Tech Over Tea - The Future Of KDE Plasma Is Wayland | Xaver Hugl

Episode Date: December 12, 2025

Today we have Xaver Hugl from the KDE Plasma project, Kwin and various things involving KDE on the show to talk about the migration from X11 to Wayland and ultimately dropping the X11 session.==========Support The Channel==========► Patreon: https://www.patreon.com/brodierobertson► Paypal: https://www.paypal.me/BrodieRobertsonVideo► Amazon USA: https://amzn.to/3d5gykF► Other Methods: https://cointr.ee/brodierobertson==========Guest Links==========KDE Fundraiser: https://kde.org/fundraisers/yearend2025/Blog: https://zamundaaa.github.io/==========Support The Show==========► Patreon: https://www.patreon.com/brodierobertson► Paypal: https://www.paypal.me/BrodieRobertsonVideo► Amazon USA: https://amzn.to/3d5gykF► Other Methods: https://cointr.ee/brodierobertson=========Video Platforms==========🎥 YouTube: https://www.youtube.com/channel/UCBq5p-xOla8xhnrbhu8AIAg=========Audio Release=========🎵 RSS: https://anchor.fm/s/149fd51c/podcast/rss🎵 Apple Podcast:https://podcasts.apple.com/us/podcast/tech-over-tea/id1501727953🎵 Spotify: https://open.spotify.com/show/3IfFpfzlLo7OPsEnl4gbdM🎵 Google Podcast: https://www.google.com/podcasts?feed=aHR0cHM6Ly9hbmNob3IuZm0vcy8xNDlmZDUxYy9wb2RjYXN0L3Jzcw==🎵 Anchor: https://anchor.fm/tech-over-tea==========Social Media==========🎤 Discord:https://discord.gg/PkMRVn9🐦 Twitter: https://twitter.com/TechOverTeaShow📷 Instagram: https://www.instagram.com/techovertea/🌐 Mastodon:https://mastodon.social/web/accounts/1093345==========Credits==========🎨 Channel Art:All my art has was created by Supercozmanhttps://twitter.com/Supercozmanhttps://www.instagram.com/supercozman_draws/DISCLOSURE: Wherever possible I use referral links, which means if you click one of the links in this video or description and make a purchase we may receive a small commission or other compensation.

Good morning, good day, and good evening. So a lot of you've probably heard that Plasma is joining on with dropping the X-11 side. And I know that a lot of people that have concerns about this are sort of worried about the state that things might be in by then. So I think the easiest way to answer that question is speak to someone from the project. speak to the person who was directly involved in making this decision. So how about you introduce yourself and then we'll go from there? Yeah. Hi, I'm Zava Hoogel.

I work on Quinn, KD's compositor, and I'm very involved in all the Wayland and X11 things because of that. I work for Tech Patelan, the KD consultancy company, and have been doing stuff in KDE for five years now, and time moves fast. Doing stuff in Gwynn and all the things related to it, like Mesa, X11 protocols, ex-Wailant, and of course, Whalen protocols. And yeah, I've been involved in the announcement and the decision to drop the exorc session.

So before we go any further, I do want to just, as we're talking about, just before we started recording, highlight the fundraiser, because the fundraiser's done exceptionally well, and I think it's pretty fair to say that the majority of this

comes from actually asking people to donate, actually prompting them where they're going to see it, and right now, as of the recording of this, at least on the previous donations, page, that is sitting at almost 120,000 euros. So from your perspective, as someone actually involved in the project, like, what is that

like? Well, it's absolutely incredible, of course. Like, especially seeing so many individual people donate. Like, usually when there's big numbers thrown around for donation, it's, yeah, some big company or a single person decided to donate a large sum. but a lot of people like I think the average numbers like 25

euros or something so a huge number of people donated and sent absolutely amazing messages along with it as well these are always great to read one thing about the fundraiser page because I was kind of

confused about this when I saw it so is it correct that they were the numbers were combined on that that page where it says like 247,000 right now yeah it's combined from all the sources compared to just the fundraiser um if i recall correctly it's not even completely everything because some source like some ways you can donate aren't captured on that page but i'm i'm not entirely sure on that one but so it is a lot a lot a lot more than last year

partially because it's combined, but it is also a huge increase either way. So, as for the thing you are, you are sure of, you are aware of, the dropping of X-11. So, I think the first place to start here is, myself included, but a lot of other people had this idea that Plasma 7 is when this would happen. Oh, can you still hear me? Yes. Oh, your camera just froze for a second. Okay, we're good. That's weird.

Hopefully that... Okay, it's moving again. Whatever. Internet problems. So, yeah, a lot of people assume Plasma 7 was the date, or the, I guess not even date, because we don't even know when that could possibly be. Where did that come from?

And I guess, why was it decided to not make it with such a a major shift like that one. So I think part of the reason everyone assumes that is because, well, a big version number increased is always when we do such breaking changes, right? That's where everyone expects breaking changes. The other one is that my colleague, Blood, he made a blog post where he wrote basically, he doesn't expect it to happen until Plasma 7, and he expects it to be very likely then. But there weren't really any specific plans for when we should drop it.

