Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

One of the comments on the article ( https://utcc.utoronto.ca/~cks/space/blog/unix/XDeathwatchSta... ) is also possibly of interest, suggesting possible technical routes for using existing windows managers on Wayland:

> There is XWayland, and apparently it will support handling the root window, not just application windows:

    https://wayland.freedesktop.org/xserver.html
> The missing piece of the puzzle is Xweston, which apparently allows you to run window managers for X on Wayland:

    https://github.com/ackalker/Xweston
> or maybe xwlnest will work better:

    https://github.com/gschwind/xwlnest


Again -- XWayland is Xorg. The xorg-server codebase contains a number of components, one of which is a front end called DIX (Device Independent X) that handles protocol-level stuff, another of which are several backends collectively called DDX (Device Dependent X) that handle drawing to the actual video hardware or other graphics layer. One of these DDX backends is 'xwayland'. So the XWayland server incorporates the entire Xorg server code base except for the hardware-specific back ends. And so if Xorg languishes unmaintained... well, sucks to be an XWayland user.

So your best bet is to commit now to switching, entirely, to Wayland. The aim is to get everyone off of X altogether as quickly as possible, and then stop shipping X (including XWayland) altogether.


>> So the XWayland server incorporates the entire Xorg server code base except for the hardware-specific back ends.

Well doesn't that mean XWayland will be easy to maintain since it has no hardware dependencies? X can completely stagnate so long as the Wayland back-end is kept up to date with any changes there - and no, people don't want to be changing Wayland protocols because every compositor would have to be updated. I think XWayland will be around for a while yet, but I'd rather have things run native Wayland.


Ostensibly, yes, but that would clash with the official narrative ("X is broken and bloated and needs to be deprecated"). If they can maintain Xwayland indefinitely, they can maintain a stub DDX that works with kms indefinitely also. What they are telling us is that they want to get rid of the entire thing. Pointing out that Xwayland would be relatively easy to maintain would be stating the emperor has no clothes.


> So your best bet is to commit now to switching, entirely, to Wayland. The aim is to get everyone off of X altogether as quickly as possible, and then stop shipping X (including XWayland) altogether.

Ah, it is the "32bit isn't needed anymore" all over again. Yeah, this will happen around the same time you wont need to ship 32bit support on Linux :-)


If the sole maintainer of the 32-bit code base decided to fuck off, and no one else is stepping up, your choices are to either stick with your old OS forever, or get a 64-bit machine.

This is where we're at with Wayland and X.


Maybe nobody is stepping up because they have no reason to do so and nobody even asked for anyone to do so? If the message was "hey, we want to stop working on Xorg and move to Wayland, is anyone interested in taking over?" instead of "Xorg is dead, we wont work on it, nobody will work on it, ever" then perhaps some people would show up?

Besides, even though i am not interested in taking over Xorg, if something breaks on it i'm more likely to look into fixing it than switching my environment to Wayland.


Problem with your suggestion is that Wayland won't be around in 15 years while Xorg will still be km kicking.


What makes you think that? The maintainers of both Wayland and Xorg have chosen to take Xorg behind the barn and shoot it.


That's not the germane point. There are applications dating back over 30 years written for X11. Thousands upon thousands of them. Compatibility matters.

Wayland may turn out to be a flash in the pan. Those applications aren't going away. I have many open source and proprietary applications which I fully intend to carry on using. They don't, and won't, support Wayland.

Unless you're using a modern toolkit which will be modified to support Wayland, you're not going to be using Wayland in a hurry. Because you'll be tied into the X11 world.


I can't help but notice that Xweston and wwlnest are both projects that seem to have gone years without an active contribution. Sometimes projects become mature and don't need additional work, but I would be stunned if anything to do with Wayland had reached that point yet.


> Sometimes projects become mature and don't need additional work,

You mean like X?

It seems to me the Wayland thing has a lot of support by people who think lack of changes in and of itself makes X completely unusable. That the software world must exist in a state of perpetual rewrite or it sucks.

I predict my minimal X setup will remain usable for a long time.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: