With Fedora KDE and Kinoite being fully Wayland by default from login (since F38) to desktop (since F34), it's now time to work toward eliminating our dependency on the Xorg server for Plasma 6.0.
Three reasons for this removal:
This will drastically reduce our support burden and give us the ability to focus on quality for the KDE Plasma stack and continue our feature-forward nature.
This means Plasma 6.0 isn't getting backported to any current release of Fedora. At this time, we are not planning on backporting it to any current EPEL release either, and it is targeted for EPEL 10.
This also does not mean that X11 applications will not work in Plasma 6.0, as we intend to still support Xwayland for running X11 applications on Plasma Wayland.
First merge request upstream for this: https://invent.kde.org/plasma/plasma-workspace/-/merge_requests/2188
This will make Plasma Wayland the default for Plasma 6 upstream unless a distro configures it otherwise.
I know I'm the new guy in the team but this feels a bit far to go [this isn't the words I want to use, I lack the words to describe it at the moment].
I'm all for pushing Fedora forwards, adopting new technologies, etc. But restricting our spin to Wayland only feels like we're excluding users, those that still need x11 for whatever reason.
I've seen other unnamed projects remove features and they cop a lot of negative feedback for it and rightfully so in my opinion.
I am not against removing it in the future, I just don't think a sufficient portion (my opinion) of the Fedora userbase can comfortably work, play and use their Fedora devices solely on Wayland only just yet. Just to clarify there are no facts behind this, merely speculation and observation.
I'd like to see a lot of discussion happen around this change before any decision is made.
Metadata Update from @justinz: - Issue tagged with: meeting
To date, there are only two major legitimate cases I'm aware of for X11 session usage:
Here's the deal, though. Neither @marcdeop, nor @tdawson, nor myself want to wind up having to commit to maintaining the Xorg server in any shape or form. We've already mostly eliminated X11 from the KDE Plasma hot-path now mostly so we can avoid that scenario. Xorg releases are already a struggle upstream, and dealing with Xorg CVEs is tough work.
@uraeus wrote on his blog back in 2019:
The reality is that X.org is basically maintained by us and thus once we stop paying attention to it there is unlikely to be any major new releases coming out and there might even be some bitrot setting in over time. We will keep an eye on it as we will want to ensure X.org stays supportable until the end of the RHEL8 lifecycle at a minimum, but let this be a friendly notice for everyone who rely the work we do maintaining the Linux graphics stack, get onto Wayland, that is where the future is.
That window is rapidly closing in on us and upstream has already split Xwayland into its own release cadence separate from the main Xorg stack.
Additionally, there are many features going into KDE Plasma now that are Wayland-only anyway (new features in Spectacle; Kwin's mixed-DPI, VRR, and soon HDR capabilities; etc.). @zzag also started upstream discussion about splitting kwin_x11 out of kwin because it's become increasingly problematic to deal with both all the time.
kwin_x11
I don't expect upstream KDE Plasma 6.0 to rip out X11 support yet, but I expect that every distribution that doesn't start preferring Wayland now to be in for a rude surprise soon. And in our case, I expect that the decision will be made for us anyway if we don't do it of our own initiative. I'd rather us purposefully do it rather than be forced to.
The one major issue I've heard of is graphics work, whether video or images. I've been told color calibration doesn't work or is not correct on Wayland just yet.
I know night color doesn't work on my machine and while that's an NVIDIA issue with Wayland, it's still only on Wayland from my understanding.
I wouldn't expect anyone in the team to pickup the work keeping Xorg server alive. However, removing it harms the userbase.
Even if we specified that X11 sessions were unsupported, people could still use them of their own choice.
This has been part of the work for HDR (defining how to do color profiles, color management, and color calibration). @sebastianwick has been working on Wayland protocols for color management and for color representation. That said, if you're using an external colorimeter, you're no worse off than you were using X11, provided the compositor supports loading color profiles.
Simon Ser (the wlroots lead) wrote a blog post about the recent HDR hackfest organized by @csoriano and @tpopela where this was discussed and planned out. @apol and Xaver Hugl were at that hackfest representing KDE. I'm fairly optimistic that problem will be resolved soon. There's a kwin issue tracking this work too.
Night color is a subset of color management (dealing with color temperature), and the problem is specifically with the NVIDIA proprietary driver not supporting GAMMA_LUT, which is being worked on. That said, I expect that given the active development to integrate support for the new firmware and interfaces into Nouveau, the proprietary driver will not be a driving concern for us for that much longer.
GAMMA_LUT
And the thing is, shipping a whole extra session that we don't support means we're on the hook for security fixes and other issues related to it. And as long as Xorg remains part of our dependency chain, then we are going to have to deal with problems caused by it.
I will also point out that Plasma 6.0 is pretty much the only "good" time to make this change. People generally expect that major experience changes occur with major version bumps, but if we drop X11 sometime later rather than at this boundary, people will be extremely upset with us.
My problem with this, as I've said on other occasions, is that session restore on Wayland still seems to be broken. That's the only reason I still use X11. Once that's fixed, I'm fine with switching, but it doesn't seems to be getting much attention.
Based on what @ngraham wrote last week, I think there's a drive to close out those issues ASAP for Plasma 6.0.
It's listed as a "true showstopper": https://community.kde.org/Plasma/Wayland_Showstoppers#Session_management
And there's a merge request prepared to add support for it: https://invent.kde.org/plasma/kwin/-/merge_requests/3024
I guess that's encouraging, however IMHO it's premature to drop X11 support before this is delivered (or did I misunderstand the proposal?).
This ticket is for tracking our work to actually get this done. If we can't pull it off, then we'll figure something out. But we (and everyone else involved in this effort) are generally quite motivated to avoid having to step up to maintain the Xorg server and all the old Xorg server infrastructure components that haven't had releases in years or maintainers for that matter.
Fedora Workstation and Fedora KDE are the only Fedora editions that ship both X11 and Wayland, so I think we should wait until Workstation drops X11. If we drop X11 earlier than Workstation does, we'll have a lot of angry users, and it probably won't help our reputation as being "second class" to Workstation (as untrue as that is)...
if we drop X11 sometime later rather than at this boundary, people will be extremely upset with us.
People will be upset if it's dropped for Plasma 6.0 too, in fact I think more people would be upset because Workstation would still have an X11 session. I can already picture all the "Fedora treats the KDE spin as second-class", "Fedora doesn't care about KDE", and other inflammatory posts that'll be on Reddit, the green site, YouTube comments, etc. If the Workstation WG and KDE SIG coordinate to drop X11 at the same time, then I can see the discussions being more about "Fedora isn't interested in maintaining X.Org because it's a dead end" instead.
Neither marcdeop, nor tdawson, nor myself want to wind up having to commit to maintaining the Xorg server in any shape or form.
I don't know why you think continuing to ship Plasma's X11 session, when it's still developed upstream, somehow also means that you or anyone else in the KDE SIG will have to step up to maintain X.Org itself, that's just silly.
I'm not saying we should keep X11 for long - I want it gone just as much as you do - but I just think it'd be a bad move to drop it sooner than the only other Fedora edition that ships both Wayland and X11.
That literally doesn't make sense, since the only overlap in maintainers for Workstation and KDE variants is me. I would love to drop it in Workstation too, but it's not my call. But we've been in agreement for a while now that we should do this for Plasma 6.0 in the Fedora KDE SIG.
if we drop X11 sometime later rather than at this boundary, people will be extremely upset with us. People will be upset if it's dropped for Plasma 6.0 too, in fact I think more people would be upset because Workstation would still have an X11 session. I can already picture all the "Fedora treats the KDE spin as second-class", "Fedora doesn't care about KDE", and other inflammatory posts that'll be on Reddit, the green site, YouTube comments, etc. If the Workstation WG and KDE SIG coordinate to drop X11 at the same time, then I can see the discussions being more about "Fedora isn't interested in maintaining X.Org because it's a dead end" instead.
We've made it abundantly clear throughout the last several releases of Fedora KDE that we are generally focused on Plasma Wayland. Bugs in Plasma X11 just do not have much priority among us, and all automated testing is on Plasma Wayland exclusively.
There's basically no good time to rip the band-aid. But at least doing it on a major version boundary makes it more "acceptable". If we ripped it out in 6.1 or later, people would be even more upset, because it's a huge breaking change in the user experience in a "minor version", even though the major version merely indicates the Qt version it pairs with, and the minor version is the "true" Plasma Desktop major version. Explaining that nuance is just not going to work out.
And upstream KDE is working hard toward making Plasma Wayland the default upstream, which will make most distributions shipping KDE Plasma 6.0 have Plasma Wayland by default too.
Neither marcdeop, nor tdawson, nor myself want to wind up having to commit to maintaining the Xorg server in any shape or form. I don't know why you think continuing to ship Plasma's X11 session, when it's still developed upstream, somehow also means that you or anyone else in the KDE SIG will have to step up to maintain X.Org itself, that's just silly.
Why is that silly? If we wind up in a situation where there's no maintainers for the Xorg stack in Fedora and we still have dependencies on it, we're basically forced to take it. I don't want to be in that position, so I will try my hardest to avoid it.
GNOME and KDE are completely independent projects with completely different architectures. In some respects, I could make the argument that KDE Plasma's Wayland session is more featured than GNOME's and evolving more rapidly as well.
My perspective is that we do not have the manpower to maintain the X.org stack. When the rest of Fedora drops it, we will have to drop it as well.
In order to not be surprised, we need to prepare for it and preparing for Plasma 6 to not include X11 support is a good way to do that.
If, and that's a big if, we still have X11 supported in the Fedora release where we land Plasma 6, then we can re-consider.
I still use vncserver [1] to have a GUI on an headless server. I'm not sure how KDE's RDP implementation on top of Wayland is -- last time I looked, keyboard didn't work.
Gnome's RDP implementation of top of Wayland does work, but it only works for sharing an already started session. Plumbing for GDM is in the works for the last two releases [2] (thanks Pascal Nowack!). Not sure about KDE's behaviour/SDDM integration.
If I'm not mistaken, removing X11 would prevent the usage of vncserver as a remote GUI provider.
[1] https://docs.fedoraproject.org/en-US/fedora/latest/system-administrators-guide/infrastructure-services/TigerVNC/ [2] https://gitlab.gnome.org/GNOME/gnome-remote-desktop/-/issues/92
There's a new RDP implementation being worked on for Plasma 6.0. This may be able to solve that problem for you.
I recently tried Wayland on Plasma again, and identified two showstoppers for me personally. The first was screen-sharing with Zoom, which I see some discussion of above / potential future workarounds. I don't see any discussion of supporting Zoom screen-sharing annotations (other people drawing on my screen) which is quite important to me as well (maybe this will "just work", but given all the complex bridges and so forth needed to make even basic sharing work, I doubt it).
I also don't see any discussion of restoring window positioning when applications are restarted. This is especially problematic for applications like Chrome and IDEs like IntelliJ IDEA where these applications will routinely have 5-10 or more windows open, and each window is carefully positioned across desktops and screens, as well as within each screen.
Currently, restarting such an application simply puts all the windows in the center of my central monitor, making Wayland extremely annoying to use for what I suspect is a common workflow. The closest I see to this is the Session Restore showstopper work but I'm not sure if window positioning on app restart is included within that bucket. I don't see any explicit mention of it.
If your workplace uses Zoom, then you should file support cases with them. But we cannot block on Zoom. They seem to only fix things when it is dire. :frowning:
Insofar as the rest of it, please discuss that with KDE upstream by filing bug reports.
Personally, I think this is a bad idea until Wayland proves it's 100% ready. I own an nvidia card, and currently I feel KDE under Wayland is only 92% ready, at best. I still experience one too many glitches and errors under Wayland to call it viable, and its not like I have a bad card or lack of the needed drivers. Sure its an older card (a RTX 2070 super) but it works well with X11 and I cant say the same for Wayland. If the fedora KDE team does this, be my guest, I will probably be using cinnamon or something until I feel Wayland is ready. So see Fedora KDE again in 2039?
Could those that are willing to do so maintain X11 packages for plasma 6?
My current setup includes flameshot and rofi. flameshot has various issues in regards to clipboards, windowing and many more on Wayland. rofi refuses to support wayland because of ideological reasons and the wayland fork crashes on startup.
While Wayland may be the future (even just because no one wants to maintain Xorg) it still lacks support from every application and XWayland isn't a perfect drop-in solution.
Plasma 6.0 isn't even out yet? Perhaps you should consider waiting and seeing how things play out a bit. That said, if you're also willing to do upstream work on the Xorg server and help maintain it in Fedora, then maybe. But that's a lot of work that so far nobody has been willing to do.
Insofar as Flameshot goes, it looks like there are only two minor issues (one of which doesn't apply to Fedora KDE at all).
rofi-wayland on the other hand seems to not be very useful in a KDE environment since its features are built into KDE Plasma already.
rofi-wayland
In many respects, XWayland performs better than regular Xorg, and there's a lot of work going on right now to improve it for even more legacy applications. Windows games are the main gap, but even that's rapidly improving as we speak.
And when it comes to being "100% ready", there's basically no way to be that for everyone until after it happens because the long tail only adapts when required to.
I'm pretty happy with where we are now, and I'm pleased with the progress upstream KDE is making on Plasma Wayland for 6.0.
In my opinion, as just an observer, I think Fedora (KDE or GNOME) should not include X11 in the ISOs, it should be an optional thing installed only after the system installation (RPM Fusion or in the repos but in preparation to be dropped from it) if some people encounter a corner case that is important to them.
I don't see why the community should maintain X11 forever to be a crutch for Nvidia proprietary drivers that can't even be included in Fedora (Wayland works fine on Nouveau).
In my opinion, as just an observer, I think Fedora (KDE or GNOME) should not include X11 in the ISOs, it should be an optional thing installed only after the system installation (RPM Fusion or in the repos but in preparation to be dropped from it) if some people encounter a corner case that is important to them. I don't see why the community should maintain X11 forever to be a crutch for Nvidia proprietary drivers that can't even be included in Fedora (Wayland works fine on Nouveau).
Yeah but I think there should be at the very least a safety net until we are 100% sure Wayland can replace xorg full stop without issues. Including an x11 session still might be a necessary evil regardless what anyone thinks about it. Nvidia cards still need something, otherwise it won't work and one can bet fedora would be inundated with complaints. In my opinion Wayland still has to get better on KDE, sure it's mostly fine on gnome but as it stands KDE is a long way off.... again at least a year, maybe two considering the issues I have faced. Nvidia will no doubt get better, but I don't see that happening until another driver or two down the road. Personally I would hate to be forced to use gnome on fedora as I hate gnome with a burning passion.
Here's the thing, even Plasma X11 has issues. Some of them are bugs, some are missing features, and some are fundamental issues through the X11 architecture. I am not going to promise that we can be "100% sure Wayland can replace xorg full stop without issues" for everyone because that's impossible. The Wayland architecture has its own trade-offs that make a fully equivalent experience not possible. The best we can do is ensure as many use-cases are supported as possible.
NVIDIA's proprietary driver does support Plasma Wayland. We have members on the Fedora KDE team daily driving it. And there's work going on to revamp and significantly update Nouveau for the last 3 generations of NVIDIA GPUs, which are 8 of the top 10 GPUs from the Steam hardware survey, including the number one GPU.
With KDE Plasma 6 upstream intending to switch to Wayland by default, there's much more intense focus on resolving the remaining important issues. I have confidence and faith in the KDE folks to pull it off.
Here's the thing, even Plasma X11 has issues. Some of them are bugs, some are missing features, and some are fundamental issues through the X11 architecture. I am not going to promise that we can be "100% sure Wayland can replace xorg full stop without issues" for everyone because that's impossible. The Wayland architecture has its own trade-offs that make a fully equivalent experience not possible. The best we can do is ensure as many use-cases are supported as possible. NVIDIA's proprietary driver does support Plasma Wayland. We have members on the Fedora KDE team daily driving it. And there's work going on to revamp and significantly update Nouveau for the last 3 generations of NVIDIA GPUs, which are 8 of the top 10 GPUs from the Steam hardware survey, including the number one GPU. With KDE Plasma 6 upstream intending to switch to Wayland by default, there's much more intense focus on resolving the remaining important issues. I have confidence and faith in the KDE folks to pull it off.
I'm not saying Xorg is perfect, I am well aware it too has its fair share of issues. But that doesn't invalidate my concerns as an end user. I mean, this isn't just an update to a DE here but an entire display server, one that still has issues on some cards and setups. I myself have a RTX 2070 suer and use 1440P resolution, and currently stuff like fractional scaling is still a little borked on that front. Not to mention the slow-downs, I never once had slowdowns on Xorg even when I was running linux on a hard drive on my old computer made in 2009. I just say it's still too early, RHEL deprecating xorg reeks of corporate BS as I have tried to use wayland more than once and still feel it's unready. Xorg may have its own issues, but it still works better for me than Wayland at this point, even in gnome where Wayland is allegedly better still has its quirks that need serious addressing.
Yeah but I think there should be at the very least a safety net until we are 100% sure Wayland can replace xorg full stop without issues.
I think that Xorg should be available for those who need it, but not in the ISO. So I think that having Xorg as installable after system install is adequate and sufficient at this point.
Nvidia cards still need something, otherwise it won't work and one can bet fedora would be inundated with complaints.
I think the community should not fear the complaints about certain things not working on Wayland, more feedback is needed to fix the problems, not people throwing the dirt under the carpet choosing the X11 session. Legitimate use cases that can't be contemplated with Wayland atm can install X11 after the system install.
In my opinion Wayland still has to get better on KDE, sure it's mostly fine on gnome but as it stands KDE is a long way off.... again at least a year, maybe two considering the issues I have faced.
I think that KDE should retain the ability to work with X11 (optionally), so you can use X11 if you choose to install it after system installation. But I think that X11 should not be present in the system install for the motives I said above. Wayland is far from an impediment to the system to work like it was some years ago, you can just boot, log in the wayland session, install Xorg, reboot and choose the X11 session.
Yeah but I think there should be at the very least a safety net until we are 100% sure Wayland can replace xorg full stop without issues. I think that Xorg should be available for those who need it, but not in the ISO. So I think that having Xorg as installable after system install is adequate and sufficient at this point. Nvidia cards still need something, otherwise it won't work and one can bet fedora would be inundated with complaints.
Yeah, but what if wayland doesn't work at all for your card? Guess time for another distro? I mean, might as well ditch fedora if Wayland works like crap for your GPU. In any case, that's why I think a fallback is necessary, at least for another year after Plasma 6 comes out. Again Wayland is close, its super almost ready but when you have a nvidia card it does make matters more complex.
With NVIDIA proprietary drivers, you should be complaining on NVIDIA's forums, not here. They only fix things when people complain en masse about it. The problems are related to their driver being incomplete, not us doing something wrong here.
Yeah, but what if wayland doesn't work at all for your card? Guess time for another distro? I mean, might as well ditch fedora if Wayland works like crap for your GPU.
tbh I never saw an nvidia card doesn't work with wayland in the sense they wouldn't show any video at all for years. In Fedora, you need to activate the nvidia proprietary drivers anyway with RPM Fusion and in the process you can install Xorg too if you want.
X11 as a fallback is unnecessary today because Wayland works acceptably already if you need to install X11 to accommodate your needs. The focus of new development and bug fixing should be focused entirely on Wayland.
A few things I am really not understanding in all this...
Given that:
Can you clarify what exactly is meant by proposing "No X11 for Plasma 6"? e.g. assuming that this was accepted, I get the impression that would mean the official spin team would only work tickets relating to Wayland+KDE, that the spin iso would only contain Wayland, and that on an out-of-the-box install only Wayland would be selectable at login...
but what about:
1. Would KDE+X11 in Fedora still be possible from a user perspective? e.g.
1-a. Could they download the spin image and install with wayland, but afterwards manually install x11 and still have KDE+X11 appear as an option on the login screen?
1-b. Is the thought to prevent this entirely for some reason?
1-c. How/does this change if a SIG materializes?
1-d. What about if there was no official SIG but someone added something on rpmfusion and/or copr?
FWIW, I don't get the sense that you are trying to actively prevent people from using x11, more just that you don't wish to maintain it or have the resources to do so, but I think people seeing other options for this discussed in a little more depth could alleviate some fears here.
2. I assume that any tickets related to X11+KDE would either be ignored or rejected if no SIG for KDE+X11 materialized. But if one did, what changes to current process do you see there being (or would it literally just be assigning a ticket to a different person)?
3. Timelines: I have not been able to find anything related to when Plasma 6 might release. I've found the upstream schedule and progress porting Frameworks and Plasma to Qt 6 pages.
Normally, based on loose notes like this and assuming that there would probably need to be time for QA, I would give a very rough SWAG based on absolutely nothing but gut feeling of "not likely in 2023" but given this I am not really sure.
Do you have any better guesstimates than this in terms of predicting how much time folks that require x11 have?
e.g. "at minimum x11 should be around until ___, but likely we would not see plasma 6 being fully ready until ___ \" ... ___ could be in terms of very rough timeframes or Fedora release targets. This is not at all about committing to a timeline and instead just about approximating the correct scale of time (weeks, months, half year, longer, etc).
For some folks, I think if they heard that x11 would definitely be around for one or several more Fedora releases, that would go a long way towards easing immediate concerns bc if nothing else, there would be a little more time for planning or for migrating their setups to wayland, as well as for more wayland bugfixes to be patched/improvements to be developed.
Update: after more searching, I also found this post from just last week
https://pointieststick.com/2023/06/18/on-the-road-to-plasma-6/
That seems to make some rough SWAGs on timeline. Looks like mine isn't totally unreasonable and with this, I would guess that even if Plasma 6 dropped early than Nate predicted that they would not be time for proper QA in rawhide prior to Fedora- 39 release targets. So my thinking is that it's very unlikely to make Fedora 39, but could possibly be out in time for Fedora 40. Would you guess similarly?
edit: fixed typos, formatting, and (hopefully) poor phrasing
edit 2: added (hopefully) more accurate timeline guesses from upstream kde
that is a change for RHEL, not for Fedora (which is upstream of RHEL, not downstream and thus not directly affected by such a change).
The same set of people maintain X.Org upstream, in Fedora, and in RHEL, so it's very likely it'll end up unmaintained when it drops from RHEL.
Fair enough, but unless I'm mistaken, wouldn't it still be supported via RHEL 8 until at least the Full Support window closes (May 31st, 2024 if this link is accurate), and possibly longer via RHEL 8 EUS? If so, then sounds like x11 would still be supported (in Fedora itself) through F39 for sure and very likely thru at least F40.
And anyway, IIRC Red Hat doesn't official support KDE either (in RHEL). I know KDE has upstream maintainers but it's not impossible SIG could materialize for x11 once RHEL is no longer officially supporting it. If that happened, probably there would be a delay though, as there usually is with other things that RHEL drops but are picked up by the community. So still curious about the questions above.
My two cents on the subject:
Replying to some questions: 1. Would KDE+X11 in Fedora still be possible from a user perspective? e.g.
I would personally aim for: no (note: I don't have the authority on this). If you need X11, stay with the previous version of Fedora and likely the next one will have support for all your needs.
no
1-a. Could they download the spin image and install with wayland, but afterwards manually install x11 and still have KDE+X11 appear as an option on the login screen? If Xorg maintenance is dropped... likelyl that possibility won't exist.
1-b. Is the thought to prevent this entirely for some reason? We don't actively prevent any of this but we are being pragmatic
1-c. How/does this change if a SIG materializes? In my opinion, if X11 is being maintained, we could perhaps give the option of an extra installation (not installed by default)
1-d. What about if there was no official SIG but someone added something on rpmfusion and/or copr? If is not officially maintain: no
In my head, Fedora KDE on Plasma 6 should aim for a full wayland experience + xwayland ( and yes, that means no X11). As stated above, if there are some things which do not work for you on wayland then, stick to the previous version of Fedora, report bugs and likely your issues will be solved by the next release).
I would personally aim for: no (note: I don't have the authority on this).
Sorry, I am not familiar with everyone here yet. Does this mean you are on one of the kde spin team / one of the teams doing related work, and you vote no for maintenance reasons but are not the lead...
or you are a user like myself participating in discussions and that you dislike x11 so you vote no... Or something else?
If Xorg maintenance is dropped... likelyl that possibility won't exist.
This seems to be the critical piece then, e.g. a) is x11 to have maintainers at all, b) if x11 maintained are there limitations in scope e.g. not just gnome issues but kde etc too, and c) if no maintainers whatsoever what is Fedora steering committee planning to do with x11 packages in Fedora.
I know for smaller unmaintained packages that they have a history of disappearing from the repos pretty quick. OTOH, I think this has a lot of far reaching implications outside of KDE as well. I think something like half of the current spins only work under X11 so I believe that the steering committee would probably take that under consideration before dropping the x11 packages from repos completely.
Definitely interesting to see what happens. For my own 2 cents, I have no big preference and care not if x11 is included on the live disc/default install or not. I do think that as long as the x11 packages are available in Fedora repos, that users should be allowed to manually install it themselves, e.g. not including it is fine but going out of the way to actively block users from setting up manually would not be cool
Saying this more bc I am generally in favor of user choice more than out of any loyalty to or desire to use x11 or Wayland either one.
If x11 packages were removed from Fedora repos tho, obviously that is beyond the scope of any one spin to deal with.
On Wed, 2023-06-28 at 07:22 +0000, Marc Deop wrote:
I've been using KDE on Wayland for over a year on 4 different systems and 2 of them Nvidia only with no issues (Specially since Jan and I (as tester) invested time in fixing screen sharing with Nvidia cards)
I'd like to repeat a comment I made earlier: how do people actually work with Wayland/KDE if it doesn't have session restore? Do you just never log out? Do you resign yourself to manually setting everything up every time you log in? What am I missing?
poc
Session restore is being worked on, but in my case, I'm the only person using my computer, so I just put it to sleep instead of logging out. It locks the computer and puts it in a low power state.
I do the same, but I also update my system frequently, meaning I have to log out and in again, even if not actually rebooting. That's when the whole thing becomes painful.
On Wed, 28 Jun 2023 at 12:50, Neal Gompa pagure@pagure.io wrote:
ngompa added a new comment to an issue you are following: Session restore is being worked on, but in my case, I'm the only person using my computer, so I just put it to sleep instead of logging out. It locks the computer and puts it in a low power state. To reply, visit the link below or just reply to this email https://pagure.io/fedora-kde/SIG/issue/347
ngompa added a new comment to an issue you are following: Session restore is being worked on, but in my case, I'm the only person using my computer, so I just put it to sleep instead of logging out. It locks the computer and puts it in a low power state.
To reply, visit the link below or just reply to this email https://pagure.io/fedora-kde/SIG/issue/347
I've been mostly okay with having to relaunch applications manually, as I'm used to it from other platforms too (I use all three OSes daily so that I can keep up with what everyone is doing).
It's fine if not everyone else is, but that's why I'm not concerned about it.
I would personally aim for: no (note: I don't have the authority on this). Sorry, I am not familiar with everyone here yet. Does this mean you are on one of the kde spin team / one of the teams doing related work, and you vote no for maintenance reasons but are not the lead...
I am a member of the KDE-SIG, I mostly do package updates :-)
how do people actually work with Wayland/KDE if it doesn't have session restore?
session-restore is not such a pain for me ;-)
Still, I would find it super easy to have a script starting all the apps I need on startup. Ain't that like very easy?
On Thu, 2023-06-29 at 06:29 +0000, Marc Deop wrote:
=C2=A0 how do people actually work with Wayland/KDE if it doesn't have session restore? =20 session-restore is not such a pain for me ;-) =20 Still, I would find it super easy to have a script starting all the apps I need on startup. Ain't that like very easy?
Not as easy as you might think.
1) You need to change the script manually every time you change your layout, instead of just clicking on Save Session.
2) More importantly, the script cannot determine the layout of the windows and virtual desktops, which is most of the problem. I=C2=B4d be happy to be corrected on this.
Let's not discuss technical details here.
The bug is recognized upstream as high priority and hopefully they will fix it soon.
With Fedora KDE and Kinoite being fully Wayland by default from login (since F38) to desktop (since F34), it's now time to work toward eliminating our dependency on the Xorg server for Plasma 6.0. Three reasons for this removal: Xorg server is deprecated since RHEL 9.0 and will be dropped in "a future major RHEL release". Graphics fallback modes are Wayland-friendly now with SimpleDRM enabled since F36. NVIDIA drivers (since v495~v515) support GBM for Wayland instead of EGLStreams. Wayland is fully supported on current NVIDIA drivers. This will drastically reduce our support burden and give us the ability to focus on quality for the KDE Plasma stack and continue our feature-forward nature. This means Plasma 6.0 isn't getting backported to any current release of Fedora. At this time, we are not planning on backporting it to any current EPEL release either, and it is targeted for EPEL 10. This also does not mean that X11 applications will not work in Plasma 6.0, as we intend to still support Xwayland for running X11 applications on Plasma Wayland.
Well I guess I'll make the switch to OpenSuse Micro because Wayland is still broke on my Kinoite/Rawhide install on my laptop (2021 HP 17-by3xxx, Intel CPU, Mesa Intel graphics, 16 GiB ram). Mainly the touchpad won't select most left clicks, unless you right click then select from the alt select menu, and other oddities and I'm not going to go thru the tedious process of listing everything demanded for a bug submit just to have impatient pros to say "that's been submitted in______ already..." and immediately close my hour long prep AGAIN and the link to the supposedly "same" issue not be the actual issue or whatever.... I've given up after several times and just use x as it works. Same issue with inability to rebase out of rawhide into a regular stable (something with glib needed by whatever and no source available). Instead of being reminded of how I'm not smart enough to follow some impatiently offered solution with Mac commands, from my Linux laptop, yet (I'm learning) I've just dealt with it. Now there is the inability to update Flatpak which I assume will eventually get fixxed as I prefer to just search and learn from someone else's solution but as close as I am to a pure keyboard only workflow, I'm not there yet, and still need a touchpad/mouse so this sucks. I get it tho.
oops. just saw the above about "no technical...". my bad.
Same issue with inability to rebase out of rawhide into a regular stable (something with glib needed by whatever and no source available)
Major version downgrades are generally not supported. Only forward rebases.
Now there is the inability to update Flatpak which I assume will eventually get fixxed
It should be fixed with the latest ostree that is now in the repos.
Could those that are willing to do so maintain X11 packages for plasma 6? My current setup includes flameshot and rofi.
My current setup includes flameshot and rofi.
@sentry for rofi see https://src.fedoraproject.org/rpms/fuzzel
rofi
Or wofi
On 26/7/23 19:41, Artem Polishchuk wrote:
atim added a new comment to an issue you are following: `` Could those that are willing to do so maintain X11 packages for plasma 6? My current setup includes flameshot and rofi. @sentry for rofi see https://src.fedoraproject.org/rpms/fuzzel `` To reply, visit the link below or just reply to this email https://pagure.io/fedora-kde/SIG/issue/347
atim added a new comment to an issue you are following: ``
Could those that are willing to do so maintain X11 packages for plasma 6? My current setup includes flameshot and rofi. @sentry for rofi see https://src.fedoraproject.org/rpms/fuzzel ``
My current setup includes flameshot and rofi. @sentry for rofi see https://src.fedoraproject.org/rpms/fuzzel ``
Metadata Update from @ngompa: - Issue untagged with: meeting
This was agreed upon in a meeting last month and reinforced again last week.
Sorry, but for clarity could you say what was agreed?
Based on the discussions with KDE upstream on their progress on Wayland showstoppers and our confidence in the Plasma Wayland development, we are going to continue with Wayland-only for Plasma 6.0.
This is now part of the Plasma 6 Change submitted: https://fedoraproject.org/wiki/Changes/KDE_Plasma_6
Metadata Update from @ngompa: - Issue close_status updated to: Fixed - Issue status updated to: Closed (was: Open)
Login to comment on this ticket.