly after Wayland was rolled out on desktop distros.

There are certain environments where remote display is the *only* display, so if Wayland doesn't have it, Wayland doesn't go into those environments.

Then in such a case I would say two things:

First, why are you using a GUI in such a situation?

Second, X11 is not going away immediately, and no one expects it to. Qt and GTK+ will remain compatible with X11 for some time to come precisely because of this. And you'll still be able to access those remote X applications via XWayland.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

This might be useful.

http://tech.slashdot.org/story/13/04/03/1219239/remote-desktop-backend-merged-into-wayland

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

not any time soon.
Problem is not with Wayland.
Problem is with NVIDIA binary drivers. They simply have no support of it and are rather in very initial stage of Wayland compatibility development. AMD open source drivers also suck at the moment when it comes to many aspects. So there is long road ahead.

Re:Will it really go the pulseaudio way?

[Tailhook]

RDP protocol support was merged into Wayland over a year ago. Wayland's original developer prototyped a remote display implementation almost two years go, before 1.0 was released. This is in addition to XWayland already providing an X server to host legacy X apps.

Wayland will have good remote display. The peanut gallery rant-fests around here not withstanding.

Anyhow. Now you know. If I'm wrong get a refund.

Re:Will it really go the pulseaudio way?

[dbc]

First, why are you using a GUI in such a situation?

Robots don't have displays. It's really difficult to get your work done if your monitor keeps skittering away across the lab. Visualization tools for various pieces of robot state are much better than text dumps -- not surprisingly. Display across the WiFi network is a requirement. Also, all the generic basic tools need to run in a headless environment.

But robots aren't the only embedded environment where Linux is popular. Again, with those it is nice to be able to display to a large monitor for development work, even though the device might have a small display of it's own.

Second, X11 is not going away immediately, and no one expects it to. Qt and GTK+ will remain compatible with X11 for some time to come precisely because of this. And you'll still be able to access those remote X applications via XWayland.

And that is what we will no doubt do when the time comes.

Re:Will it really go the pulseaudio way?

[bluegutang]

Anyhow. Now you know. If I'm wrong get a refund.

While I think Wayland remote display will end up working just fine, "get a refund" is exactly the wrong attitude, and one that is doing a great deal to hold back open source. Don't like your Firefox buttons switching places every two weeks? Get a refund. Unity's window management for retards driving you up a wall? Get a refund. Newest GNOME version missing half the features you depend on? Get a refund. Guess what? Nobody is going to ask for a refund. They are going to switch to a different piece of software. And if every open source application has been trashed by developers, they'll switch to closed source.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

I'm really surprised. You're one of the nastier Linux zealots on Slashdot who has done a fair amount of trolling over the years, yet you have a positive attitude towards Wayland. Maybe you're growing up. Better late than never, they say.

Re:Will it really go the pulseaudio way?

[jones_supa]

Maybe I'll love Wayland too? Has anyone here actually seriously tried this thing before bashing it?

Not many, probably. Currently it's very hard to set up and there's not much you can do with it.

Re:Will it really go the pulseaudio way?

[boristhespider]

I've had to remote Firefox too, chiefly to access papers that are behind paywalls my university has access to but I have no access to at a university I'm visiting, or at home. You're making Firefox over X sound a lot better than it actually is.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Except that the applications that send remote drawing commands over the network are those that perform well on a DSL or lower, while those that send framebuffers are those that have several seconds of lag all the time.

SDL sends framebuffers - and only the changes, if the right functions are used. Yet, a simple solitaire program of mine written using SDL is slow enough that if I had the time, I would rewrite it to use XPutImage, to get rid of the multi-second lag.

Whatever remote system Wayland may get is likely to be as crappy as VNC (which also sends the framebuffer, rather than drawing commands). Welcome back to Windows 1995, where VNC is the best we can do.

Re:Will it really go the pulseaudio way?

[Jorl17]

But that's what I meant. Can't running X inside Wayland still allow us to use it as an X Server to point client apps to, thus only kind of changing X's backend?

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

And if every open source application has been trashed by developers, they'll switch to closed source.

So for a project like Firefox where the number of non contributing users counts; this makes a difference, but for the vast majority of projects it makes no difference as the only people that matter are contributors. A library I work on has on average a few thousand downloads but we have 6 contributors. Aside from those 6 people, everyone else is just contributing bandwidth costs. What do I care if those other people are satisfied with it ?

For projects like wayland etc.. the end user doesn't really get a say except through choosing a distro. Distros will either ship Xorg or wayland, possibly both and if the distro doesn't work for use case xyz then people won't use it.

If remote display isn't there for day one who really cares ? Those who need it will continue using X until someone pays for or wants the development bad enough.
If remote display is _that_ important then the commercial distros will absolutely support it as they will get funding for it.

Re:Will it really go the pulseaudio way?

[GauteL]

If the app is built on top of GTK+ (I assume Qt is the same?), the app will support both and the backend will be selected at runtime.

I.e. GDK_BACKEND will be wayland by default, but if you log in via ssh and set X forwarding, it can be set to "X11". It should be completely seamless.

Re:Will it really go the pulseaudio way?

[DrXym]

Wayland isn't heavily criticized from a technical standpoint. Most of the "criticism" seems to be from people moaning that it's not X. As you say, X will run on top of X and QT / GTK will be have backends that support X and Wayland. It's simply Ludditism from people who think everyone should suffer under a arcane windowing system to support their esoteric (and still supported) use cases.

Re:Will it really go the pulseaudio way?

[DrXym]

It does support remote display (and not just one protocol either) and widget sets will support X backend as long as there are volunteers to maintain it.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Virtually none of the features (remote drawing) that X provided over the network are used today (line/polygon drawing) and tool kits like Qt/GTK+ have you shipping framebuffers across the network

I'm using Motif and tcl/tk over tunnelled X every day you insensitive clod.

And when I do have to use a bling app, 'ssh -CX' generally tames the beast, even web browsers with horribly inefficient and unwanted fade in - fade out effects.

Check out some supercomputer cluster management software some time. Bling doesn't matter one iota. Having the damn thing do its job matters.

grumble grumble lawn grumble grumble hipsters grumble grumble cold dead hands.

Re:Will it really go the pulseaudio way?

[DrXym]

There is already a reference RDP implementation in Weston. So to answer your question, it's happened already.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

It's highly likely that Wayland's remote display will beat X.

Promises, promises. From the same bunch that failed so miserably and so serially getting their X implementation sped up -- where others had less trouble. To me it looks like they're blaming their own failure on the protocol --whatever its failings-- to have an excuse to do it all over again from the ground up, like a good little mozilla.

And now you're pulling a redmond by insisting it'll be sooooooo much better once it finally appears. Well, it's not there now, is it? Not the 3.0 that'll work a little, not the 2.0 ramshackle notional beta, not even the 1.0 public proof of concept pre-alpha we've come to know and love so much. So what use are your promises to me now then?

Re:Will it really go the pulseaudio way?

[DrXym]

Wayland is the protocol that clients talk with the compositor, not the compositor itself. The reference compositor Weston already implements an RDP server and does so in a remarkably small amount of source code.

As for it's performance, it will be no worse than X (or Xvnc) on modern apps because as everyone has stated, most modern apps are pushing pixmaps around anyway. If anything performance has the potential to be better because the remoting protocol can be asynchronous (unlike X) and the server doesn't have a handful of X and extensions processes with all their context switches to worry about.

Re:Will it really go the pulseaudio way?

[thegarbz]

No one is asking for feature refunds. They are simply bitching about users who demand every piece of software be 100% feature complete the moment it's first alpha team is announced and then continue to spew crap about it long through the development process.

Yes Firefox has abandoned geeks in favour of more simple users, well guess what there are many other packages out there that de-crappify the interface. Funnily enough that is EXACTLY the stance Wayland developers have taken from the very start. Design a flexible light weight modern protocol that does away with X's cruft and offloads stuff to the client. The users demand remote. Well if it matters that much to that many then the compositor can be written to support that. That is the flexibility that is missing from X.

The attitude was fine early on, but seeing every other bloody post on slashdot spewing the same crap, even after the Wayland team have announced remote desktop is possible, and even after the Wayland team have demonstrated code that does that, what do you think the answer is going to be?

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

I don't follow Wayland devopment outside of the occasional rant-fests on Slashdot like we are having now.

So you're saying you're wilfully uninformed yet that doesn't deter you from ranting about the very subject you know nothing about here on Slashdot?

Re:Will it really go the pulseaudio way?