A lot of people, of course, read that and went and ran with her story. Yes, this is a promise. We won't drop it until Plasma 7. Thankfully, most people seem to understand that it wasn't really like any concrete promise or a decision made by the team. but yeah so that's where it comes from as for why we decided now where you don't know when plasma 7 happens we don't either it probably would take quite a few years still for Q7 to release and then probably another year or a few years until we switch to Q7 and then plasma 7 so If we're unlucky, that would be, like, five years of plasma still needing to support X-11,

which, well, I'm sure we'll talk about why that's not a great situation to be in later. Just for context, how long after QT6 did Plasma 6 come out? well long enough that we needed to make our own cute five branch to continue supporting it i mean i can i can look it up we don't need the exact exact number of it was it was i think we started with cute 6.5 so cute 6 was out for quite a while Right, right, right. So the main point is there is like a sizable gap between those periods, like a year and a half, two years before you would even, because it's not just when you're designing a desktop, it's not just, hey, new version comes out, everything just magically works. Like, it's not that simple.

Yeah, especially when porting to new library versions and Q'd is, of course, the one big set of libraries that. we build everything upon it takes a while to first port to the new APIs and then figure out all the like more sneaky changes and regressions that caused by this and then also change our own APIs because we want to like make the backwards incompatible changes at that point right because we won't get another chance for a decade right and so last time like it took more than a year between the last plasma 5 release and the last plasma 6, the first plasma 6 release for us to make all these changes. And I'm sure there's some people who have ideas about a big breaking changes they want to

make and then it's like how many breaking changes do we want to have right now? Do we want to bring it in at like a 6.1, a 6.2 and then sort of deal with it then? because I'm sure no one like every KDE person I've spoken to who was around then doesn't want to see a repeat of KD4 where so many breaking things were changed and that's

a whole mess under itself Yeah changing too many things at once kind of catastrophic consequences and in plasma we have a lot of freedom changing things between versions because it's not so much a set of libraries as it is, well, the desktop and as long as the components of the desktop are updated and sync, it's fine if we do breaking changes, but with our frameworks, there is no backwards and compatible changes ever. Like if we can at all avoid it, that just doesn't happen. So when we port, we need to be really sure about.

all the changes that we want to have. And that can take a long time. Right, right. So what does the current timeline for dropping X-11 look like? And I'm sure this could be subject to change, depending on what happens when you get there. Yeah, so in terms of plasma version, the first plasma version that will drop it is plasma 6.8,

and that will be released towards the end of, 26, but 6.7 will still be supported throughout the whole lifetime of 6.8. That's at least what we're targeting. We don't have the exact release schedule written down yet. And we have changed the schedule a bit to add more bugfix releases, which we wanted to do anyways. But now we certainly make sure that it happens before we drop the X-Ox session. We're going to address this once

and I'm never going to mention it again because I don't want to talk about it again How many people were aware that announcing that something's going to happen with version 6.7 is going to have a lot of annoying people spamming 6'7 Well,

I didn't see too many comments about anything like that I saw one comment being disappointed that we didn't announce it with plasma 6.6.6.6, which is a shame, but nowhere.

Good. We're never talking about that again. I just, I had to read it up because I know someone's going to ask about it. So, okay, once the version comes out, my understanding is there's going to be support for that version

up till 2027, is that correct? Yes. I would estimate February 2027 is when the last bugfix release for Plasma 6.7 will release. The exact date will find out once we get to the release of Plasma 6.7. So why announce it so far out? You know, like a lot of, you know, you could just be like, okay, well, next version, we're just going to drop it. that's, three months away, it's done. Like, why give a leeway for it?

Yeah, so aside from the few people that got confused about, hey, we're going to drop this, and then immediately thought, oh, this is happening like next week. Right. For most people reading the announcement, it is a warning. Hey, we're making this change that might be affecting your system. In a year, so you have time to prepare, time to annoy us about things that you think use their need that aren't supported or aren't working for some reason or another. It's also time for companies to update their applications. Like there are some stragglers that don't support the Wayland session properly because as an end user, if you use the application, you can't just log out.

log back into the X-11 session and then it works, even if you have some other downsides from that session. A corporate customer doesn't really care that much if you're a little bit inconvenience. But with a year in advance, I hope that's enough time that everyone can prepare. I think a better question then is, why do it now? And let's not, let's say like not three years ago, for example. what was stopping it happening earlier? So there were some people advocating for doing it with plasma 6 already

because big version change, breaking changes are expected. But we didn't feel like we were quite there yet. Like the wheyland session wasn't even the default yet upstream. Like Fedora had it by default. for a while then and it's been working mostly fine, but when some plasma developers still run the X-Rex session and some distros, even after Plasma 6, still defaulted to the X-11 session,

that is a clear sign that you weren't quite ready yet. At Academy, however, I went around and I asked people, like, what do you think about this? drop the X-Ox session and, well, I think there was one person in the room that was like, they use the X-Ox session for session management, I think. That was it. Like that was the only concern. And everyone else was, yeah, I'm looking forward to this.

And in terms of application support, we are of course also at a point where most apps also do support the Wayland session, like Discord support screen sharing in the Wayland session now properly through the XTG desktop portal, Firefox and Chrome support it properly. We have now an experimental bit of CRETA that does color management on Wayland, and it is so much better than it ever was on X-11 because you don't need a lot of manual setup to make it work. So we have some feature lists that we maintain. One is the significant plasma when it issues page. And the other one is an issue on the Quinn repository.

We kept a list of features that we know these, some of these might be really niche, but We need to support pretty much everything before being able to drop because while some people will always be inconvenienced by a big change like this, we need to minimize the damage. And I think it's especially important as, like, from all the metrics I'm seeing, plasma is on this sort of, this like upturn in usership. Like, there's, um, on, on Arch Linux, you have packaged stats. And I just checked it the other day. And it's just been gradually, slowly climbing. Everything else seems to indicate the same thing as well. And it seems like people are mostly happy with the direction that plasma has been going over the past, you know, five or so years.

So putting a stop to that doesn't really, at least to me, seem like the best of ideas. Yeah, I think if we would have dropped it with Plasma 6, like especially where the Nvidia situation wasn't like completely there yet with some remaining issues like explicit sync not having been solved, we would have lost a lot of users and that would not have been great, like, even just, like, as a desktop environment, we don't necessarily require, like, acquiring the maximum amount of users. Like, that's not really what we're doing, but it would make life a lot harder, unnecessarily for a lot of people.

Sure, yeah, yeah. You don't want to, like, you don't just want to make people angry for the sake of make people angry. Like, I understand that it's not operating like a lot of. business but you do have there's some level of responsibility when you have a large user base to not just change things for the sake of changing things yeah there's always the need to consider the trade-offs between what we're doing how much additional work we need to put on to support some use cases and how many people

we were not just like anger but like break their workflows and in some cases cause serious issues like if you had flickering with nVIDIA that can be really really problematic so we needed to be absolutely sure and cover as many things as we can so with these significant issues is the goal to get everything on this page completed before release or are there some things that can be you know pushed on until later i mean that's the goal i somehow like i don't think we will necessarily get everything completely fixed like session restoration depends a lot on application support yeah so i doubt like on x11 it's not very different a lot of apps supported a lot of apps don't support session restoration

So we will have some, like, sliding scale of how much we have solved a specific point. But the list should definitely be pretty much empty then. And it is so much shorter than it was. And even recently, at Academy, we cleaned out a lot of old points that had been fixed ages ago. And a bunch of points on that list, we have already. the, like, solutions pending. Like the whole screen section. Yes, that one we're targeting.

Some people at Google have recently, because of these announcements, started poking us about, hey, they need to support it here, what changes we are missing, what features we still need to add to make it work. The whole screen section,

I want to remove in plasma 6.6 already, like output mirroring I've been working on for quite a while, and the result is, it already works so much better than it ever worked on X-11. So I think, like, there might be one or two points left on that list, but most of it should be on. Yeah, I remember bring this up when Nate was on, honestly, last year, and I do recall there being considerably more issues on this list. When it comes to the headless RDP problem, my understanding is that is tied into the new login manager being worked on.

Yeah, so there are some, there are multiple possible approaches, and we're, I don't know which one will be like the first to be implemented. One of them is you have a background demon that runs in the login manager and that allows you to log in remotely. And when you then log into your session in the login manager, it like transfers the connection to the demon in the session. The other approach is to just start the demon. And the demon currently gRDP can allow you to log in with your normal username and password. So when you do that, it could directly start a session for you without leading to go through a second login manager. I don't think we have any prototypes for either of these approaches yet.

as yet, but they're all things we're considering, and hopefully in a few months we can announce that we have something that works better. Because this is one of the issues that does get frequently brought up any time I talk about any time, I'm sure it's just as bad for most of the KDU here. Any time I talk about Wayland, I get lots of, lots of fun comments. Some of them there's an interesting problem when it comes to anything related to

Wayland or anything relatively new because there's a mix of users who are not aware of what is going on and a mix of users who are technically correct because they're running an LTS distro where they don't actually know the current... It's the same thing

in the Gnomes space as well. Ubuntu 2404 is very popular and it really skews the state that people think things are in. So I'm happy to know that headless IDP is actually a problem, like, legitimately being thought about, because for a lot of people, that is a really important thing. Yeah, it's one of the strongest feedback points we got from the announcement.

Before that, like, we were aware, yes, this is a thing we need to fix. But it was like, yeah, it's a really small niche case. But from the responses we got, it's one of the biggest reasons for people to stay on X-11, or at least use the X-11 session for the headless login. So, yes, that's very much considered important. From your perspective, sort of how do you deal with that feedback coming from the LTS users? because obviously they're going to have a, the older the LTS gets, the more skewed their perspective is going to be. And you can't go back and fix problems on the LTS.

So like when problems come in from those users, how do things like that get handled? How do you try to like inform the user on what, you know, what the thing you're using is supposed to be? Yeah, it depends on the, you know, the, you know, the thing. user, like a lot of people providing feedback are thankfully just not on LTS distros. I feel like most people that are interacting with upstream actually use a somewhat recent version. Some of the people I see in comment sections on Reddit and stuff are just, I tell them, yeah hey we fix that it works fine and then most are just yeah okay cool then move on no problem um i feel like sometimes the lTS and like debiln stable do shift the like to change the

image of plasma a lot because a lot of these don't even ship our bug fix really So when it comes to, hey, missing features, that's a completely separate concern. It would be nice to just have bug fixes. But features I don't think we have too many, we don't get too many complaints directly. It's not the main thing to talk about here, but obviously that's part of the reason why things like KD Linux are being worked on. are being worked on as a way to be like, hey, this is KDE. Here you go. Have a look at it.

It's good. But obviously, distros like Fedora do a good job of packaging things. Arch Linux, even if they have packaging issues sometimes, they tend to be relatively up today and, you know, Gen 2. And there are places you can see modern plasma. It's just a lot of people are clustered around those either point releases or the longer-term ones. yeah uh like kubuntu 24 lTS is still on plasma 5 so there's that right um from what i hear though kubun the next lTS version won't even come with an x11 session out of the box like you

i think you could still install it but like it's not officially supported uh So the distros, even the LTS distros, are moving faster than us in this regard. So that's also a part of a very nice feedback that we got before making the decision. Like Fedora has been there for a long time, but that's normally the case. But other distributions. Like if even Kubuntu is making this move, we are certain this is the right one. Yeah, some argue Fedora makes the moves a little bit too fast. I think on Workstation, they defaulted to Wayland in like 2015 or something, which was a little early.

Yeah. Yeah, that was, I think that hurt the reputation of Wayland quite a bit in the beginning. Like, to be fair, Nome also changed the default upstream way too soon. Like, if you can't take a screenshot, how can I take that seriously? But, well, they got there eventually, and I think Fedora helps a lot with just kicking people into finally doing something about technologies they should have perfected a long time ago. So what about KDE apps? Is there anything changing on that front? If a KD app developer still wants to do X11 stuff, can they keep doing X11 stuff?

Yeah, so most KD applications, I don't expect they will drop X11 support for a really long time. Thankfully, for most of them, it is really well abstracted away. Like, there might be a few extra checks in the main CPP file, but that's mostly it. like we have our frameworks which has really long API guarantees and that will support X-11 for a really, really long time. So I don't think anyone has to worry about it that much. Some applications might change, like Spectacle, I would expect, will drop X-11 at some point, but it is also part of plasma.

Right. So is there a different sort of case there for the core, plasma applications or is it just specific ones that don't really make any sense to have on x11 anymore that might drop it yeah so spectacle is pretty closely tied in with plasma so there it's a bit of a different story but like dolphin will definitely support it for a long time to come system settings doesn't really have any reason as far as i know to support any specific uh window system. But if one does come up, it might just drop it because it doesn't really make sense to use it outside of plasma. Sure, sure. So you mentioned there's a lot, like a lot of

guarantee with frameworks. Is there any discussion there on when the X-11 stuff could be dropped? Is that like Plasma 7? Is that whenever Qute drops it? I think that discussion will happen when the KD Framework 7, like porting to Q7, will happen. I don't think it would make a lot of sense before that. But if Q'd 7 takes 10 years, maybe we'll see it before that. Right, right, right. So, why, I guess, why drop X11? like what is having it around causing issues with slowdown making not possible

like why not just let it be in the state it's kind of been in where it just kind of sits there and just rots away yeah so two big reasons i would say one is actually the code and i will get to that the other is bit rotting code will break like just in general and I have to like when making this decision one of the worries I had is if we just leave it we will be technically supporting but actually not supporting in any way of like the X11 session and when users use it and everything is broken that hits us of Of course. And that's not a user experience. We want to give you this. I recall that Nome released and like one Nome release.

I think it was Nome 48 released with broken Xox support until they noticed, like because no one tests it. And then they decided when to drop it. And I did not want that to happen for Plasma. because, well, that's just shitty for the users, instead of being switched to something that works. You just can't log in or it crashes when you log in or whatever. The other reason is on the code side,

we have a lot of code for X11. In Quinn, we've sort of mitigated that by splitting Grin, but we have a screen locker. We have a case screen for the display settings. plasma shell itself is a huge amount of code and has to deal with all the annoying X-11 specific bits. And from a development perspective, having to deal with that is really terrible. But more importantly, makes all the code more complex. And complex code is generally much more error prone and causes bugs for Wailant users.

What it also does is it brings us down to supporting like the smallest overlap between Wailand and X11. So if on X11 we can't do a feature, but on Wailand we can, then you either need to split up a bunch of code and see that. that both keep working separately. And then nobody tests the X-11 version, of course. And in some cases, it's just way too much effort to do that. So some features just don't get implemented at all or get implemented in a worse way.

So to give a few concrete points of what we would like to do once X11 is dropped, we have I've considered splitting the wallpaper into a completely separate window or even process. So right now the wallpaper is rendered by plasma share and the folders and widgets and stuff are rendered on top directly in the process. So what we would like to do is split it into its own window. So Quinn could then, when you lock the screen, reuse the same wallpaper instead of the lock screen loading its own wallpaper and then like taking more time to look at the screen obviously but also having its completely separate settings and less capable settings compared to the normal wallpaper that has been a real big source of confusion for a lot of users because they

set the wallpaper and it doesn't apply to the lock screen and it can't automatically apply because the lock screen has fewer features for the wallpaper like you can't set a wallpaper per screen because it's the security critical component right we wanted to keep it as dumb as possible so if we can split up things in a way that is at least challenging to do an x11 especially if you need to support the no compositing case where transparency isn't that thing we will be able to hopefully improve start-up times with that and make the wallpaper features work transparently across the desktop and the log screen. There are a few other things here and there, like radically simplifying the lock screen.

