User: Password:
|
|
Subscribe / Log in / New account

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 8:31 UTC (Sun) by marcH (subscriber, #57642)
Parent article: The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Disclaimer: I do not understand anything about any technical details.

What I know is: at work we are supposed to use VNC but many of us use Xming instead (or Cygwin X).

The VNC / Remote Desktop / "desktop over desktop" user interface is completely broken. Two window managers on my screens? No thank you.

I've used remote X for as long as I've been using Unix. I do know how this works but it does work. I'm happy to believe that X11 is an unmaintainable engineer's nightmare fixed by Wayland. But if Wayland can't do remote X in some way then it's completely useless to every work place I've been to, big and small. And no I don't care a bit about fancy 3D graphics.


(Log in to post comments)

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 9:36 UTC (Sun) by Frej (subscriber, #4165) [Link]

I've haven't used X/linux in years.. but it's quite simple to understand?
How do you come to the issue that must have multiple window managers? You you just assume it, and then you even state you don't know, and complain anyway. Not cool ;) Think of VNC as the abstract method of sending images vs. a protocol of drawing commands (Core X). Why does either method limit you of creating anything you need? Free your mind :-)

PS: You are probable using a crappy implementation of VNC (as in method) in remote X either way. As the drawing primitives by toolkits today consists of mostly sending images.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 18, 2013 11:15 UTC (Tue) by nix (subscriber, #2304) [Link]

I think of VNC as 'incredibly slow'. I have never used any implementation of VNC nor whatever the MS Terminal Server protocol is called that could scroll an Emacs window with text in it at more than about three to five screens per second, with a noticeable lag -- over a local network! And this applies to scrolls by one line as much as to scrolls by whole pages, which is a bit tough if you want to scroll up five or six lines, line-by-line, as I do quite often. Meanwhile under X, even remote X (over an Ethernet LAN) scrolling is latency-free: it scrolls so fast it is a blur.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 12:27 UTC (Sun) by cmrx64 (subscriber, #89304) [Link]

RDP and Spice both support "seamless" mode which has windows independantly floating around, not desktop-on-desktop.

As Eric says, nothing in Wayland precludes remoting equal to or better than X (it'd be better, because of the protocol's semantics). But nobody has implemented it yet, because they're working on other things.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 14:25 UTC (Sun) by mebrown (subscriber, #7960) [Link]

Why would you continue to make this criticism of Wayland when literally every single article (including the linked one) specifically says that they are working towards a solution for VNC/remoting, that the solutions are more technically elegant, and that there are already prototypes?

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 16:20 UTC (Sun) by marcH (subscriber, #57642) [Link]

> Why would you continue to make this criticism of Wayland...

You are assuming that merely stating a (non-negotiable) requirement is a critic.

> they are working towards a solution for VNC/remoting, that the solutions are more technically elegant, and that there are already prototypes?

Here it comes again: the good, old and infamous "more technically elegant prototype". Thanks for answering your question yourself.

I have an old & ugly product that just works. Why the hell would I switch to an elegant prototype? I'm not an X11 engineer, I'm just an X11 LUSER. God preserves my favourite distro from switching.

I feel really sorry for the Wayland developers that X11 raised the user expectations that high.

I mean, come on: we are talking about an important feature here and not asking for some obscure and crazy backward compatibility hack
http://www.joelonsoftware.com/articles/APIWar.html

I like LWN. I pay for it. I love the technically accurate and insightful articles *and comments*. The only thing that I find sometimes annoying is the inability of a large number of people in this crowd here to think outside the engineer box and put themselves in the shoes of a plain luser who wants important features not to break and to Just Keep Working.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 16:39 UTC (Sun) by Cyberax (✭ supporter ✭, #52523) [Link]

> I feel really sorry for the Wayland developers that X11 raised the user expectations that high.
In many cases Wayland developers _are_ X11 developers.

>I mean, come on: we are talking about an important feature here and not asking for some obscure and crazy backward compatibility hack
http://www.joelonsoftware.com/articles/APIWar.html
You're wrong on so many levels...

First, the old style X11 is not going away. It's going to be supported on Wayland for any foreseeable future, so you can "ssh -X" to your heart's delight in the next 10-something years.

Second, Wayland is not just something out of the blue. It's a logical evolution of real-world X11 usage patterns. It solves very real problems with X11.

Third, trying to preserve X11 will continue to make it more and more complicated and hard to support. Which will (at some point) make it impossible to add new features.

> put themselves in the shoes of a plain luser who wants important features not to break and to Just Keep Working.
They have. So X11 won't be broken. Your apps won't go away. Keep calm and continue working.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 21:05 UTC (Sun) by dig (guest, #91108) [Link]

Too many promises... X isn't easy peace of software, especially the one to emulate it on totally different concepts Wayland uses. Only the future will tell...

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 21:08 UTC (Sun) by Cyberax (✭ supporter ✭, #52523) [Link]

X on Wayland is easy. You just run a rootless server with a Wayland backend. Exactly as you would do it on Windows and Mac OS.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 11, 2013 8:28 UTC (Tue) by mmarq (guest, #2332) [Link]

So what is the point of Wayland ?

IF a dev codes for X, and he continues, wouldn't be much better to run it directly on X instead of a server on top of another ?

And yes i read all the arguments, but i see 2 trends developing for applications. One is Kronos the other is HSA, and they are some how very close related in many points.

Even Nvidia will eventually join(both) with their ARM push, and the same with Apple(LLVM is the dev base of HSA and Apple is LLVM). More dubious is Intel and Microsoft. The strength of X is not what lays ahead but the enormous baggage that lays behind... Why change ?... in a pure app POV what isn't broke don't fix it, the DS is *orthogonal*... so again why change ?

Is any of those DS , X, Wayland or Mir, going in the direction of the industry ?

(if i was an app dev right now, i would want it on Mac, Linux, Android and Windows etc... if possible... all this discussions and promises don't give me any incentive)

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 11, 2013 17:18 UTC (Tue) by Tobu (subscriber, #24111) [Link]

Devs code for toolkits: SDL, Gtk, Qt, or even more high-level. One of the toolkits' jobs is to provide portability (though apps will use platform calls when the toolkits are weak on that front). I don't think a brand-new Kronos standard will displace them.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 22:04 UTC (Sun) by daniels (subscriber, #16193) [Link]

> X isn't easy peace of software

I know. And the only thing harder than doing X is continuing to shoehorn new features into it as it goes along.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 15:08 UTC (Mon) by drag (subscriber, #31333) [Link]

> Too many promises... X isn't easy peace of software, especially the one to emulate it on totally different concepts Wayland uses. Only the future will tell...

You don't know much about X then.

X11 runs perfectly well on many operating systems besides Linux. Windows and OS X are both examples that are supported by Xorg's Xservers. In fact one of the biggest contributors to Xorg is Apple and they use that code directly in their X11 related Apple products.

I find it extremely unlikely that Wayland will be much harder to support then Windows or OS X is.

In fact how Wayland is designed it should simplify the functionality of Xservers for Linux massively.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 15:33 UTC (Mon) by alankila (guest, #47141) [Link]

Hm. What X11 related products do Apple have now? They aren't even shipping an X server anymore.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 16:26 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link]

Apple ships X server as a separate download. I'm using it extensively to run wireshark.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 16:46 UTC (Mon) by drag (subscriber, #31333) [Link]

Yep, http://support.apple.com/kb/ht5293

There isn't many people clamoring for X11 support in any OS except Linux (or the BSDs), but it does exist.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 11, 2013 22:49 UTC (Tue) by marcH (subscriber, #57642) [Link]

X servers with few applications, how come?

Because these servers are useful to run applications remotely of course.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 17:32 UTC (Sun) by tpo (subscriber, #25713) [Link]

> God preserves my favourite distro from switching

:-))) ! Thanks,
*t

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 9, 2013 21:16 UTC (Sun) by Mook (guest, #71173) [Link]

> Why would you continue to make this criticism of Wayland when literally every single article (including the linked one) specifically says that they are working towards a solution for VNC/remoting, that the solutions are more technically elegant, and that there are already prototypes?

Pretty much it keeps being described as a solution for VNC. From a user's point of view, VNC (/RFB) is horrible in that it is presented as a giant rectangular hole in which things show up, completely disconnected from the environment around it. In contrast, X can be forwarded without a root window so that things integrate well. If people start consistently describing the future Wayland solution without comparing it to VNC (even if it does do pixel scraping instead of sending drawing commands over), people like me will stop hating on it. As far as I can tell, this is the direction it's actually going anyway; it's just been a horrible communication problem because it's actually more VNC-like in a way that isn't quite as important to the user.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 1:21 UTC (Mon) by jzbiciak (subscriber, #5246) [Link]

Yeah, I cringe whenever I see the parallel to VNC brought up, because, as you point out, it makes all the wrong associations.

The VNC user experience sucks. And by that, I mean the end-to-end user experience of someone who fires up a VNC client and VNC server overtop of the existing windowing environments. The one where you get a window that behaves as a (crappy) monitor and doesn't integrate well at all with a desktop environment. (For example, getting tons of spurious "Cannot empty clipboard" errors in Excel on my Windows box whenever I leave something in the selection buffer on my remote VNC'd X connection.) Wayland, so far as I've read and understood it, does not seek to replicate that in any way.

Using communication model with async, timestamped requests and a focus on shipping bitmaps for windows instead of drawing primitives seems like a reasonable model these days for shipping a window from one machine to another. This model bears some resemblance to how VNC forwards an entire screen from one end to another, and keyboard/mouse inputs back the other way, but as I understand it, Wayland will just do this for individual windows.

Sounds entirely reasonable to me, and perhaps provide the best of both. Also sounds like it's less likely to hit snags, such as my fonts disagreeing between two machines leading to software looking weird.

I'm personally avoiding getting too excited (positive or negative) about Wayland until it gets to a point where it's worth it for me to give it a spin around the block, and that means finding it in a distro somewhere. I'm cautiously optimistic. I'm a little skeptical about some things, but I'm reserving judgment until later.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 9:16 UTC (Mon) by blackwood (subscriber, #44174) [Link]

Yeah, that's it. Wayland networking will combine the sound network protocol ideas of VNC (having a local server for the remote client and a local window on the local display server and forwarding bitmaps/input events asynchronously) with the nice integration of forwarding an X client. The end result is pretty impressive.