[dbIII]

Apparently something good enough to be a bullet point on that was done last year but the fanboys don't know any details other than an announcement they point to. Maybe a Wayland developer will comment on one of these Wayland stories some day apart from the fanboys and we'll get some real information instead of links to a video of a half finished powerpoint presentation with no Wayland screenshots.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Robots don't have displays. It's really difficult to get your work done if your monitor keeps skittering away across the lab. Visualization tools for various pieces of robot state are much better than text dumps -- not surprisingly.

If you're running the UI on the robot and sending it over the network rather than just feeding the data necessary to a visualisation tool on your PC, you're doing it wrong.

I'm not saying that there aren't legitimate uses for UIs in headless machines, but I struggle to find one and the one you gave certainly doesn't seem to be one either.

Re:Will it really go the pulseaudio way?

[dbIII]

"Modern toolkit" is only portions of the most recent bits of gtk+. The rest doesn't do anything so stupid.

Re:Will it really go the pulseaudio way?

[dbIII]

Use an older firefox or something else without a recent gtk+.

Re:Will it really go the pulseaudio way?

[Beck_Neard]

I've built a few open source projects and been heavily criticized for my design choices but you know what? I agree with this. A lot of developers are too stubborn to make changes and it drives people away then they wonder why no one is using their project anymore.

But the flipside is true too. A lot of the time 'flaws' are actually sober and sane design choices which you have to get into the internals of the system to understand. People often don't get this and then bitch and moan about why something hasn't been done the way they like.

The Wayland devs seem pretty sober and sane so far, and I think they've made a lot of nice design choices. The problem of displaying graphics on a PC is an inherently ugly problem (and X is an ugly piece of software which visibly reflects that). If they can make it just a little bit better, it will be worth the wait, in my opinion.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

An SSH tunnel could alleviate your latency woes considerably. But w/e if you prefer waiting for 1s for the response to everyone of your clicks, I'm sure x-forwarding firefox is a good solution.

Re:Will it really go the pulseaudio way?

[paulatz]

I've had to remote Firefox too,

You're doing it wrong! Just set up an ssh tunnel and tell firefox to use it as sock proxy. This works seamlessly

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

I suspect it's because he'll want to access it from many devices, not tied down to a box with said visualisation tool.

I do this with my raspberrypi I can run an x application on the pi and have it all drawn to my laptop desktop n900 without having to start x on the pi itself. This is why it is X is so good.
ie login to pi from any where with X running using ssh, interact with it graphically by running the program on the pi with graphical output drawn on my local display, if I keep the output simple (ie not opengl ) I can do this across continents. vlc and framebuffer solutions etc struggle because they are trying to do to much...

Note the advantage to this is not having to fetch a particular app for the machine I am using.
It's similar to running a small web server on an embedded device where the remote accesses through a browser.
(Heck I have seen pic 16-bit microcontrollers with a firmware ethernet in them which can interact with a web browser).
This works graphically and at the same time uses relatively few resources (ie just a small program accessing a library). The heavy lifting is done by the machine with the remote display.

Re:Will it really go the pulseaudio way?

[jones_supa]

Yes, but will anyone ever develop it into a usable product?

Re:Will it really go the pulseaudio way?

[Blaskowicz]

Can it work without OpenGL or a GPU on the remote computer?

Re:Will it really go the pulseaudio way?

[Blaskowicz]

or elinks, or dillo.

Re:Will it really go the pulseaudio way?

[TheRaven64]

Multi-second lag on X sounds like misuse of the protocol. Unfortunately, this is largely the fault of Xlib. The X11 protocol was carefully designed to be asynchronous, but then they wrapped the whole thin in a synchronous C API. This means that you end up sending a small amount of data to the server then blocking while you wait for the reply. Used correctly, you'd send all of the display updates sequentially, get the updates as the arrive, and later handle any errors that appear. Similarly, input events would be handled as they arrived, rather than synchronising everything. Unfortunately, although XCB fixes this, most toolkits are designed around the synchronous model so they don't really take advantage of it.

Re:Will it really go the pulseaudio way?

[squiggleslash]

Probably never.

Remote Display is something that Wayland's developers do not consider important in any way shape or form whatsoever. When these projects started, anyone saying "Hold on, one of the major features of X I like is the network transparency" was summarily ridiculed by the Wayland group as an out of touch nerd who doesn't know what "real users" want or need.

And, FWIW, the GP is completely flat-out wrong given everything I've heard thus far. "Wayland Remote display" will be, according to what I've read so far, if ever implemented based on H.264 (or something similar), essentially a slightly more efficient version of VNC (which had an MPEG-1 mode) running in a per-window mode. Because nothing says "better" than something that sucks up CPU on both sides of the network, introduces a GoP of latency, and requires megabytes of throughput every second to show a full screen word processor.

Thanks Shuttleworth. Yes, you. Yes, I know you're proposing Mir. But if you hadn't said Ubuntu was switching to Mir at some point in the future, the debate would still be "kludgy but proven X.org vs untried, poorly thought out new system", not "untried poorly thought out new system vs another".

Re: Will it really go the pulseaudio way?

[Anonymous Coward]

Yes there is a software back end someone cooked up.

Re:Will it really go the pulseaudio way?

[DrXym]

Quite obviously yes.

Re:Will it really go the pulseaudio way?

[jones_supa]

Nothing is obvious in open source. :D

2 years later: "Yeah, Wayland has RDP support... But it's kind of broken and buggy. It requires these patches to be applied, manual creation of this 1000-line configuration file, and the sessions have always to be started from command line. Also, you have to use this certain distro. We are waiting it to be fixed, but the component has insufficient manpower and resources for the task to ever be actually completed anyway. Just use VNC, sigh..."

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Yes, it can run on a framebuffer and on headless devices.

Re:Will it really go the pulseaudio way?

[serviscope_minor]

There is already a reference RDP implementation in Weston. So to answer your question, it's happened already.

Is it in Weston or Wayland? If the former, than that's a whole pile of yuck if you can only get remoting if you run the right window manager (or compositor in Wayland-speak).

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

First, why are you using a GUI in such a situation?

Because I'm a sysadmin and sometimes GUI tools need to run on the server and I don't need X11 (and gdm and xfs and ...) running all the time. So "ssh -X/-Y" is a handy tool to have.

I also sometimes use a Mac for my monitor and keyboard (i.e., fixing something from home when I'm on-call) but admin Unix-y machines, so SSH X11 forwarding works there as well so I don't have to spin up a VM and work in a tiny window.

VNC/RDP can work sometimes, but having a rootless display makes the app I'm running just another window on my screen.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Have everything served through The webserver as, i dunno a webpage? It even has built in visualization for free, because well THATS ITS JOB.

Re:Will it really go the pulseaudio way?

[AC-x]

Robots don't have displays. It's really difficult to get your work done if your monitor keeps skittering away across the lab. Visualization tools for various pieces of robot state are much better than text dumps -- not surprisingly.

For this use case wouldn't it be a lot more appropriate to stream raw data from the robot to software running on a desktop machine to represent it visually? Surely it's a waste of CPU effort drawing a GUI on the robot itself?

Re:Will it really go the pulseaudio way?

[johnsie]

If it's an alpha then why have they given it a version code higher then 1.0? I would much prefer alphas to be 0.xxx

Re:Will it really go the pulseaudio way?

[thegarbz]

Reread. I didn't say Wayland was in Alpha. I said the network transparency bitchfest by the uninformed started back at the alpha releases and even now when remote application display was shown to be working in Weston the bitchfest by the uninformed continues.

People somehow believe that every new effort needs to be 100% feature complete and ready to go from day dot. Guess what, there's a reason compatibility layers are written when new software is released. Magically making everything out of nothing is not how programming works, attempting to make sure everything works before release is exactly how GNU Hurd got its reputation.

Re:Will it really go the pulseaudio way?

[DrXym]

It's a reference implementation, a proof of concept to demonstrate to people complaining that yes wayland can in fact be remoted assuming a compositor provides with the support. It's not wayland's job to do the remoting, it's the compositor's. I see no reason to think that once wayland is switched on by default in a dist or two that compositors will explicitly support remoting, or there will be dedicated compositors for that purpose.

Re:Will it really go the pulseaudio way?

[DrXym]

Wayland is just the API and protocol for clients and servers to allocate & release buffers and pass input and display events around. It doesn't say how those buffers get shown to the user - that's the compositor's job. Weston provides a a reference implementation of remoting demonstrating it is possible. It doesn't mean it's perfect, or comes with a pile of advanced options.

I expect it will improve and there will also be dedicated compositors for headless servers and the like.

Re:Will it really go the pulseaudio way?

[jones_supa]

I hope so.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

the only one I've seen made with any muster has been network transparency - but even that could be solved rather easily given the way Wayland works with framebuffers

Not really, the one you saw about network transparency is some cretin who can't get through his stupid skull that modern apps on X are not network transparent.

Re:Will it really go the pulseaudio way?

[serviscope_minor]

I expect it will improve and there will also be dedicated compositors for headless servers and the like.

So... you';d have one compositor (on the headless machine) which does nothing but send the bitmap over the network and a client program on the display server machine which collects that and opens windows in whatever compositor you're running?

What I mean by this is I don't understand at all why it would be in Weston: it sounds like one want a couple of utility programs instead to do that job, not a whole compositor with window management built in deal.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Exactly.

This is what ROS is for ( http://www.ros.org ). You publish the needed data on a ROS topic and any computer on the network that has ROS installed can visualize the data as it sees fit. Don't be deceived by the name: ROS is not an operating system, but mostly just a message passing system with standard message formats and tools. It is not very efficient, but it is quite flexible.

Re: Will it really go the pulseaudio way?

[BitwizeGHC]

Remote display on Wayland will be much better and more modular than X11. X11 mandates a chatty, slow, obsolete protocol for remote display and applications MUST be network-aware. With Wayland, you can run a compositor on the remote server that doesn't display its clients on the screen but rather transmits streaming video of the clients back to a Wayland client on your desktop which decodes and displays these streams. And neither your local Wayland compositor nor the remote clients need be network aware.

X11 is morbidly obsolete.

Re:Will it really go the pulseaudio way?

[kenaaker]

I've seen this, (a while ago it's true). I was curious about why some GTK applications had such terrible remote performance. When I ran the applications through xscope, before the first window materialized, I saw thousands of calls to XGetGeometry(). What made that doubly silly is that any window geometry change is provided to the client as an XConfigureEvent. So, a local copy of window geometry information can be easily kept up to date.

Re:Will it really go the pulseaudio way?

[Wesley Felter]

There are two different ways to do network display: the RDP way and the right way. With RDP you're sending the entire "screen" over the network, so all the windows have to be composited first. Thus RDP requires a fully featured compositor like Weston on the remote end.

The right way is to send each window over the network, which should require a lightweight compression proxy. No one appears to be working on this.

Re:Will it really go the pulseaudio way?

[MMC Monster]

I'm sorry. I didn't realize that Wayland 1.5 was an alpha release. Presumably it will be complete when it hits 1.0?

Re:Will it really go the pulseaudio way?

[Kremmy]

Because the webpage data will likely have a higher footprint. Once you pack on the full distribution of a web application that does everything you need it to onto the embedded device, does it have any space left? There are certainly ways to save space with web applications, but for the most part we're talking distributing and executing raw source code which is often orders of magnitude larger than compiled binaries. This matters in the embedded space.

Re:Will it really go the pulseaudio way?

[Kremmy]

It started back in Alpha because it had to.
It's still not ready to go, at version 1.5.
I still need remote display functionality.
It's still not ready to go, at version 1.5.
It doesn't have to be feature complete in alpha.
It's still not ready to go, at version 1.5.
Cool RDP patches, bro. How do I use them?
Have they made it as easy as ssh -X?
Don't argue the merits of half-baked software that is incapable of providing equivalent functionality to the pre-existing solutions as if that's acceptable, that's FUCKED UP. I need a solution that works, not something I get to sit around and wait for them to develop the functionality for.
I'm getting really tired of watching these systems be slowly dismantled in favor of doing it a different way that isn't necessary better, and doesn't provide the functionality that I've come to expect from my computer systems. This isn't some groundbreaking new territory that the Wayland guys are treading, computer history has covered this ground. I'm all about new ways of thinking about things, I'm not okay with sacrificing my workflow for the new hotness when I've been using roughly the same stable stack for damn near 20 years now and the new shit doesn't provide equivalent functionality.
It's not progress.

Re:Will it really go the pulseaudio way?

[Kremmy]

It's an ugly problem, but it's one that isn't half as ugly as they want to make it.
In fact, I would go so far as to say they're actively trying to make it uglier.
At the end of the day, we're just drawing colored rectangles on the screen.
Is it too much to ask that they not fuck that up?

Re:Will it really go the pulseaudio way?

[DrXym]

On Windows I can login into a PC, walk over to another machine and remote login to that same session. For that you probably want the window manager to host the remote session. But if you have headless machine that is only used in a specific way then perhaps something simpler will do. I don't see that either is precluded.

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Just tried gvim from my Debian box with XFCE over a 150Mbps wlan and I could not see any lag on it. Do you use some fancy transparent window manager or similar effects which cause the whole window content to be transferred?

Re:Will it really go the pulseaudio way?

[Anonymous Coward]

Line/polygon drawing are still used, via XRender. X11 even has some OpenGL remoting support.

Re:Will it really go the pulseaudio way?

[Beck_Neard]

It's ugly because there are a lot of different ways of rendering those graphics to your screen, and they are highly variable depending on your graphics hardware and what exactly you intend to do. If you're just drawing a bunch of browser windows you can get by with communicating with your graphics hardware on a per-pixel basis. But if you want to get into games and so on you have to provide an interface for your software to communicate with the hardware's OpenGL implementation. And you have to support a system where all of these things are happening simultaneously (like, say, an OpenGL context in one window overlaid on a browser window, or even a WebGL app running inside a browser window). And all of this has to be wrapped around a common interface, which has to be fast, reliable, and secure.

Re:Will it really go the pulseaudio way?

[thegarbz]

Then don't use it yet. Use it when it has the functionality you need.

In other use my computer won't send me to the moon yet. Clearly it's functionally not ready and I need to bitch about it on facebook every day.

Re:Will it really go the pulseaudio way?

[squiggleslash]

(Original has disappeared because of abusive moderation. FFS. The desire of Wayland fanboys to silence any opposition is beyond ridiculous.)

Probably never.

Remote Display is something that Wayland's developers do not consider important in any way shape or form whatsoever. When these projects started, anyone saying "Hold on, one of the major features of X I like is the network transparency" was summarily ridiculed by the Wayland group as an out of touch nerd who doesn't know what "real users" want or need.

And, FWIW, the GP is completely flat-out wrong given everything I've heard thus far. "Wayland Remote display" will be, according to what I've read so far, if ever implemented based on H.264 (or something similar), essentially a slightly more efficient version of VNC (which had an MPEG-1 mode) running in a per-window mode. Because nothing says "better" than something that sucks up CPU on both sides of the network, introduces a GoP of latency, and requires megabytes of throughput every second to show a full screen word processor.

Thanks Shuttleworth. Yes, you. Yes, I know you're proposing Mir. But if you hadn't said Ubuntu was switching to Mir at some point in the future, the debate would still be "kludgy but proven X.org vs untried, poorly thought out new system", not "untried poorly thought out new system vs another".

Re:Will it really go the pulseaudio way?

[spiralx]

Actually, you were moderated down for your overly vitriolic rant full of factual errors, logical fallacies and heroic assumptions in the guise of fact. A perfect example of the "peanut gallery rant-fest" as a comment above yours mentions. And seriously, you got moderated down, then threw a strop, and then posted it again with a hissy fit preamble? How long have you been posting here?

Re:Will it really go the pulseaudio way?

[Kremmy]

Except that there's no reason you would expect your computer to send you to the moon. I'm not about to ride my bike across the sea. It's not an argument against Wayland, it's an argument against pushing for things that Aren't Ready and pretending like they're suitable replacements for what they're intended to replace.
If the product is not up for the task, it's not up for the task. There are a lot of potentially good reasons for Wayland but every single one of them falls short until it matures to the point that these arguments are unnecessary.

Re:Will it really go the pulseaudio way?

[squiggleslash]

I'm sorry, but WTF are you talking about? If there were factual errors or logical fallacies (and what are "heroic assumptions") you'd correct them, as would others.

And generally, I've been posting here long enough to know that some people do abuse mod points, and the only workaround is to repost what's been abusively modded down if you feel the moderation was undeserved. Hissy fit? How? It's a statement of fact.

The truth is every time I mention problems with Wayland, from criticizing the decision to start at "feature incomplete" for the architecture to the fact they're throwing out a perfectly good system in favor of an ideological rewrite (pretty much guaranteed to waste time and resources on producing something that will have the same results), I get modded to the floor. Not "Now hang on Squiggy, in this case this argument may be wrong for this reason", but just blatant abuse.

Wayland advocates are too thin skinned to handle criticism. And that, on top of the fact it's throwing out a working body of code, on top of the fact that they're omitting critical features from day #1, on top of the fact that the design workarounds mentioned thus far (like the effing stupid RDP H.264 bullshit) are just about the worst way of pacifying those that have said they want features the designers don't give a shit about, is why the project is destined to produce something destined to fail as an adequate X replacement.

If you can't listen to criticism, you're doomed from the start.

Re:Will it really go the pulseaudio way?

[spiralx]

I don't run Linux, and I've never had to deal with X, so this isn't an emotive issue for me. But your post was clearly not a serious of facts, more of a backlash against perceived slights from Wayland developers and/or "fanboys", the latter being a useful way to smear people who disagree.

Your post might make more sense if the existence of Wayland meant that X no longer existed, but as that's clearly not true it's hard to take claims of "throwing out a perfectly good system in favour of an ideological rewrite" seriously. And ideology? There have been plenty of technical arguments since the beginning, here's one set that was posted on /. a while back, and it's just the first one I found.

As far as I'm aware X currently does act like VNC in most cases, except without any compression at all and a synchronous API - so nobody uses it directly because of the performance issues, instead using ssh as a tunnel. Even having it act like per-window VNC with H.264 compression would be an advantage. But anyway, that's all part of the compositor, which now has RDP as part of the core, and I've yet to see any explanation of why or how X forwarding is different or better than rootless RDP.

The assumption that forwarding is a critical feature is based on the idea that your personal requirements are the only important ones. If a piece of software doesn't do what you want, don't use it. As it turns out, they are supporting it (as I'd read from pretty much day one), it's just taken time to get to that point... as you'd expect from alpha software.

Re:Will it really go the pulseaudio way?

[allo]

some simple thought: Do you think slashdotters here are the only ones thinking about "X Forwarding" for wayland?

I guess they know its easy to do and best to be done, when the other stuff is api-stable and maybe working stable as well ;).

Re:Will it really go the pulseaudio way?

[allo]

wayland is just the way to go. You see, that they are taking it serious.
They made a concept, they made a reference implementation, they are still calling it alpha, they are waiting for the toolkits and main desktops, they are reacting to feedback, they are looking at the distros and graphic card vendors.
It seems the concept was well thought from scratch instead of building on X11, clean and without old cruft.

So i guess wayland is like IPv6. Incompatible for a good reason, but using the fact that its incompatible anyway to implement stuff which would be impossible to add to X11.
And like IPv6, the transition is not easy, but i guess the compatiblity layers are easier to do.

Re:Will it really go the pulseaudio way?

[renoX]

> RDP protocol support was merged into Wayland over a year ago.

That's incorrect: RDP support was merged into **Weston** if you use another Wayland compositor it may have or not RDP support..

I agree with your point about XWayland: as long as the toolkits doesn't remove support for X, Wayland's remote display capabilities can only be superior or equal to those of X.
Now, the tough question is: how long will the toolkits keep the X support?
The developers of desktop tech have been called CADT among other things..

I would be more or less wait for other desktops

[thieh]

So it will be more widely adopted and as such have a more unified troubleshooting across different DE's. Or maybe that's just me distro/DE hopping.

I hate FPs like this.

[pla]

I hate FPs like this - Yes, enough that I feel a need to complain about them in the discussion instead of just moving on.

I use Linux. I've rolled my own kernels (as in actually writing code, not just a custom config and build of the stock tree). And I have never heard of Wayland or Weston. And out of three links, could you have included one going to "what the hell is Wayland"? No. You have a release announcement and a PR page.

Wayland may well rock the world. But when writing up an FP about something obscure (yes, it is - I don't care how many of your friends run it), you would do well to link to an intro-to-obscure-thing page.

Just sayin'.

Re:I hate FPs like this.

[nighty5]

I don't write Linux code, but even I know what Wayland is.

Re:I hate FPs like this.

[Anonymous Coward]

Which troll cave have you been living in the past 4 years?

Re:I hate FPs like this.

[Anonymous Coward]

I give up...what does "FP" stand for? Can't be "first post"...

Re:I hate FPs like this.

[Anonymous Coward]

If you're just writing useless/toy software (which you are, since you're clearly lost in around 1995), you're not better than people who just do custom configs.

Re:I hate FPs like this.

[Anonymous Coward]

That's the problem, it's been broadcast to all of the caves and he was under a bridge instead

Re:I hate FPs like this.

[jones_supa]

Suckah! Wayland is widely known in the Linux world and requires no further explanation at this point.

Re:I hate FPs like this.

[wonkey_monkey]

Slashdot is not "news for the Linux world," and even if it was, not everyone in "Linux world" is so deeply involved as to keep up to date on every developing piece of software.

All a summary writer has to do is drop in a brief, casual couple of words about what (roughly) it is, and those who need informing are slightly better informed, while those who are already informed don't notice and aren't offended.

Ever notice how the BBC will often refer to "US President Barack Obama," or drop in a reference to the team a famous footballer player plays for, even though one would think those would both be widely known facts among the readership of such articles? Chances are, you didn't notice and didn't care.

Re:I hate FPs like this.

[jones_supa]

Hahhaa! Yeah, right.

Re:I hate FPs like this.

[johnsie]

The BBC get more than $1 billion of taxpayers to spend. Slashdot doesn't.

Re:I hate FPs like this.

[Anonymous Coward]

What is lwn.net is for, read it and weep...

Again?

[Anonymous Coward]

How many stories is Slashdot going to run about Wayland before it's actually a viable product? Apparently it's already up to version 1.5 and it's still not there...

Wayland is nothing until

[morgauxo]

Wayland is nothing until there is good remote display support.

Re:Wayland is nothing until

[blackpaw]

Oddly enough X gets along fine with crap remote display support.

Re:Wayland is nothing until

[Anonymous Coward]

Real admins master the command line.

Re:Wayland is nothing until

[martas]

You're right. That's why nobody uses Windows or OSX.

Re:Wayland is nothing until

[UltraZelda64]

Have fun watching YouTube in Lynx.

I personally make a distinction between "using" and "administering" a machine, and as a user, I tend to run X11 (these days often with a tiling window manager). When I want to perform some administrative tasks, I'll often just run a terminal emulator within that environment. Face it, while great for many things, the command line--especially in its raw, no-X11 form, is pretty limited in many areas from the point of view of a typical user.

Don't get me wrong though; I'll often use wget instead of Firefox to download files, do basic file system operations in a terminal, even play an occasional podcast in mplayer. But really, it is not optimal to use the CLI 100% for everyday use for semi-normal people.

Re:Wayland is nothing until

[blackpaw]

Well windows has Remote Desktop (RDP), but I think that supports your point. I use xrdp for remote gui connects to my linux boxes as it performs a hell of a lot better than X, which is unusable on anything except a local LAN.

Re:Wayland is nothing until

[Anonymous Coward]

i'm confused; are you saying windows/osx don't have remote display support? [Win+R]mstsc[Enter] .. yep, works fine.

Re:Wayland is nothing until

[martas]

Yeah, I mean, X forwarding is one of those things that you very rarely need, and when you do, it either doesn't work at all because [inexplicable reasons that can't be fixed without admin on a machine you don't own or having to restart a machine that's 200 miles away], or it sort of works but makes you wish it didn't because it's too slow/glitchy. Typically ssh does the trick, sometimes combined with forwarding gnome-open to view images/pdfs/whatever (which is terrible due to the aforementioned reasons), because scp'ing every time is even more painful. All my needs would be perfectly satisfied with something that combined ssh for some cli, and some kind of automated easy to set up and disable file updating utility (like dropbox without the third party server). Since I only ever need this stuff once in a blue moon ad hoc, setting up svn or something like that is just overkill. Example -- I want to run some code on a remote machine that plots a figure or pdflatex's some file and quickly view the output, without downloading a copy of all the code/data. I have yet to find a way to do this that isn't extremely awkward...

Re:Wayland is nothing until

[Anonymous Coward]

And how often do you need to watch youtube on a remote desktop when administering a remote computer?

Re:Wayland is nothing until

[Anonymous Coward]

I'm not entirely sure if this is what you want but have you looked into sshfs? I use it quite regularly without issue.

Re:Wayland is nothing until

[multi io]

It seems the real remote drawing / display technology these days is HTML/CSS, carried via HTTP. It even supports running client-provided code locally on the display server, as did NeWS and DPS 20 years ago, to render animations and depressed beveled buttons without incurring a server roundtrip (and Javascript is generally much nicer than PostScript -- you can even run the whole program on the display server if you want). And the protocol is a bit backwards in that the display server (aka "web browser"), rather than the client (aka "web server"), initiates the protocol requests.

[scnr]

Re:Wayland is nothing until

[UltraZelda64]

And how often do you need to watch youtube on a remote desktop when administering a remote computer?

That brings me back to the second sentence in my post.

"I personally make a distinction between "using" and "administering" a machine,..."

Did you even read the post at all? Obviously, you are talking about administering a system. But I can guarantee that I am not administering a system while I sit here wasting time posting crap on Slashdot, and I have this funny feeling you're not either.

Re:Wayland is nothing until

[Anonymous Coward]

That, to give the impression of working without unacceptable latencies, it might need a dedicated 100Gb/s connection.
It is painful over 1Gb/s for CAD/CAM work (and no choice of running it locally because the software is node locked to the machine where it runs).

Re:Wayland is nothing until

[Knightman]

and some kind of automated easy to set up and disable file updating utility (like dropbox without the third party server)

BT Sync? http://www.bittorrent.com/sync

Re:Wayland is nothing until

[martas]

Huh, that's actually pretty interesting. Thanks for the link.

Re:Wayland is nothing until

[Anonymous Coward]

In that case, I'd suggest solving the cause of the problem rather than shitting up next-gen display protocols with legacy crap that next-to-nobody needs.

Re:Wayland is nothing until

[Anonymous Coward]

How the f... do you get X not to work without admin rights on the remote machine? Everything is in the user program and libraries, unlike inferior competitors like VNC or RDP, which require setting up a forwarding service on the remote machine (RDP is preinstalled on Windows, but still requires setting up access).

As for SSH doing the trick, you should be using SSH if not on the LAN anyway (for security reasons - the X protocol is not encrypted), and even on the LAN it may be preferable to copying "magic cookies" around or doing stupid things with xhost.

Re:Wayland is nothing until

[Anonymous Coward]

Real admins use whatever tool gets the job done. It's only bitches and zealots who will only use a hammer or a cli.

Re:Wayland is nothing until

[thegarbz]

The Wayland protocol will never have remote display support.

The compositor on the other hand already has it. I find it funny that Slashdot users don't understand this given how the article about it was posted on slashdot.

Re:Wayland is nothing until

[martin-boundary]

But really, it is not optimal to use the CLI 100% for everyday use for semi-normal people.

Your reductionist dichotomy is too naive. In the real world, there are many more shades of gray where people who use computers have expertise to handle complexity, and where CLI use isn't just for "administration".

But your worst mistake is to assume that somehow, we need to choose only one thing that is optimal for semi-normal people, whoever these people are. Here's a clue, in FOSS we don't need to choose at all. We already have a working Xorg system, and there's no need to replace it. It works. If someone wants something else, by all means they should build it for like minded people, but not by depriving existing users under the dubious claim of optimizing for semi normals, and if you're not semi normal, too bad.

FOSS isn't about concentrating resources for the greatest return on investment for the greatest market share. FOSS is about writing software that solves a problem. Throwing away software that already solves a problem just because it's old and unhip isn't FOSS, it's the kind of commercial decision a Google would make.

Re:Wayland is nothing until

[Beck_Neard]

And what if a particular program has no command-line version, or the command-line version sucks and is extremely limited (e.g. running a MATLAB instance on a remote cluster)?

I don't know if you're serious or not but if you're serious you should know that no one enjoys using slow, crappy remote desktops when there is a choice. The problem is that we often have no choice.

Re:Wayland is nothing until

[martas]

And how often do you need remote display support while using a machine?

Re:Wayland is nothing until

[jones_supa]

Wayland is nothing until there is good remote display support.

There is already remote desktop support in Wayland. It uses RDP and works for individual windows too.

Re:Wayland is nothing until

[jones_supa]

Using remote desktop is the practical solution for his problem.

Re:Wayland is nothing until

[Blaskowicz]

A long VGA cable run could be nicer, or these days there's beaming of HDMI on CAT6 for a bit more investment.

Re:Wayland is nothing until

[jones_supa]

True. If the machine is in the same office, those might be viable options too.

Re:Wayland is nothing until

[Anonymous Coward]

TIL a windows protrick. Hope I'll ever find a use for it.

Re:Wayland is nothing until

[Bill_the_Engineer]

Have fun watching YouTube in Lynx.

It's not that bad once you get used to it. What you now see as random characters falling on the screen will soon become something you'd recognize like that hot blonde in the smoking red dress walking down the crowded sidewalk.

Re:Wayland is nothing until

[Bill_the_Engineer]

Nomachine (NX) removed a lot of the pain that I had with X over VPN. I used it several years ago on a project and the performance was such that I barely noticed that I wasn't on the host's LAN. I liked it better than VNC at the time because I was still allowed to open a single X window without having to put up with an entire remote desktop.

Re:Wayland is nothing until

[Anonymous Coward]

Were you successful getting the individual window mode working? I tried using it with the freeRDP client on both OS X and Linux and the only thing I could reliably do was an entire desktop.

Re:Wayland is nothing until

[jones_supa]

I have not done any testing.

Re:Wayland is nothing until

[bill_mcgonigle]

I tried using it with the freeRDP client

can you point to the Wayland mailing list thread you posted about this? I'm not being [merely] snarky - I want to know if the developers helped identify the problem or not.

Per-window RDP is OK by me and then I'll tell everybody like the GP to shut it and complain about something else. But if it's just checked-in-but-broken then that's a different story.

Re:Wayland is nothing until

[praxis]

I personally do on occasion. I sometimes run R on an EC2 instance to be able to use a beefy box and use it remotely from a device of my choice. I don't think that's a particularly odd use case for the types of people that use X today. If Weyland/Weston are not targeting those people, then so be it. If they are, then the remote display use-cases are going to suffer.

Re:Wayland is nothing until

[AntiSol]

I tried this on my bosses mac but couldn't wind the 'win' key - can you help?

Re:Wayland is nothing until

[spiralx]

You might be interested in this podcast if I'm reading your post right... they talk quite a bit about using js-git to mount GitHub repositories as file systems, so that you can mount a repository, copy files into it and then run a commit and have the stuff you've copied automagically pushed to GitHub.

It's certainly interesting stuff even if possibly overkill :)

Re:Wayland is nothing until

[allo]

> Have fun watching YouTube in Lynx.
Have fun with youtube via x-forwarding.

Re:Wayland is nothing until

[allo]

shouldn't such tools have some kind of client-server model?

Re:Wayland is nothing until

[praxis]

shouldn't such tools have some kind of client-server model?

They do. Hopefully Weyland/Weston will as well and hopefully it will work as good or better than the solution we have today.

Re:Wayland is nothing until

[allo]

i speak about a client-server model of having the gui here, and the server doing the work there, not about having the compositor here and the gui there.

for example have a look at mysql-workbench. You have a nice gui. on your pc. It connects via tcp or tcp-via-ssh to your server, where the db is running. nobody wants to have this program running on the server, forwarded via X11, rdp, NX or VNC.

so, mysql has the client-server model? What about having something like this for R and the display of its charts?

Re:Wayland is nothing until

[praxis]

There's no reason we couldn't have both. If the application developer does want to provide their own protocol and a thick client and all that, that's great. That doesn't mean the display layer should not have a network layer to allow remote displays. They solve different problems and are useful in different situations.

Re:Wayland is nothing until

[allo]

jep, thats okay. But i think a thick client is more suitable for many tasks, especially where the server part runs as privilgeded user.

Again?At least it's better than Mir

[Anonymous Coward]

And like it or not, X DOES need to be replaced. ASAP. Technical users who need all it's (hacked in) functions can stick to X, but the general public just needs a smoother, faster, less buggy solution. And that is/will be Wayland.

And Mir is a stinking pile of dog crap. Canonical needs to stop worrying about "not invented here" and actually work with the Linux community if it wants Ubuntu to become the most popular/most recommended distro again.

Re:Again?At least it's better than Mir

[Anonymous Coward]

Right now they're all stinking piles of dog crap. Eventually one of them will stop being dog crap, at which point I would like to know about it. Maybe it'll be Berlin. I hear lots of good things about Berlin on Slashdot.

This steady stream of "Wayland 1.X released; it's still dog crap" just makes me think it's vaporware.

Re:Again?At least it's better than Mir

[Anonymous Coward]

And like it or not, X DOES need to be replaced. ASAP. Technical users who need all it's (hacked in) functions can stick to X, but the general public just needs a smoother, faster, less buggy solution.

Replacing X with Wayland will solve very few real problems, and it will almost certainly create some (remember the pulseaudio and systemd hype before they were actually forced on users by all major distributions ?). Wayland cannot possibly solve problems like these listed below, simply because they are not in X:
- poorly written, buggy, slow, and outdated (in terms of API version compatibility) OpenGL drivers - that is, basically everything other than the Nvidia proprietary driver
- bloated, slow, and buggy - but unfortunately widely used - GUI toolkits
- bloated, slow, and buggy applications
- desktop environments and window managers with the same flaws as above
User-unfriendly configuration and bad (or non-existent) documentation will likely also continue to be issues, as they tend to be with open source software, since they are a low priority for programmers who mainly code for themselves, rather than as a real job. To summarize, games will continue to perform notably worse than on WIndows/Direct3D, KDE and GNOME will still suck in all their usual ways, major Linux desktop applications will still continue to be inferior to their Windows counterparts, and users will still end up having to hack configuration files when things (not uncommonly) go wrong. There will just be some additional PITA related to migrating from one windowing system that works to another.

Alright guys...

[Anonymous Coward]

This one looks almost, kinda-sorta useful and on its way to more adoption. This is obviously bad and against everything we stand for! I'm starting up a new WaylandX project on github (https://github.com/waylandx) to build a better one of these. Come and join me!

Re:Alright guys...

[AntiSol]

I like some of the changes you 've made, but I disagree with your philosophy of standing for things. Due to this, I'm starting up a new WaylandX-ex ((https://github.com/waylandx) project on github to avoid these political issues. Join me instead!

Wayland is nothing until

[Anonymous Coward]

What's the problem with VNC ?

And not...

[Anonymous Coward]

a fuck was given

Clipboards?

[serviscope_minor]

Does anyone know if Wayland has the nice dual clipboard system like X? Or are we going to be stuck with something hideously primitive like other well known operaing systems?

Re:Clipboards?

[dbIII]

Clipboard? It's a framebuffer with a compositor on top. Clipboards are a client problem (as are many other things).
Now do you get why the "X sux" stuff from Wayland fanboys is annoying? Wayland is designed to be something different to X with different goals. Those of us that "want to run software from 1996" are made fun of in Wayland presentations, which would be fine if they were not also telling us to stop using X.

Re:Clipboards?

[serviscope_minor]

Clipboard? It's a framebuffer with a compositor on top. Clipboards are a client problem (as are many other things).

Well, no, it's not. It's also a keyboard and mouse input system.

It also deals with copy and paste and drag and drop:

http://lwn.net/Articles/491509...

Because it's a windowing system and it turns out that just a compositor alone isn't enough (who knew, eh?). It's also interesting. Apparently Wayland implements passing of data by just passing a file descriptor, apparently instead of reimplementing 10 pages of ICCCM grot. The thing about the 10 pages of ICCCM grot is it's really REALLY well specified and a random person from the internet can come along, read the ICCCM, grok it (yes, I have actually implemented copy/paste and XDnD from the specs) and get it working. It's not that hard.

The wayland one seems poorly specified by comparison. For example they don't specify teeny-tiny details liekl whether the FD must be seekable, for example. So, do you have to write a local file, or can you pass a socket? Who knows! It's really easy to have a short, simple spec when it's full of ambiguity and people haven't had 26 years to beat it into a definitive, unambiguous state. Anyway, I digress.

Now do you get why the "X sux" stuff from Wayland fanboys is annoying?

Yes, but it's more annoying when it comes from the Wayland author FUDmonsters who understand X11 and yet still make silly claims about it. For example, from the link above, Packard claims:

X was created before there was MIME or Unicode, so there are many pages expended in the X specifications to do things that are more easily handled with MIME types and UTF-8 these days. For cut-and-paste and drag-and-drop, Wayland uses MIME-labeled UTF-8 encoded objects.

Well, that sounds all like OMG X sucks we need MIME and UTF-8. Well the thing is, in order to list types from a copy/paste transfer, applications exchange a string (i.e. atom) with the type name(s) available. And guess what? Almost everything these days except for plain text is exchanged using MIME types. If the MIME-type specifies UTF-8, then the data will be in UTF-8 format. So basically, X names types with a string, just like MIME, and MIME works *perfectly* without modifying or respecifying anything.

You can verify this easily: download and install a copy/paste debugger/sniffer and look at the list of types available that programs offer.

The ICCCM also specifies a few (non-MIME) types that you might like to support, such as TEXT, which maps perfectly on to text/plain and is all of 1/2 a line to implement (if(typeAtom == TEXT || typeAtom == textPlainAtom)...). And X11 sends arbitrary data (including NULs) because it represents data as data+length not a string, so you can exchange anything, such as UTF8.

Anyway, KP implies that that doesn't work with X11 copy and paste, whereas in truth it works perfectly and without any faff or hacking.

Wayland is designed to be something different to X with different goals.

Not so much. It's designed to replace X wholesale. It does windowing, compositing, input, copy/paste/DnD, and a bit opf inter client communication.

Those of us that "want to run software from 1996" are made fun of in Wayland presentations,

Yeah us with our legacy programs. From stroustrup:

"Legacy code" often differs from its suggested alternative by actually working and scaling.

Meanwhile, I shall keep using legacy programs productively. XTerm works amazingly well, still. gvim works great---though I find I sometimes have to compile it with GTK disabled and with XAW (seriously WTF??) support because GTK can't seem to get its shit in order with fonts and everytime ubuntu updates itself/reboots, the font size changes. Xfig is old but works really well within its domain for producing simple, effective figures.

etc etc blah blah.

I also use some more modern programs too. And they all w

Re:Clipboards?

[thegarbz]

You seem to be all for the X11 side of things. Yes it's well specified and there are many benefits.

Do you care to put as much detail into the downsides? Such as that it's so well specified that things like the volume keys on the keyboard can't be passed onto a music player because the laptop lid is closed and the lockscreen is fullscreen and the way X works means that any fullscreen program completely monopolises the input system? Yes I'm all for specification, but if network systems were specified the same way as the X11 protocol then we wouldn't have 7 layers, everything would be TCP with fixed sized headers and one port on the other end, and you could kiss most of the flexibility of the network aside.

What Wayland is trying to achieve is exactly the opposite. Make a display system by specifying the absolute minimum and leaving the client the flexibility to implement the remaining features as it sees fit. This is a bad thing if you want 100% certainty of functionality however limited it may be or interoperability, however it's fantastic for systems where requirements change as technology advances, such as the desktop.

Re:Clipboards?

[serviscope_minor]

You seem to be all for the X11 side of things. Yes it's well specified and there are many benefits.

Yes, but I restricted myself to only talking about the anti-X FUD.

Such as that it's so well specified that things like the volume keys on the keyboard can't be passed onto a music player because the laptop lid is closed and the lockscreen is fullscreen and the way X works means that any fullscreen program completely monopolises the input system?

Works for me (TM). It's set up that my WM does stuff when the volume keys are pressed. It must have a grab on those keys because it always works, lid closed or not.

Actually, this is another piece of FUD which has been floating around: wayland is better because the WM/screensaver interaction on X is crap but in Wayland, the idea is to have the compositor do that. Well, there's nothing in X that stops one having the WM run the screensaver directly, and there's nothing in Wayland that stops one having the screensaver being separate.

Yes I'm all for specification, but if network systems were specified the same way as the X11 protocol then we wouldn't have 7 layers, everything would be TCP with fixed sized headers and one port on the other end, and you could kiss most of the flexibility of the network aside.

Uh, TCP headers ARE fixed length. You're beginning sound like one of the FUDdy types because you've moved seamlessly from a legitimate complaint into random bashing. And well specified doesn't mean inflexible. Your anti-example of TCP(/IP) is in fact extremely well specified to the point where many independent implementations seamlessly interoperate, yet it is very flexible.

What Wayland is trying to achieve is exactly the opposite. Make a display system by specifying the absolute minimum and leaving the client the flexibility to implement the remaining features as it sees fit.

So the idea is that Wayland merely provides the mechanism allowing things to be done, but not the policy on how specifically to do them?

This is a bad thing if you want 100% certainty of functionality however limited it may be or interoperability, however it's fantastic for systems where requirements change as technology advances, such as the desktop.

I'm not sure I follow but it sounds like you're saying that Wayland is sacraficing interoperability for rapidly chasing the latest things? That sounds like a massively dumbass idea (imagine a system where GUI programs didn't interoperate! It would be like taking the desktop back to 1995, or using Android). I also don't believe that's the case with Wayland. At least I hope not.

Re:Clipboards?

[Bengie]

Uh, TCP headers ARE fixed length.

Actually, they are variable length for TCP, and you must pad the TCP header to a 32bit boundary if you decide to add extra fields.