So it starts up faster in general and it crashes less often. Thankfully, it just doesn't crash as often, like lately anymore. We improve that a lot, but I would still be much happier if the lock screen wasn't. So such a gigantic pile of layers and top of layers with everything being super brittle and terrible. So obviously, most people don't enable. KDE telemetry, but from the, I don't know if you've looked at it recently, but from the numbers that are there, how many people have been using the X11 session? Obviously, again, these can be skewed because it's not named by the fault, for anyone

argues about statistics. What do the numbers say? Yeah, so accessing telemetry is a fun topic because I don't think I have access. No. Okay, well, at some point I talked about it with David Edmondson and like, there's a website I can go to, but all the actual pages on that website where you can look at the statistics, I don't know, aren't unlocked for me or I need to figure out how to add the correct pages or whatever. But the last numbers I heard was that over 80% are in Wayland, over 80% of plasma 6 uses. Right, right. And roughly a bit less than 20% on X-11. And those numbers are a tad skewed by the fact that there are some distros that still default to X-11.

One of the biggest ones, of course, Steam OS. Yes, yes. That is changing. That's getting fixed. But yeah, as far as I know, we unfortunately do not record the distribution in the telemetry data, which is, and adding that is a whole, yeah, kind of firms that we can't really open because of how the telemetry system is designed. So my estimate would be 10 or 5% maybe that are on X11 by choice rather than by distro report. Right. And that's a big distinction. That's a really important distinction because it's kind of the same way with Wayland, where a lot of people are on Wayland, not by choice. It just happens to be what their distro uses. And it's good enough that they don't think about it. Yeah, so most people don't change the defaults as a general rule unless they find something that is broken. I imagine some part of the X-11 users that chose to opt into X-11 are also like they opted into X-11 two years ago or like one year ago.

because something was broken maybe they used an invidia driver that had a regression and they switched back and never thought about trying well and again because if x11 session works fine for them why should they change which i can't understand like most people use the computer for work and not for figuring out window systems but yeah it would be it would be interesting to see the numbers with a better telemetry system and up-to-date numbers as the straggling strows switch to when and by default. We'll see how much data we can collect about that before we actually just drop it. Then it will be a very easy statistic at least. This is kind of just an anecdote, but I do know people who when the

A Winsplit happened, who were using X-11, because Arch didn't just install it for them, they were just dropped into the Wayland session. And that was a moment to rethink whether or not they should even bother going and using the X-11 side, because they just hadn't tried it in a couple of years. And it just, it was better than they remembered. It, well, not just better than they remembered, it is, it actually is better now. I would hope so Definitely hope so

But I can't speak for anyone else But I'm sure some people For them that actually was a moment To rethink what they were doing Because it did force them for a moment To just try out the Wayland session And just Actually try it

I think that's the problem for a lot of people A lot of people have these opinions on Wayland Without legitimately giving it a shot or without giving a shot in any recent times. Yeah. I've seen a lot of messages from people that are like, yeah, so they recently found out they've been using Wayland for months and they didn't notice.

So that's the best case, of course. Like, if you even know about that, but you still don't notice, then, yeah, we're doing a perfect job then. But, yeah, a lot of feedback, like the very unactionable vague feedback about, oh, when it is bad, comes from people having tried it years ago and people trying it on LTS distros now, or even on like, hey, I've tried this on Linux Mint, the Wayland session is experimental, wonder why some things were broken. That's crazy.

Yes. Like, I hope Cinnamon go as well with getting their Wayland version ready. And they tend to be fairly conservative when they make, you know, the UI has looked the same for the past 10 years. They're still on a mix of GTK. It's mostly three, but I think there's still some GTK2 apps in there somewhere. Cinnamon will get to it when they get to it. But on the topic of trying things out a while ago, invidia support

invidia support is a big one because just up until really two years ago it was pretty bad where are things now so assuming you are using a GPU

that Nvidia still supports which means some really old GPUs you need to use the open source driver you can't get around it. Nvidia has all the things in place that should make it work smoothly. Like there's explicit sync support, which makes sure you don't get flickered to death. And all the necessary graphics driver bits are there.

And they've been helping out upstream as well, like actually implementing all the things that we want them to support. actively looking for feedback and even contributing to compositors in some cases like explicit sync support for pipe wire for screencasting was something that Nvidia implemented for us that we didn't need to do ourselves so I think they're doing really well and I haven't really found any issues on my laptop with an Nvidia GPU so I'm pretty happy with them what do you happen to have in that A RTX 3060.

Mm-hmm. So when it comes to the... A laptop version. Yeah, yeah, sure, of course. Yeah. That was pretty chunky laptop. So when it comes to Nvidia, my understanding is old cards with the open drivers work pretty well.

2,000 onwards works really well, but there's like a few things in like that weird middle period that's pre-GSP but not old enough to be well-supported but because there was that period where they were moving stuff from the drivers into the firmware if I recall correctly and there's a couple of cards in there which still don't work at least don't work well that's that's my understanding at least I could be wrong yeah so there's the the 1000 series which has the firmware locked so you can't use it with the open driver but the proprietary driver supports it pretty well like they are missing some specific timestamp information

that they just couldn't use with the proprietary kernel modules but generally that just means we have like a little bit increased latency it's generally not that big of a deal and everything should work fine but you can never expect Nouveau to work properly on that driver that's not going to happen It kind of all comes back to what I was saying before about like LTS and point release distros

but it's like you we still I think we still need a couple of years onwards for those to really be resolved because there are people who are going to be running even if it's just a kubuntu

you're still going to be a little bit behind and hopefully by the time we get the next version it should be at a point where they're shipping something sensible but obviously you're going to still have the Debian users and you know all that stuff

where yeah there's going to be who are still trying to run older stuff who don't you know they're just not getting the fixes basically and yeah yeah but the

the way and only version of Kubuntu will be in like 2028 I think like so they still have a lot of time and I think that's

going to be fine on Debian well I personally would strongly recommend against using KD Plasma on Debian because they like I said they don't ship any bug fix releases so if you have issues you're

on your own, your bug reports will be automatically closed in our bug tracker. At least use Debian testing or something if you want to use Debian because that will ship at least a newer version. But yeah, if you are using an LTS distro, you've got to rely on your distribution to make the right choices for you if it doesn't maybe switch distributions. what do you personally recommend to people i personally use fedora kitty because they are always quite responsive in fixing issues and also keeping plasma up to date uh like they are often the very fastest to update plasma like even faster than arch um and but they also make sure that things don't just break

