Mir 2.16 Released With Seamless Boot Integration, Improved X11 Client Support

[phoronix]

Phoronix: Mir 2.16 Released With Seamless Boot Integration, Improved X11 Client Support

Canonical this morning released Mir 2.16 as the latest version of this open-source software for helping to build out Wayland compositors...

https://www.phoronix.com/news/Mir-2.16-Released

[M@GOid]

I wonder if they will improve Mir to the point of using it on the Ubuntu desktop.

[JPFSanders]

Can MIR be used for example to run WindowMaker, Fluxbox, ICEwm, etc on top of it and have both X and Wayland apps running alongside together?

If not, what's the point? Why is nobody focusing on that use case? Want to get rid of X? Provide a way forward.

[AlanGriffiths]

Can MIR be used for example to run WindowMaker, Fluxbox, ICEwm, etc on top of it and have both X and Wayland apps running alongside together?

If not, what's the point? Why is nobody focusing on that use case? Want to get rid of X? Provide a way forward.

With Wayland, there is no separation between window manager and compositor, they are the same program.

Mir recognises the desire for customising window management and makes it easy to customise the window management part (by implementing a window management "policy"). (But that does require some coding.) It also includes support for X11 applications alongside Wayland ones (using Xwayland, the same as other Wayland compositors).

So a way forward would be to take your favourite window management paradigm and implement it using Mir.

[ehansin]

Can MIR be used for example to run WindowMaker, Fluxbox, ICEwm, etc on top of it and have both X and Wayland apps running alongside together?

If not, what's the point? Why is nobody focusing on that use case? Want to get rid of X? Provide a way forward.

Okay, going to take a shot here, but could get it completely wrong. The window managers you mention are X11 window managers. Within these window managers, my best guess is you are never going to run Wayland apps as stand-alone within them alongside stand-alone X11 apps (I could be proven wrong here.) That said, you could run these window managers on top of Wayland by running "rootless" (not to be confused with running as root, but meaning no root window) XWayland on top of this Wayland (and then of course whatever above window manager on top of this. Again, could have missed the mark here.

What you can do today is run one of the various Wayland-based window managers and run both Wayland and X11 (via XWayland) side by side. Now admitted the ecosystem of Wayland-based window managers is not as extensive as what exists for X11, but that should improve.

I did my best here, open to critique.

[JPFSanders]

I did my best here, open to critique.

No worries, no need to critique, I appreciate your post. Precisely you illustrate the point I wanted to make, that the whole Wayland "Edifice" should have included a migration plan where everything X11 would run on top of Wayland as it was and that would have allowed to replace parts with Wayland specific stuff bit by bit.

Running an entire X11 DE as root has only been made possible recently, and it is still incomplete.

This idea of Running X11 applications inside Wayland completely separated (not talking about isolation) and not being able to use a X11 window manager as "it is" to carry the at the time existing X11 DE's was as short sighted as making IPv6 incompatible with IPv4 and running a dual stack. Because in practice this is what happened to Wayland.

But well, this is what we have. I'm not a Wayland hater nor a X.org lover apologist, I want Wayland to supersede X.org once and for all, but won't happen for me until there is full parity with X.org in my personal use case, I need the computer for work, and Wayland as it stands in functional terms is lacking in so many areas it is not funny. I hope with the latest push from RH adoption will finally accelerate Wayland desktop development in the functional side of things.

[ehansin]

JPFSanders Sounds like you understood the situation, so my explanation was not needed. That said, good for me to sort out as well. I get what you are saying, though I also understand sometimes not easy to make all these things work. Just my take, but I'm happy with Wayland, though was never some sort of X11 "power user" by any means. Transitions are often tough, whatever the context. I am seeing so many pieces evolving in the Linux space. PipeWire, in a way merging the consumer and pro audio space as an example. And although I am just a layperson, my bias is positive towards Wayland right now in the graphical layer sphere. I buy the arguments that X11 just isn't suited for modern hardware and given many of the people behind Wayland were also in the X11 camp, gives me more confidence. Is there possibly even better solutions out there? I'm sure something will come along, but right now I am behind Wayland for today.

To end on a slight tangent, I have said here before I am okay with the Windows 10 UI (other things I dislike about it, trust me.) I can work well in it, and have to for work. But the more I have to help out with Windows 11, I am convinced MS is screwing the pooch there as well. Add in the telemetry, federated "pipes" to all the outside services and distractions (news, etc.), pushing AI integration - happy to get out of that space as much as I can. I am booting more and more into Linux by default, using Wayland-based desktops and window managers. There may be a lot of disagreements, but I am glad activity is happening in the Linux space.

[JPFSanders]

ehansin Yep, I agree with you, I am positive about Wayland, when it works well and one doesn't hit a rough edge it is fantastic. Like you I'm very optimistic about the future.