Red Hat Expecting X.Org To "Go Into Hard Maintenance Mode Fairly Quickly"

[thechef]

Wayland is in no state to replace Xorg yet, but it will be. Once it is better in all regards and isn't doing any wrong "most users dont care about xyz" assumptions anymore, it will do just fine. The numbers of issues are easily quantified and listed. Just go through the list and fix it all (as is happening already). I think about at the point where as much man years was spent on Wayland as was spent on Xorg and its ecosystem that will also be the turning point where the advantages of wayland will set in, but no earlier. The assumption that a better design will replace a legacy design all by itself is naive. Of course its all the work around that which matters.

[Sonadow]

That's a non-sequitur. Java IDE's can and often are used for developing server or web applications. They don't imply the presence of popular java consumer desktop applications. Java on the desktop is essentially dead. You haven't really provided any good counter examples for consumer desktop applications. Oracle's last try was JavaFx and they have abandoned that effort as well.

So what does this mean? That OpenJDK with its X11 dependencies for launching GUIs are going to be unusable when Red Hat stops maintaining X and lets it bitrot? My employer has a very liberal BYOD policy which allows me to use my own Linux desktops and laptops at work, and we have a large number of network appliances that use JNLP for their management. Will I no longer be able to use Oracle Java 9 or OpenJDK9 to manage them for the next five years?

I'm concerned about how the transition from X to Wayland will be as different distributions have various problems with Wayland. Here's an example with my self-compiled builds of the current Chromium, Libreoffice, Firefox and Thunderbird versions on different distributions:

OpenSuse 15.0
Firefox Wayland And Thunderbird Wayland launch on Gnome and Weston, with serious problems. Closing a dialogue box with 'Enter' or 'Esc' locks up the whole GUI, forcing a hard reset or a Sysrq reboot
Both won't launch in Plasma Wayland.
Libreoffice Wayland only works with the GTK plugin but is painfully slow. When used with the QT5 plugin, any document with more than 1 page of images (or more than two worksheets) freezes up the whole application. Only works best with SAL_USE_VCLPLUGIN=gen, which forces LO back to x11 mode.
Standard Chromium build works in X only. Ozone (Wayland) version of Chromium works in all desktops and Weston, but tends to crash after some time.

Debian 10
Thunderbird Wayland launches in Gnome Wayland, Plasma Wayland and Weston, mostly usable.
Firefox Wayland won't launch in any desktop environment or Weston. Immediate segfault.
Chromium and Libreoffice not built and not tested yet.

I don't mind if Red Hat decides to abandon X and let it bitrot, but XWayland is just a bandaid and if abandoning X = abandoing Xwayland I'm going to be righfully peeved because a significant portion of my software is going to be unusable.

[Guest]

Unfortunately, lots of people are still very insistent on running and using X, especially NVIDA fans. And Valve for some reason is working a lot on improvements to X rather than working on Wayland support.

[skeevy420]

Unfortunately, lots of people are still very insistent on running and using X, especially NVIDA fans. And Valve for some reason is working a lot on improvements to X rather than working on Wayland support.

As an AMD user, being insistent on running and using X applies to us too....KDE users as well. And probably Intel users. It applies to most of us; all of us who play proprietary games.

As for Valve, fix and improve what works for games now and not bothering with Wayland until it's more widely adopted just makes sense. I doubt that Valve will bother a whole lot with Wayland until Wine, Gnome, and KDE all offer a good OOTB Wayland experience (and likely with XWayland).

If a part of your companies' Linux gaming backend doesn't support Wayland and no major desktop works great with Wayland, Wayland isn't something a gaming-oriented company should focus on unless they're working on Wine and desktop environments. Basically, they all need a well working Wayland environment with general desktop usability so they can actually develop on Wayland so they can then focus on using that environment for a specialized task like gaming.

I say "they all" above because that applies to Valve, developers/publishers like Feral, Gnome and KDE and other desktop environments, Crossover & Wine, GPU driver devs, and a lot more. Until Gnome, KDE, and XFCE all get their shit in order so we and "they all" have proper Wayland development environments, Wayland adoption will just continue to be slow.

[stormcrow]

Unfortunately, lots of people are still very insistent on running and using X, especially NVIDA fans. And Valve for some reason is working a lot on improvements to X rather than working on Wayland support.

And that's going to continue to be the case because Wayland is only a "Thing" on Linux and even then it's problematic. X.org runs on every other Unix & Unix-like OS out there and Wayland has only given at best lukewarm reception to upstreaming ports to other operating systems. If you want things to work on more than just Linux (and even on Linux) then you target your UI for X.org if you're running native.

X.org is illustrative of what happens when a single entity has too much control over a core infrastructure project. This isn't about whether or not X.org is sustainable, this is about RedHat wanting to push it's singular vision of a Linux UI framework and just sitting on the competing project it wormed its way into a major position of authority. RedHat has been insidious in this manner over the years: Gnome, GTK/Glib, Wayland, X.org. With IBM essentially in control of RedHat, I predict this will only get worse as IBM has an order of magnitude more resources to sink into the embrace-and-take-over strategy.

[oiaohm]

And that's going to continue to be the case because Wayland is only a "Thing" on Linux and even then it's problematic. X.org runs on every other Unix & Unix-like OS out there and Wayland has only given at best lukewarm reception to upstreaming ports to other operating systems. If you want things to work on more than just Linux (and even on Linux) then you target your UI for X.org if you're running native..

Really X11 is not that native. You did mention Redhat and IBM. But where is the other Unix & Unix-like personal in x.org development.
https://www.x.org/wiki/BoardOfDirectors/ this is the board of directors. Notice all Linux people.
All the people who put there name forwards last 8 elections for the board have been all Linux people. There has not even been a single person from the BSD or commercial unix camp put their name in the ring.

https://www.x.org/wiki/Membership/ to be able to put you name in the running all you have todo is contribute code/documentation to come a member and put your name forwards.

X.org is not example of when a single entity gets too much control over a core infrastructure project at all. Its example of something core where most of the possible effected parties are not putting their hand up to take part in the management so the ones that do have to do the best they can and of course a bias to their own interests. This is more since it works it no our problem to make sure it goes the way we want.

Redhat is being very good declaring in advance how long they are planning to take care of x.org X11 server part. If it to stay alive the other parties need to put up resources. Leaving Redhat/IBM holding the maintenance bag why should they not do what is in their own best interest when they are the one paying.

[starshipeleven]

X.org runs on every other Unix & Unix-like OS out there

None of which is installed in any significant capacity on workstations, apart from MacOS where I suspect you won't be running Xorg anyway.

[Space Heater]

RedHat wanting to push it's singular vision of a Linux UI framework and just sitting on the competing project it wormed its way into a major position of authority. .

Ah yes, they "wormed" their way into authority. Being the biggest contributor/supporter of a project for years is certainly a good way to do that.

[RahulSundaram]

So what does this mean? That OpenJDK with its X11 dependencies for launching GUIs are going to be unusable when Red Hat stops maintaining X and lets it bitrot?

Both Openjdk and Xorg are open source projects. So it up to the community including other vendors how long Xorg is going to be maintained and how the dependencies are going to be handled at that point. If you are paying customer, let your support team know what issues you care about. If not, file bug reports or patches

[9Strike]

The future is swapping laptop with an AMD graphics one. People shouldn't make the wrong choice and then demand other to fix something that could be easily avoided.

Yes, AMD laptops are surely the better choice for a Linux System - at least in theory. But the hard truth is, that even with Ryzen most laptops before 2018 have a Intel (+ nvidia) combi. And the 2018 Ryzen laptops had back then a lot of Linux problems. I personally did want to buy a laptop with AMD APU, but back then there was no *good* option for my use case. Now I'm stuck with a nvidia gpu at least for the next 5 years, because I can't afford a new laptop just because nvidia is a dick to the FOSS community. And at least they proposed the fix themselves, which is a nice move.