What's holding things up is simply that other stuff is currently a higher priority (like plugging the holes in the input layer for input methods and fleshing out the window display mode a bit). But I guess it'll all be in place when desktops environments have working Wayland code, together with XWayland and all the other pieces we need to have for a well-working full-fledged Wayland desktop.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 18, 2013 11:20 UTC (Tue) by nix (subscriber, #2304) [Link]

The end result is pretty impressive.
Except if you want to scroll windows full of text, since that means repainting the entire screen rather than erasing a line from the top, shuffling the rest up, and painting a line at the bottom (or vice versa), since it has no understanding of the semantics of scrolling.

I am told that in order to get such a rare and obscure use case to work I have to wait for toolkit-level remoting, which is as far as I can tell a complete mirage: nobody is working on it, nobody is planning to work on it, if people do work on it their work for distinct toolkits will be totally uncoordinated (natch), why aren't you happy with VNC-style bitmap shuffling nobody needs to scroll windows full of text anyway.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 19, 2013 0:14 UTC (Wed) by daniels (subscriber, #16193) [Link]

> Except if you want to scroll windows full of text, since that means repainting the entire screen rather than erasing a line from the top, shuffling the rest up, and painting a line at the bottom (or vice versa), since it has no understanding of the semantics of scrolling.

Again, this is completely false.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 13:21 UTC (Mon) by drag (subscriber, #31333) [Link]

> The VNC / Remote Desktop / "desktop over desktop" user interface is completely broken. Two window managers on my screens? No thank you.

There is absolutely no reason why you can't have individual applications remoted. If people want Wayland to be configured to be able to send individual display buffers over the network rather then a composited desktop then that should be very possible.

> I've used remote X for as long as I've been using Unix. I do know how this works but it does work.

It works sorta.

For your specific use case, really.

At my work we use a wide variety of remote desktop and remote application using Citrix and Windows and the idea that Linux/X11 is competitive with what you can accomplish using something like Microsoft Windows is ludicrous.

> I'm happy to believe that X11 is an unmaintainable engineer's nightmare fixed by Wayland. But if Wayland can't do remote X in some way then it's completely useless to every work place I've been to, big and small. And no I don't care a bit about fancy 3D graphics.

X11 can't even do what you _BELIEVE_ X11 can do. For most applications you really end up just shoving huge textures over the network similar to what people do with VNC, only much worse.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 14:35 UTC (Mon) by dps (subscriber, #5725) [Link]

Fortunately X can do better than that---I have run 3D X clients using *faster* on a remote server with better 3D acceleration (and more CPU and faster memory). Any ideas of making the remote client do software rendering on the slower remote system would be a huge step in the wrong direction.

NX takes doing work on the *server* side much further: it does extensive server side caching and has amazing performance other links with high latency and low bandwidth.

X has its faults but storing things on the server so things like cut paste between clients running on different boxes work is not one of them. Nor is allowing clients to actually use the acceleration available on the server.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 14:51 UTC (Mon) by mjg59 (subscriber, #23239) [Link]

X provides very little remote support for any OpenGL feature past 1.5, and nobody seems especially interested in fixing that.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 18, 2013 11:22 UTC (Tue) by nix (subscriber, #2304) [Link]

Actually there have been suggestions (from Daniel Stone, IIRC) that might well get indirect GL working much much better with much less development overhead (IIRC the idea was to fall back to bitmap shuffling in situations in which you currently abort: yes, it's slower than hardware acceleration but at least it *works* and lets you do simple 3D stuff remotely which is just using it to do analogues of 2D-analogous stuff).

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 19, 2013 0:12 UTC (Wed) by daniels (subscriber, #16193) [Link]

I suggested to do the rendering locally. Very few GL-using apps (apparently I need to disclaim this now, so let me be clear that I don't think they're marginal or pointless; just not the majority) these days have heavier geometry usage than textures. It comes down to Blender/Maya/etc, and CAD apps really. At that point, bandwidth-wise, you end up sending more over the wire than you would just sending the final post-composed image.

Then the more significant hurdle is updating GLX beyond GL 1.5, which as others have noted, is a hell of a lot of work.

The Wayland Situation: Facts About X vs. Wayland (Phoronix)

Posted Jun 10, 2013 14:58 UTC (Mon) by drag (subscriber, #31333) [Link]

> Fortunately X can do better than that---I have run 3D X clients using *faster* on a remote server with better 3D acceleration (and more CPU and faster memory). Any ideas of making the remote client do software rendering on the slower remote system would be a huge step in the wrong direction.

When the ability to get accelerated indirect rendering working on X I experimented with playing games over X11 protocol remotely.

It worked very well graphic-wise. But the controls were very laggy. So while the graphics rendered fast enough, the controls make it relatively unusable except under very ideal situations (ie: from one side of a lan to another)


Copyright © 2017, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds