Lots of great points, I wish I knew a place to post this so all the creators of all Linux Distroes & GSDs could read this.
2 Likes
Most of the engineers that designed, implemented and supported X11 for the last 40+ years think that Wayland is a very good idea because those same engineers desiigned, implement and maintain Wayland.
13 Likes
Well, it is AI slop with the goal of product placement for hardware that hitches on the current announcements for transitions to Wayland.
Unfortunately AI slop generates a lot of inaccuracies and packages that in a very nice form.
So people might actually believe the pseudo content and end up proliferating it further.
13 Likes
The Wayland implementation up to each compositor was a horrible idea as now if I develop something for say KDE KWin there is no guarantee it will work correctly in GNOME Mutter as nothing says both compositors have to support the same extensions and they can even introduce their own, the more compositors the more potential problems.
Do the BSDs use Wayland?
That was already the case.
KWin and Mutter did not magically split off the same code base when they introduced support for Wayland, they have always had different implementations.
Compositors, even âsimpleâ window managers, implement dozens of protocols and standards, desktop shells implement many additional ones.
Due to different programming languages and/or software stacks these implementations are often not shared.
Same on the client/toolkit side.
It is understandable that people who are not involved in development would not realize how many different pieces are involved in their system and how many of these components do not share much or any code even if the serve the same purpose.
Unfortunately the resulting misunderstanding/misconception can easily lead to do the kind of inaccurate content the linked video contains.
Adriaan de Groot blogged about running Plasma/Wayland on FreeBSD earlier this year.
7 Likes
Yes, Iâm daily driving Plasma6/wayland on FreeBSD 14.3 here, two machines run great, one with an older intel chipset is sluggish as hell and remains on xorg, which is ok with drm/modesetting xorg.
But FreeBSD(or derivates like Nomad or GhostBSD) is just one, of the very different to each other, BSD/Unixoid OS.
There is an uptodate plasma port for OpenBSD, which I tested on xenocara, Openbsdâs security fork of xorg. When I asked on Mastodon about âif plasma6 going all in on Wayland then does it mean for OpenBSDâs port deprecation?â, apart from the usual snarky remarks from the âwho cares, donât hold us back nowâ move-fast-break-things folks, I learned there are efforts to bring the wayland protocol to openbsd, both from âporting is hard, because wayland relying on linux solutions (?)â to âI already run a tiling wlcompositorâ.
I donât know if plasma6 or wayland, libinput, drm, dbus etc has already been ported to NetBSD or Dragonfly. Same for Illumos/Opensolaris. Others might know more.
tl,dr; So, yes of the BSDs FreeBSD should be mostly ok. Iirc 15.1 will come with a plasma6 desktop target as an option from the installer. OpenBSD might be or not, others might be ??? ymmv
(sorry for the longer reply, feel free to add or correct, if I got something wrong)
iâm very concerned if not to say worried. using libvirt/qemu/kvm in a client-host system still needs x11 as protocol. wayland is not that compatible that it works. especially all the services that rely on "âremoteâ-actions do not work properly and often arenât working at all. wayland on a single desktop computer is okay-ish but what about virtualised client-host systems in business ? wayland is promoting too much and overestimate the efford they have done. itâs still too early to skip x11.
4 Likes
I am just bummed that I have a good KDE setup and that when I try Wayland, things I rely on to be productive turn out not to work. The KDE end of life statement on x11 states that the âvast majority of our users are already using the WaylandâŚâ Am I really that unusual of a KDE user because I use x11? I would be interested in how this was assessed. I had no idea that I was so odd for using x11.
3 Likes
Plasma 6 defaults to a Wayland sessions (same with Gnome). A lot of distros also default to wayland by default. Anything that requires X11 is quite niche at this point since almost all things work better in wayland and are much safer for end users.
Thatâs very subjective, and unfair to folks like me using stuff that does not work with Wayland. Weâre ânicheâ so we donât count.
A bit more respect, please.
6 Likes
How was that disrespectful? I just gave you the facts and data.
Most problems with Wayland has fixes on the way and will be resolved by the time X11 is dropped (2027)
This is the sole issue I have with Wayland: Productivity features I rely on are simply not there, or not fully implemented.
Thereâs no argument that Wayland is the future, with most X11 developers working on it. Itâs not only a good idea, itâs GREAT in daily use for most of you. However, itâs not yet ready to replace X11 for those that rely on these ânicheâ capabilities daily.
The shortcomings are well-known and tracked, by this point only a handful of differences remain from the X11 session. These arenât whimsical requests or niche cases, theyâre heavy hitters for anyone that use persistent multi-window apps, headless remote access, etc.
Itâs time to get specific, whatâs missing in Wayland that you use in X11? Is it already on the show-stopper list? We need to pound it in now, come out of the woodwork with every ânicheâ use case so we can make sure thereâs a Wayland equivalent.
And TBH, leave the conspiracy garbage at the door. Wayland makes sense on paper, but itâs just not finished for the remainder of us. Once multi-window and remote access are 100% indistinguishable from my X11 session, Iâll gladly migrate.
1 Like
Would you like it if your concerns were dismissed as âquite nicheâ, and implicitly ignorable?
2 Likes
I can quite relate as I am testing a Wayland session every now and then while primarily running the X11 one.
However, each test the delta is decreasing, with the delta in the other way around increasing.
I think for me the point of transition comes when the window restore protocol has been finalized and enabled by default. At which point it will surpass any comparable option on X11 due to its capability of restoring activity association.
My guess would be several sources of statistics.
@dphaldes already mentioned distribution defaults and thus their respective package installation statistics.
Then Plasmaâs own user feedback statistics and maybe those of applications for which they are publicly available.
It wonât be a perfectly accurate picture of reality but should allow for rough estimations and detection of trends.
Realistically there isnât much difference for a lot of users so they use whatever the default is and have likely not even realized they made a transition.
I am using Plasmaâs window handling capabilities to an extent that is far beyond what most other people do.
At this moment I have 25 windows across 6 virtual desktops. And that is just one of my 5 activities!
I would probably need a separate user account for each activity if activities did not exist and I have no idea how I could possibly work without virtual desktops.
Yet many people work without those, even on platforms where such features are non-existent or have not existed until recently.
If, for example, you are not using activities at all, you might consider that a niche case and compared to the total of all users it most likely is.
So any of use could have a use case that is important, potentially even essential, yet not as widely used as we would imagine.
Personally I think a good way to improve my odds is to improve awareness of activities.
2 Likes
well iâm in linux for a long time now and wayland has such a long history of announcements that failed or had been delayed. imho linux is on the brink to get very popular in homeusers and business (thanks to m$) but it has to get more reliable. wayland is still a playground and yes itâs the future but plasma/kde should focus on delivering a working-system on any circumstances. using plasma enabling x11 or wayland makes it very reliable even in business. it will be a hassle if plasma/kde throws this flexibility under the bus by switching to wayland under all costs. wayland always promised things on schedule that they always missed. 2027 ? i donât think so. x11 is a protocol that is used for decades now and to upgrade all the applications and services to wayland in a couple of years ? thatâs a too optimistic point of view.
3 Likes
I run libvirt on my Plasma Wayland and inside it I run desktops that use Wayland.
Where did you see that x11 is required for libvirt?
2 Likes
youâre running it local on your pc. i already explained this is related to client-host systems where a server hosts external pcâs.
Are you referring to qemu+ssh?
Given that virtualization host and guest share the same hardware, i.e. âlocal PCâ, it might be helpful if you describe your setup in more detail rather than vague generic terms.
Assuming âclientâ in your description is not the virtualization guest, what is it?
How does this client access the virtualization host?
Or does it access the virtualzation client?
In either case which remote access technology is being used?
2 Likes
I watched most of the video and I can say it is substantially wrong, not necessarily that its facts are wrong (some probably are) but because it is the wrong direction.
First, there is always the following option for old hardware: use old software. No one will stop you from using the last supported X11 version of a DE.
âBut that wonât have updates and work with the modern web and there wonât be security updates!â
Well the lies the problem. Now the community is being asked to focus its resources not on the future but on the past. Perpetually support hardware that is fading away. Continue to build for things with a constantly diminishing user base who increasingly will have no nostalgia for the old devices. Build for things that are architecturally prevented from being properly secure.
If a community really values keeping these project supporting X11, they can fork it and do it themselves. The real objection is that the limited resources of modern developers be used for what I want vs what you want.
Lets suppose all the problems with Wayland are true - and some of them are. There are still plenty of tools for remote desktop, KVM, and other tools that are still buggy and not ready for prime time.
Well, for a long time developers have been able to say, âIf it doesnât work just use X11 for now.â Developers who canât say that anymore will have a fire lit under them to get fixes in place or to make their case to upstream projects.
Ultimately it is not enough to say, âYou are obsoleting my hardware!â No, the environment around you has changed. If your hardware worked before, just stop updating it. If that is not sufficient, it is because other things have changed. This is why if you boot an old Windows XP computer and use Internet Explorer it struggles to handle modern web pages. If your goal is to get the modern internet to work on XP, that is a niche project, not something that needs the support of the whole community.
All that being said, there is still a lot of Wayland work to be done. We should not minimize that. But you have to rip the Band-Aid at some point.
3 Likes