Options (Variable 0–320 bits, divisible by 32): The length of this [header] field is determined by the data offset field.

IPv4 also has a variable header

The second field (4 bits) is the Internet Header Length (IHL), which is the number of 32-bit words in the header. Since an IPv4 header may contain a variable number of options, this field specifies the size of the header

IPv6 actually has a fixed header, but it can have a reference to data to extend the header, but that "extended" information is not part of the actual header.

Re:Clipboards?

[thegarbz]

s/TCP Headers/MTU Size. The point I make is the same yet you seem to grasp only the equivalent of spelling mistakes to make your counter argument.

Over specification is a problem for something as fluid as display technology. I'm not talking on the GUI level but on a system interaction level. Just have a read about the Xinput multitouch extension and the problems in creating it. In the past three years we've gone from mouse to multi-touch inputs as an emerging computer trend. Yet due to over specification X11 is still having trouble dealing with multi-monitor support out of the box something that hasn't bee a problem in other systems since the turn of the century. Multitouch support in X11 can even now really only be considered experimental, i.e. you can experiment on which distributions you MAY get it to work. Everything since the year of the lLnux on desktop (LOL) you CAN get it to work, but you shouldn't need to. It should just happen. Every time I attempt to use linux on a desktop it seems some things are broken, need tweaking, and all in all it's a few years behind and just an effort to use. Similar to Pulseaudio, yes it's broken, but at least it provided some functionality which simply didn't happen before. Try getting a bluetooth speaker to work without it.

Back to the screen saver point. The screensaver / lockscreen grabs all the keys. On the rare occasion when it works is due to some screensaver having a nasty workaround hack. That is exactly the problem. It shouldn't be up to a program to lock out other programs because it didn't hack its way around some limitation in the protocol.

Underspecification has it's problems too, but it leaves a lot more scope to implement changes.

Re:Clipboards?

[dbIII]

Well, there's nothing in X that stops one having the WM run the screensaver directly

For example that's the way E17 does it.

Re:Clipboards?

[serviscope_minor]

The point I make is the same yet you seem to grasp only the equivalent of spelling mistakes to make your counter argument.

No. You were insisting that well specified is the same as inflexible. This is not the case. IP and TCP are extremely well specified, for example and yet have done very well.

Over specification is a problem for something as fluid as display technology.

OK, fine. What's overspecified about X11?

Just have a read about the Xinput multitouch extension and the problems in creating it. In the past three years we've gone from mouse to multi-touch inputs as an emerging computer trend.

Uh? Doing multitouch right is hard. But, this has nothing to do with overspecification. Yeah, I read about it too. X11 is actually designed (not perfectly, but it's there) to have almost arbitrary extensions, you know. The prolem was coming up with the mechanism, choice of events and design of the API in a way that would be useful. i.e. the usual problems with software design.

ut that's not just "X and multitouch" you're talking about. You're actually talking about a specific API. If you wanted multitouch before, there were still raw calls to get the state of the touch pad, and ad-hoc multitouch could be implemented that way. I know this because I've done it in the past.

Yet due to over specification X11 is still having trouble dealing with multi-monitor support out of the box something that hasn't bee a problem in other systems since the turn of the century.

lolwot? X was quite late to the table with a user friendly solution, though even in the early 2000s it was actualy technically better at multimonitor than a number of other popular poerating sytems. It sucked to configure and had some issing features, but it was generally superior in spanning across multiple disparate graphics devices from different manufacturers.

So, tell me, what is wrong with multimonitor support at the moment? I haven't had trouble in a while. You can use any xrandr based tool to switch on and off external screens and rearrange them all at will. Even nvidia finally got with the program a few years ago and allow the whole shebang to controlled with xrandr.

But anyway, that's besides the point. That has nothing whatsoever to do with overspecification. Old ones, i.e. ones before xrandr was developed still work unmodified. New programs can make the new API/protocol calls. So, this has nothing to do with your point.

Every time I attempt to use linux on a desktop it seems some things are broken

Are you talking about "X being overspecified" or general problems with the Linux desktop now? But whatever, I've yet to encounter a system which doesn't need tweaking to make it usable. Linux requires the least for me and takes the tweaking better too.

Back to the screen saver point. The screensaver / lockscreen grabs all the keys. On the rare occasion when it works is due to some screensaver having a nasty workaround hack. That is exactly the problem. It shouldn't be up to a program to lock out other programs because it didn't hack its way around some limitation in the protocol.

Ah the old "limitation in the protocol" canard. As I pointed out and you ignored, if you imlement the screensaver in the WM instead, these things are not problems and it's semantically equivalent to implementing it in the compositor. As dbIII pointed out, Enlightnment does this.

Re:Clipboards?

[serviscope_minor]

For example that's the way E17 does it.

Interesting. I always knew it was possible. I didn't know anyone had actually tried it in practice. Good to hear the E17 nutters are at it again---and I mean that in a good way :)

Due to concept that is unlikely

[dbIII]

The cases where people have argued such a thing are the strawman you've taken up on as streaming bitmaps, which is not always (in fact very rarely outside of gnome3) the case. So there's equivalent performance (streaming bitmaps the same speed in both) or better when you have a situation where one can do something other than just stream bitmaps. Of course every time this gets mentioned we get the "only dinosaurs want remote access/shaped windows/whatever feature of X does not apply on phones - then the distraction - hey look how slow gedit starts on X so obviously X is crap and not gedit" so this discussion usually ends up at a dead end.
We've already got VNC and Wayland is not planning anything better remotely so we may as well focus on what it gives us as a local framebuffer, then screenscrape as best as we can later. With a dumb framebuffer the plan is to trade complexity and flexibility for speed.

Has it got network transparency?

[martin-boundary]

No? Then stop spamming slashdot with this "news". It's not news until there's working network transparency.

Re:Has it got network transparency?

[thegarbz]

No, what IS news is self-righteous experts wanting a feature they don't use while not knowing an equivalent feature has already been merged into the compositor. I mean fuck it's not like there was a Slashdot notice about remote desktop support being added to Weston last year sometime. Oh wait ....

Please take your bitching elsewhere.

Re:Has it got network transparency?

[jones_supa]

No? Then stop spamming slashdot with this "news". It's not news until there's working network transparency.

Well, it happens to have network transparency already, so I guess that's a green light to continue spamming with Wayland news.

Re:Has it got network transparency?

[Anonymous Coward]

I'll tell you what *would* be news - if someone posted an actual legitimate use-case for a network-transparent desktop app that wasn't a bodge to get around some archaic CAM app that management would probably replace if your time was actually worth anything.

Re:Has it got network transparency?

[TheDarkMaster]

I can't understand why network transparency is something SO important, especially when a lot of ugly hacks are necessary to make it work and compromising the performance of much more important parts like the presentation of local desktop.

Re:Has it got network transparency?

[thegarbz]

It's not. People heard the fancy term "transparency" and went oooooooh I'll use that without realising that network transparency does not mean the ability to display an app on a remote desktop. The same people also think that the very specific term "network transparency" which has a very specific meaning still applies. It doesn't. Most Linux desktops have lacked network transparency since the mid 90s in favour of nasty fallback hacks and rendering in different ways depending on the target server.

Modern X11 over the network is nothing more than a slower implementation of something like VNC except it doesn't support even basic things like compression. It is one of the worst performing remote desktop solutions there is.

Re:Has it got network transparency?

[Parker Lewis]

Because this is Slashdot, and people always need something to complain. Wayland will be clean, fast, and, more important, maintenainble. But, OMG, if it don't offer remote desktop in its bones, it'll be crap.

Re:Has it got network transparency?

[Anonymous Coward]

I can't understand why network transparency is something SO important, especially when a lot of ugly hacks are necessary to make it work and compromising the performance of much more important parts like the presentation of local desktop.

Probably because you're a retard who just uses Linux to watch Youtube kitten videos.

Right here on my desktop, I have two copies of Eclipse running from a VM using X forwarding, because that VM runs the old version of Linux that our old servers have been using for five years, while my development machine runs the new version that our new servers run. Do you really think I want to do that via some crappy VNC crap that has to resend half the screen every time I scroll down a line?

Why, oh, why are the Shiny Kids so desperate to remove the features that have made Linux a much less crappy OS than Windows?

Re:Has it got network transparency?

[Anonymous Coward]

It's because the Shiny Kids are smarter than you. They recognise that the solution to your problem is to make it so you don't have to run 2 different versions of your IDE, rather than lug a load of X11 legacy bullshit around for decades so you you can continue with your bodged-up edge use-cases.

Re:Has it got network transparency?

[PPH]

Most Linux desktops have lacked network transparency since the mid 90s in favour of nasty fallback hacks and rendering in different ways depending on the target server.

Nope. I've used Linux desktops since the 90s with X clients running on numerous platforms. Many of which were developed not even knowing what Linux was. So the need for server specific 'hacks' could never have been satisfied. And Linux works just fine, thanks.

Re:Has it got network transparency?

[TheDarkMaster]

Let me see... At least for me, the big problem with the network transparency is that to get it working, the X server needs to follow a number of architectural decisions that are highly detrimental to the performance of a local desktop and you can not work around this without a lot of (broken or unstable) hacks. That said, how many users use the transparency and how many users needs an efficient local desktop? Why I should hurt the performance of the entire server to meet a situation that is used for only 2% of the total userbase?

Re:Has it got network transparency?

[Uecker]

Ah, the "X11 is not network transparent anymore" FUD. How is this moderated insightful? I use it everyday - between different machines with different versions of UNIX or Linux and it just works. People want to break this for no good reason, and this is really really sad.

And yes, applications which do not use direct rendering (i.e. 99%) do render in exactly the same way (using XRender).

Re:Has it got network transparency?

[PPH]

People want to break this for no good reason,

Oh, they've got a good reason. Marketing wants to reduce the expectations of the users. 20 years ago, I could run video over networked X and run a remote CATIA session on my Linux desktop. So why can't you run your precious Auotocad or Adobe suite apps over X? Per seat licenses.

20 years ago, when Microsoft was a joke in the engineering world, things worked fine. And then they (and other vendors, no fair picking on only MS) tried to convince management that every seat needed an office productivity suite. Some poor slobs ended up with two systems on their desk, or dual booted Windows/Linux. And then our IT department got smart. Citrix offered a product that could export a Windows session from an NT server as an X client. So the few times I needed to run Word or Excel, I could just start a Windows desktop as a single client on my Linux system. Our IT people loved it. MS systems could be administrated at the central server location. Management loved it. For most of us (~5000 engineers) aproximately 1 license was needed per 10 people. But Microsoft (and others) shit themselves because it cut into their sales.

For all the persuasion end users are getting to go Wayland, multiply that by 10 to see the pressure app developers will be under to drop X and go with the local only solution. One license per seat. And once X is forgotten about, management won't have an example to hold up and ask why vernors can't come up with a more cost effective solution.

For most of the neck-beards playing video games in their parents' basement, it doesn't matter. But this is where I think Linux distros need to fork and provide one platform for the kiddies and one for business.

Re:Has it got network transparency?

[Blaskowicz]

What's funny is we've been slowly getting into a situation where all PC are fast so we can afford the waste of using X even more.
On the other hand we're now down to three graphics vendors and the drivers are improving.. but at lot of time using the GPU for the GUI will result in less stability, potential overheating or lock up, and instead of Xorg using your CPU it will be a graphics driver and its OpenGL implementation.
All so that a fraction of the userbase can look at windows flying around and zooming in/out etc.

I don't want to buy a new graphics card to read text, view movies and play minesweeper.

Re:Has it got network transparency?

[thegarbz]

And Linux works just fine, thanks.

For you. Didn't you hear? Last year was the year of Linux on desktop, and it doesn't work very well. If open source is to compete with proprietary then you need to actually provide functionality that users want. You know really complicated edge case crap like plugging in a projector and expecting it to work out of the box.

Re:Has it got network transparency?

[thegarbz]

Well it's not. As said network transparent has a specific meaning. You use it every day? So do I. But I'm under no delusions that the application is entirely transparent to the system its rendered on. Mind you it still just works, and it's quite right for you to expect it to just work in the future under Wayland.

But if it "just works" why would you care about the details of how the protocol works?

As for no good reason. You really should try some other form of remote application. There is a very good reason to do something differently from the slow archaic way that X11 does. Thinking otherwise is simply ignorance, accidental or willful.

Re:Has it got network transparency?

[martin-boundary]

Nobody cares about the local desktop screen performance any more. We've been living in a poor man's network transparent world for 10 years, and you don't even realize it. Nearly everybody now runs software on a machine half way around the world and displays its output locally, *very* slowly. It's called HTML and web apps. And those who work with computers instead of playing FPSes all day run software on beefy shared multicore machines over the local network, while wanting the output to show on the local machine. Using network transparency, so the programs don't have to be written in some god awful complicated way.

If the idiots take network transparency away from us, the only realistic option we'll have left is the even slower and totally ridiculous web app paradigm (yes, it truly is ridiculous - it's ten levels of hacks just to keep a bit of state on top of a stateless protocol).

Re:Has it got network transparency?

[martin-boundary]

My theory is that now that Microsoft screwed the pooch with the latest Windows, all the Windows weenies who no longer have a usable system are coming over to Linux wanting to recreate what they lost. It's an invasion of the barbarians, basically.

Re:Has it got network transparency?

[TheDarkMaster]

Wrong, you NEED to care. Because the best way to do work on a desktop is still the old-fashioned local application, ideally with old-style compiled aplications. Truth, you have to go through the step of install the application locally on the desktop, but in return you have all the resources you need at the maximum speed that the hardware is capable, without lags caused by bad connections (connections available 100% of the time and gigabit speed only exist in the first world, not here) or bloatware caused by a crapload of hacks to turn HTML into a "desktop".

That does not sound like a real example

[dbIII]

Basically all that. Even over GigE simple things like gvim are a dog.

To me that sounds like complete and utter bullshit unless gvim is now seriously broken. In my workplace complex interactive geophysical packages with a lot of graphical information are used remotely over X by dozens of people at once to (in some cases) substandard MS Windows implementations of X without running like a dog - even over wireless to laptops, so how is your gvim over GigE example even possible unless somethign else is going on? It appears to fail the reality test. Did you make it up or was the machine you were running it on under very heavy load at the time so it would be slow in all cases? If you made it up - why - what is motivating you to make such things up about what you see as opposition instead of praising what you see as good in Wayland? This X sux rubbish that fails the reality test is annoying and doesn't do Wayland any good while Wayland is still making progress.

No GTK2

[Blaskowicz]

That's nice but what you describe is for GTK3, and not GTK2. Seems like the latter is still used a lot, and frankly GTK3 has gone rogue, deleting features, adding ones only Gnome developers will use etc.
Developers of applications run away from it and migrations from GTK2 to GTK3 seldom made (though there are dual mode GTK2/GTK3 applications where you can select the UI).

Recently with GTK 3.10 they removed icons in menus and the highlighting of letters to help you with keyboard navigation (e.g. Alt-F opens File menu). It's the Slashdot Beta of the toolkit world.

xload

[Anonymous Coward]

xload! xload xload xload! I R serious admin and i need ze lxoad

Fuck remote display

[ArchieBunker]

So use VNC if you need a remote display. This need to keep 30 year old features unchanged has got to stop. 99% of the people using the GUI are running it locally.

Wayland is the wrong place for remote transparency

[John Allsup]

The most common use case today is local applications. This must be optimised for. Have a separate server and protocol to network transparency for the classes of applications that network transparency is useful for (simple GUIs, text editors and suchlike, rather than nonlinear video editors and 3D games). Likewise with audio, there is a need for a simple high performance backend for some applications, and network friendliness for others. In both cases there should be two layers, a fast light low level backend and a network transparent application layer for applications that want to use it.

I hate FPs like this.

[Anonymous Coward]

This is slashdot, it assumes (correctly) a certain level of knowledge. If you don't know what Wayland is, get off your ass and search google.

The X11 SPoF is dead, long live the Wayland SPoF!

[Anonymous Coward]

The only major engineering fault with X11 is that the X server is a Single Point of Failure --- whenever it hangs or dies, which is not uncommon owing to the unstable state of graphics card drivers, all your running X11 applications collapse and terminate, often many hundreds of them at once on a powerful destop.

X11 badly needed redesign or replacement to eliminate this SPoF. Did the Wayland team make avoiding this single most important X11 design fault a priority for their X11 replacement? Of course not. Their eyes are not on the need for robust engineering, but entirely on eye candy.

A huge opportunity for improving the engineering quality of FOSS graphic systems has been missed.

Raspberry Pi

[Anonymous Coward]

This is VERY good news for people who want a video-accelerated gui on their Raspberry Pi. Wayland supports the video chipset on it - wheras X does not (and is horribly slow!)

How do I use it ?

[frankie_guasch]

I managed to compile and install it . But I can't find instructions of how to run it. I was thinking I should shutdown the dm and then start it somehow. Release 1.5 looks like a production name, am I missing something ? Any links to the docs or howtos ?