Wayland Preparing For 1.0 Stable Release

[phred14]

Try nx to see how inefficient the default X forwarding is. Yes, nx is considered an ugly hack but it shows that there is room for much improvement. And so far I would just wait what the wayland hackers can do...

Can you furnish a pointer to instructions to set up free NX? I'm in a situation where it needs to be unencumbered, and while I get the impression that that's possible, I've found a "twisty, turny path of directions, all different." I used to use dxpc back in the day, but that has withered, and I understand that that was one of the starting points for NX. I recently tried to dust it off. The source would build, but it crashed before doing anything useful. I've been interested in NX for some time, but never taken the plunge.

[johnc]

Agreed, we need to embrace new technology. I wish X11 would go away already, it's a piece of crap that has been slowing us down for a long time.

+1 for Wayland, Linux and systemd.

If you think that Wayland is going to usher in the mythical "year of the Linux desktop", or that X is even a top 10 hurdle that's holding Linux back from widespread adoption, you are seriously out of touch with reality.

What I do know is that where Linux and UNIX environments have prominent roles -- such as businesses that actually need distributed computing -- well-executed network transparency is a non-negotiable. Web apps and VNC-like do not cut the mustard. So as long as Wayland can step up to the plate, it will be embraced in the areas where Linux actually matters. Now if you're the type of person who's only vision of Linux displays is twirling desktop cubes FTW and posting videos of it to YouTube, then please, stop. Please stop now. We don't need to be cheer-led by people who have such a narrow field of view.

[phred14]

What I do know is that where Linux and UNIX environments have prominent roles -- such as businesses that actually need distributed computing -- well-executed network transparency is a non-negotiable. Web apps and VNC-like do not cut the mustard. So as long as Wayland can step up to the plate, it will be embraced in the areas where Linux actually matters.

I think that was the flavor I was trying to give in my post. I routinely use X11's network transparency at work, at home, and between work and home. If Wayland + whatever can fill my needs, great. But I think I've pretty well defined my little corner of the world in terms of network transparency, and I don't think it's outrageous. It all works and works well today. I think there's a pretty heavy burden of proof on anyone who tells me that I should give up what I'm doing. The layers under can change, but the net operation is needed.

[johnc]

This.

Screen-scraping is never efficient. Perhaps that's what johnc meant by having no design for network transparency; sending frames, server-rendered, is a terrible bubble-gum solution because the protocol does not allow for more efficient methods. Because it was not designed to do such.

I missed this earlier but yes, yes, a thousand times yes. There may be times when server- (remote-)rendering is the preferred option, but in many cases you actually want the local box doing the heavy work. It's not an issue of network bandwidth. It's a matter of having several persons on a single box hitting a single (or dual... or quad...) CPU and GPU. That just doesn't work. You need to distribute that workload out over several client boxes. Looking back on my previous business environment experiences, if we (as software developers) did not have that kind of functionality we would have been screwed. We might as well have used Citrix on Windows at that point.

But it's okay... we can relax. We have web apps now.

[moonlite]

If you think that Wayland is going to usher in the mythical "year of the Linux desktop", or that X is even a top 10 hurdle that's holding Linux back from widespread adoption, you are seriously out of touch with reality.

Now if you're the type of person who's only vision of Linux displays is twirling desktop cubes FTW and posting videos of it to YouTube, then please, stop. Please stop now. We don't need to be cheer-led by people who have such a narrow field of view.

But it's okay... we can relax. We have web apps now.

What's with you and all the strawmans? This is (or rather could have been) a friendly discussion.

[johnc]

What's with you and all the strawmans? This is (or rather could have been) a friendly discussion.

What strawman? We have the guy who thinks that just using web apps and VNC should be enough to satisfy the "tards" that believe network transparency is actually useful. It's a fair point to criticize that idea as absurd, because it is absurd.

And then we have the guy who wishes X was dead and buried long ago because it's "holding Linux back". And he hopes we can push the proprietary drivers right out of the picture too. Because, after all, he's successfully using nouveau now, so who could possibly need the proprietary drivers? Where's the strawman? He clearly has a severely limited view of Linux uses and why X has actually been a very important part of Linux / UNIX success. "Holding Linux back" from what? Playing Mahjongg at 120fps? X is hardly a long pole in the tent.

[Nobu]

Say you're running a Wayland desktop (no X); what's wrong with something like this:

Xephyr :1 &
DISPLAY=:1 someapp &

...start your X forwarding solution of choice, then connect remotely?

When a native solution is available, it could be as simple as this:

cust_wforwarding_comp -- someapp

...then connect remotely. If you want to forward the whole display, you just need to attach to the main compositor and forward events and regions over the network, right?

[Scias]

Really, there's no point arguing about network transparency flaws and such in Wayland...

It's not like all apps, drivers and toolkits were ready to drop X11 support the day Wayland is going to be stable, and it's really not like that companies (who need this feature the most) will switch on it in the foreseeable future...
So, regular users will be able to get the very nice progresses of Wayland over X11, because let's face it (excluding network translucency) X11 is horrible concerning modern Desktop performance.
And in the meanwhile, nothing prevents users/companies that rely on network translucency to still use X11. Like I said, Wayland won't magically replace X11 from one day to another, so there's no need to cry about it really...

[leeenux]

I'll attempt to go for some light here, and turn down the heat. But first, such a nice, enlightening term, "network-transpara-tard" - good to know one of my many attributes.

I'm typing this on my lunch break at an F500 company, on my laptop running a company-tweaked RedHat 6.2. The laptop is beefy enough to do most of my job, but every now and then I need more. So on another desktop I'm ssh'ed into a beefier machine, with the DISPLAY exported back here, running the company-mandated VLSI CAD tool suite. (It runs on my laptop, but not the particular jobs I need to be spawning from it right now.) This is normal, for me and for my co-workers. Keep in mind that this is "industrial" software, and most of the place is running a tweaked RedHat 5.x, and the software was developed on something at least that old.

This weekend I'll likely do my usual Saturday maintenance on my 4 active Gentoo machines scattered around the house. For my convenience I'll sit at one machine and ssh into the other three, exporting the DISPLAY back. It's mostly text mode, but now and then I fire up a GUI text editor, use "diffuse" to update configuration files, and I like to use "xconfig" to configure my kernels. There are 4 tabs on my terminal, and once set up they all work alike.

Sometimes I work from home, and that takes 3 forms. Sometimes I run locally, including the aforementioned VLSI CAD suite. Sometimes I ssh into a machine at work and export the DISPLAY back. Sometimes (usually a weekday evening) I ssh into my laptop at work, start "x11vnc", and use vncviewer at home to get access to my complete work environment. There are different reasons to use all 3 methods of remote access. That's based on how long I'll be working from home in a session, how graphically intensive that session will be, etc.

All of this "just works" today, and I don't think there's anything "tard" about any of it. I'm getting a job done, getting paid for it, and I'm using the right tools.

How does Wayland fit into this?

I can see that in the future my home Gentoo machines would migrate to Wayland ahead of any corporate machines. Assuming that the corporate machines will be very late on any conversion, I guess I need answers to 2 questions:
1 - Will I be able to do my "Saturday Gentoo maintenance" with full user interface symmetry on all 4 machines - 1 local and 3 remote?
2 - Will I be able to use all 3 modes of work-at-home?
2a - Run legacy X11 applications, including but not limited to VLSI CAD suite locally?
2b - Run legacy X11 applications, including but not limited to VLSI CAD suite remotely, with some equivalent of exporting the DISPLAY back?
2c - Run something like x11vnc to get my active desktop at work displayed locally?

The only one I've received absolute assurance on is #2c, but earlier in this thread it's suggested that all of these can be made to work. Perhaps the most problematic is #1. I could see getting to the point of having native Wayland tools that I'd normally use, but is there some layer in them or that can be added to get the remote display? As mentioned, symmetry is the key, so that all 4 systems act alike as I do my maintenance.

Obviously, the Wayland folks aren't going to overlook the ability to graphically access a machine remotely, since VNC/RDP/etc... are so popular in business environments, but FFS, can we please abandon the idea that a window manager should have network transparency? It is 110% preferable to just stream the whole desktop ala VNC, LTSP, or maybe even a brand new protocol that's more efficient, rather than using all of these nifty little X11 over SSH tricks. Wayland will also probably be an excellent first step towards improving the performance of remote desktops on Linux, which still lag way behind the performance and smoothness of Windows RDP.

Wayland addresses a real need for 99% of users, and that is replacing the awful performance, esoteric API, and legacy bloat of X11, with something modern that's designed for the way people use computers now. People who need legacy features of X11 should just commit to doing things in a way that will actually be supported, or just resolve themselves to using old computers running old OS to do things the same way they were done in the 90s. Meanwhile all of the young whippersnappers will use that newfangled Wayland thing on their shiny new computers.

[liam]

Obviously, the Wayland folks aren't going to overlook the ability to graphically access a machine remotely, since VNC/RDP/etc... are so popular in business environments, but FFS, can we please abandon the idea that a window manager should have network transparency? It is 110% preferable to just stream the whole desktop ala VNC, LTSP, or maybe even a brand new protocol that's more efficient, rather than using all of these nifty little X11 over SSH tricks. Wayland will also probably be an excellent first step towards improving the performance of remote desktops on Linux, which still lag way behind the performance and smoothness of Windows RDP.

Wayland addresses a real need for 99% of users, and that is replacing the awful performance, esoteric API, and legacy bloat of X11, with something modern that's designed for the way people use computers now. People who need legacy features of X11 should just commit to doing things in a way that will actually be supported, or just resolve themselves to using old computers running old OS to do things the same way they were done in the 90s. Meanwhile all of the young whippersnappers will use that newfangled Wayland thing on their shiny new computers.

Remoting a gui is easy (SPICE will let you render on either end).
The reason that they haven't determined the exact solution to networking is b/c the two problems are unrelated. Wayland simply doesn't care. Frankly, using X for input,rendering,networking should send up a flag to any true believer in UNIX. In the UNIX tradition we shouod be writing small components that each do "one" thing. Wayland is the attempt to start doing this on the display side.
So when people say THEY haven't designed a networking solution yet thus we are all screwed, they (meaning the complainer) are, probably, missing the point. Networking and display are two different domains, and that has been factored in by the Wayland devs.