I think they're doing a really good job yeah everyone I've spoken to at Fedora it seems like they really care about doing stuff I've spoken to Neil many many times yeah it seems like it seems like the projects in good hands and I've not heard of any major breakages on Fedora in a while

I'm sure if something happened I probably would have heard about it by now but it seems like things have been in a good state for a long time. So when all this was announced, I was what was like two or three so at this point, what was

the expected response? So we looked a bit at the response to the Nome announcement and used that kind of as a reference and And I hoped it would be somewhere in that direction because that was surprisingly positive.

But I also feared that, well, a lot of people that were responding negatively to the Nome announcement went like, okay, I'm switching to plasma because of this. And I kind of feared that a bunch of these people would then go crazy because, well, we are avoiding their plans. But it went pretty positively. We got a lot of good feedback. And except for the DDoS on KDE infrastructure. I did hear about that. Yeah, that was the one big incident, which our Sissatman team handled really well. Thanks to them for like taking care of of that, but it's sad that they had to do anything

other than that it was more positive than I expected. Yeah, I don't know what they were thinking with the DDoS, like, oh, if we DDoS them, they're just going to revert the change? Like, what was, I would love to speak to the person

who thought that was an idea they should do. Yeah, it was targeting specifically the announcement page. from what I saw. So maybe they hope to just take it down and I don't know what they wanted to achieve with that, but, yeah. Yeah, okay.

Because I heard about it. Neil sent me a message about it because a bunch, obviously, you were all talking about it happening. I didn't know it was specifically targeting the page. I was like, okay. Well, like, is it just a badly time one? No, yeah, obviously.

it's targeting the page, yeah, it was very intentionally about this. Yeah, that was dumb. That was really dumb. I don't I don't know what I was to say about that. It was just stupid. Ow.

So, there are, how do I how do I describe it there are this there's this contingent of people that feel as if Wayland is being forced upon them I think it's dumb

but how do you respond to that well I wish all the rumors were true and Red Hat would pay us tons of money to do this but unfortunately I have not received any money from Red Hat so I have to refute those claims about like Wayland being false than them I can sort of understand the attitude like of some people there not necessarily how they

express it but I can get the feeling thing is we are not a big cooperation with infinite money that goes and just forces like an AI assistant on you because we hope to conquer all the markets with our AI and like crash the stock market with it. We do that because we think that's the best thing for our users. Like in the end, part of it is our users and part of it is our community. Like, we need to do things that are good for us because, well, that's kind of the point of a free software movement. And that's good for our users. And in these tradeoffs, sometimes some people just lose.

And that sucks, but we can't. We don't have infinite money and can't make everything perfect by everyone. Not even companies with infinite money that actually try to do it can achieve it. I think we have pretty much the same perspective here. Like at the end of the day, no one's trying to like eliminate the exorg code from existence. And no one's trying to like delete all of the like all the window managers and all of the desktops that are still on X-11. So if you really do

support X-11, like, distro is going to have X-11 stuff package for a long time. Arch is not getting rid of the package. Who knows when? And even if they do, it'll be in the AUR. Debian, it's going to be their minimum of another 20 years,

so, like, I think you'll be fine on that front, let alone all of the little wind managers they have. Then there's Gen 2. Obviously, if you want to, you could compile the code yourself. like there are options and I think sometimes at the end of the era someone has to write the code and you're not always going to get your way

if the people writing the code don't align with your goals and if that's if that's the case like there are other environments you can use go use those distros go support those environments If you think X-11 is the way we should do things and Wayland is dumb, actually go and put your money where your mouth is and go and support those projects and show that there is a drive behind that. That's honestly the best thing I can say about it.

Yeah, I doubt that like open box or like X-11 window managers will go away anytime soon. And to go back to the rumors about Red Hat, pushing anything, Red Hat is still maintaining the X-Ox server. They are still working on it. They're maintaining it, not like adding features, but just keeping sure it works the same as it does before, which is, I think, the most important thing to everyone still on X-org because if they wanted new features

instead of it working exactly as it does they would switch to Whalen I hope or some other alternative I guess so in the past there have been forks of KDE there are things like Katana

there's Trinity do you expect to see a serious fork of Lysma 6 pre the X-11 dropping actually happen. I know obviously people will be like, I'm gonna fork, I'm gonna but like, you know, people

don't actually do it. It's already happened. So it's called Sonic DE. I don't know too much about it. Like obviously they are allowed to do it. It's free software, just

if you want to do something else with the code go nuts. it's all right um the one thing i will say about it is that they seem to have some association with x libri so that doesn't fill me with confidence about the project but like i said i haven't taking a too close of a look maybe it goes in the direction that some of the users that insists on x11 would like to see and i wish them the best of luck Yeah, like it's, there's a lot of projects that will fork and then maybe stick around for like, you know, two or three months. But like, it's still a question of whether it's going to be around by the time you guys even drop X-11.

Yeah, only time we'll tell if they're serious about it. And if they will be sued for the name, we'll see. Like, obviously legal stuff there, but like, if you're going to fork a project, just rename it. Like, that's, that's the simplest thing. If you're going to do, like, an actual, you want to do a hard fork, just pre-name it. Like, that's, that's, just, yeah. Yeah.

They did, like, they did need reminding that they can't use our trademark name for, like, their project. I'm talking about the new name, which I doubt that. anyone, like, will sue them for the name from the Hedgehog. But, yeah. Yeah. That's out of your hands now.

Yeah. They renamed away from KDE. So from that point, like, we don't have a problem with it. So, okay. Um, but there are things outside of, the major blockers, which are often brought up as like issues with the Wayland. An issue, not necessarily with the Wayland approach, but with the state things are currently in. There was obviously the recent in video talk where there was the discussion of how there's

the screen capture dialogues, which can get in the way of, you know, automated capture and stuff like that. And I guess that we kind of brought this up a bit earlier with the whole remote desktop stuff. Is there some thought on some pre-authorization that could be done here? Any sort of approach that could be done to allow for some automated. I know why it's like this. So, you know, you can't just write a Python screen capture that takes like three lines of code to do. But is there any thought on like how to do this in a safe way? So with the remote desktop portal, we already have something built into KD for that.

So you can in system settings, just select the application in the application permissions, KCM, and just say control, pointer and keyboard and share the screen without prompts. And then it will never ask that application to, or will never ask you to, to, approve for screencasting it's a bit more involved because for the application like a permission is easy like it's just a yes no but which windows which screens should it actually share this is something that we may eventually look into with the remote desktop portal as well because it would be nice if you could stream all the windows or just one window instead of having to stream the entire screen, like RDP can do that with an individual window.

Right now, how it works with the screencasting portal, the user selects which windows or screens or whatever, and the application can store a key and use that the next time it requests a screencast, and then the compositor will go, I know you had this. window the last time you have the screen and just restore all of these. How much that works for automation is of course a very different story where you might want to script some things. And I think we still need to look into what APIs to expose there. How do you want to identify a window? How do you want to identify a screen there? There are some things we can do like where we can be clever about it, and there are some things where I think in the end we just need to

provide you with a command line utility or just some API that you can call and then do whatever you want with it pretty much. Because for automation, that's what you need. That's what everyone wants. Well, we're at a point now where when I open up OBS, it remembers the screens that I captured. applications I understand harder because if it's the same application session it can grab it but if you've done a you know you've closed the window you've done a reboot in between even if it's the same application like you know it's another Firefox window for example it's not that same Firefox window so it still prompts you again for it yeah so our desktop portal is actually like somewhat clever about it

It will check, like, window names and which application you actually selected. So it should, in general, restore the correct window, even if you close the app and open it again or reboot in between. If there's a case where it doesn't, make a bug report, please. Like, in general, it should just work without any user interaction. Okay, I have to give us some more. test and just see where it goes weirdly then. Yeah.

So we talked kind of on session restoration earlier. What is kind of what is kind of the issue there? You've mentioned the toolkits, but my understanding with that one is there's like kind of wide agreement that session restoration is kind of a good thing,

but it's kind of complicated to like get all the pieces in place for it. Yeah. Yeah, so the most difficult bit is we need a way to identify a window after you close the app and then open it again. Because from the view of the compositor, you open five Firefox windows. It knows there are five Firefox windows. It does not know which one is which, which tabs are in it. it can maybe guess depending on the window title but especially with a web browser it loads a page it changes the window title depending on the page it might change the window title like every five seconds it's a difficult thing so how the protocol basically works is the app test the compositor give me a session and add these windows to it and then the application

gets a unique handle for that session next time it starts up it can tell the compositor here's the session I want you to restore these are the same windows as last time here's their IDs and then the compositor can do the rest there's the second part that is actually starting applications however which on X11 that works by the app telling a system service an arbitrary command to to run on startup and that arbitrary command will then have by the app like the session id and like the correct command to use that session id in it obviously that's really terrible and there's no way you can sandbox anything with that approach so there has also been a lot of thought gone into

how to do that in a less shit way so one of one of the ideas was you just use flat pack like if the application has a flat pack you run the arbitrary command the application provides but you box it up first so when it runs in the flat pack it can only do as much damage as the app can do anyways when you started normally the other idea was well maybe we don't do that and standardize on some key in the in the launch command like dash dash session which would conflict with some applications however that have their own session arguments there will be a solution eventually And both of these issues have like some implementations already. You Q'd, I think, has for Q'd 6.11, the experimental session management protocol

merge and Quinn has it. So then you will be able to set an environment variable. And it should hopefully, by that point, work with all KD applications. we see how much needs to be changed to work with the Wayland approach versus the X111 and how many other applications

adopted. Right, right. Because it's easy enough if Ghanome suggests hey, we do it, and then you get all the Gnome Circle stuff in, but there's going to be things outside of GtK and outside Gnome and outside of KDE that now you have a whole a whole another series of applications

that have weird different toolkits are built directly on OpenGL and yeah that's my understanding why it's like kind of complicated yeah so we are still considering maybe adding like some clever detections about like using the window title

and the application ID to maybe at least start the apps and put them in the right spot, even if the actual content in the app might not be restored, like not the correct file open or whatever. But it's very difficult when apps can have five windows open, and we just don't know which is which. And yes, especially like electron and we have, of course, chromium and Firefox. These are the most important ones.

which we'll need to support this. And Firefox, I'm confident, will, like, be very quick because they do X-11 session restoration on Wayland. So, yeah. Okay, okay. It's a whole separate, like, protocol from X-11. It's completely crazy how all of that works. but yeah it might take some time

like Discord screencasting I think took six years from the screencasting portal being added to Discord supporting it so supporting the session restoration which is an optional like super nice to have feature instead of core functionality might take a decade

we'll see well my understanding with why disco took so long is they have like one Linux guy? Yeah, I mean, that's, I think the problem with a lot of the apps from big corporations. If they even have one Linux guy, or if it's just like with Spotify, like one or maybe multiple people in the team that just do it on the side as a hobby and can't really use work. hours for it. Like, that's what I understand the Spotify situation is like, at least.

Yeah. I'm glad that things are slowly moving on that side, but, you know, when you're dealing with these proprietary apps that just don't care about Linux first, there's only, look, there's only so much you can do there. You can provide all of the tools to make it happen and provide as much documentation and make it as smooth as possible. but at the end of the day, they have to write the code. You can't go and submit patches to fix it for them.

Yeah, I mean, sometimes we can, like with Q'd. Oh, sure. A bunch of applications just automatically got at least decent violent support. And I hope that Electron can add session restoration as well. So we don't need Discord to do anything. Only update the Electron version, which can be only a matter of years. So I hope that we can get a bunch of apps to support it that way. But at some point, like with Session Restoration specifically, you need to identify Windows. If the toolkit doesn't

know which window is which either, then you're just screwed. Right, right, right. That's the case with Q and GDK, by the way. There's no window ID, no nothing. Cute can, I think, generate something that's not completely reliable in some cases. But it's probably like there's like a niche IPI for X-11 session restoration that we're probably going to just reuse or maybe add something in KD frameworks. I don't think it's all wrapped up yet. So you mentioned not bringing back the content to the window. So there's this like distinction between the fake.

and real session restoration. And I'm actually not using plasma right now. Do you guys have the fake session restoration in there? Because I know there was a discussion a long time ago. Yeah, so we start apps that were open the last time. Right. We don't currently try to restore the positions. That's something I want to look at into at some point,

but with how challenging that can be. and then working for half the apps, but not the others. We'll see if it's actually feasible. About the distinction between the content and just the window, it's more like that one, the application has to support directly. We can restore the window position, like even when we can do that fine with the Waylon protocol. It doesn't mean the app like saves all its internal. kind of state when it quits.

If the app just throws it away every time it quits, it probably will not do perfect session restoration. Yeah, there's obviously apps like browsers which do a really good job at that, but like a terminal, for example,

most terminals don't remember their session because it really doesn't matter. Yeah, I think console can at least restore the tabs and the like directories. but obviously if you had a command running in there it will not start a command for you you will have to go and do stuff yourself

I know so many people that don't use plasma but constantly tell me how great console is so clearly I don't know who works on console but whoever does they've done it a fantastic job with it yeah it's quite good so a far more controversial protocol will say

the whole issue with the positioning now I know that through like K-Win scripts and stuff and like obviously you can do the window rules to place windows like there are ways to do it in plasma but I don't know how much you even want to get into

like the whole the whole multiple thousands multiple years of comment that is that discussion? Yeah. So I think it's a very important distinction to make there between the user positioning Windows and the app positioning is Windows.

Like the user positioning Windows is something we 100% support. Sure, yeah. And yeah, so like, that's the entire point of having a compositor, right? Sure, sure, sure. Or window manager in the X-11 word. that the user is able to customize that to any degree they want.

And with Quinscripts, you can do absolutely crazy things, like a replica of the scrolling compositor thing. Yeah, yeah, I have seen that. That's super cool. Obviously, if things like Cromtite for tiling and a whole manner of crazy things you can do. Yeah. applications positioning themselves is however a lot more controversial partially just because

well lots of people confuse it with session restoration like I need to clarify idea that the way in session restoration isn't just about you log out and you log back in but also you close an app and you open it again. It's restoring the application session and the window properties it had the last time, not just like your whole computer session. I feel like most of the people are asking for apps to position themselves are actually asking for session restoration. Like they want their apps to be where they were the last time when you open them again,

which is completely reasonable. and we're on a good way to make that a reality. Applications positioning themselves is generally, at least the cases that I would say are reasonable are like scientific apps that do super special stuff. You have one window with a button in it, you click that. It starts a Python process with some script

that's not even using the same toolkit as the normal app. and then spawns a window that should be, like, right above that button or something in that direction. And the controversial bit is mostly that if we make these applications, like, possible with native wayland, well, Electron will just add support for that, for normal apps, right? Are you, on that note, are you aware of the W3C discussion regarding letting browsers spawn additional Windows? Oh, yeah. I don't fully understand why it's happening. To put everything in the browser, that's a lot. Every browser I know blocks apps from popping, like from opening new pop-ups and Windows to a degree where it sometimes breaks bad

websites. Yeah. Yeah. So that seems like the next extreme level of don't do this for websites. I know they want to have like super crazy productivity apps in the web browser judging by how well Autodesk Fusion works in a web browser I would say it's a bit of

a moonshot like but yeah in general we I don't think that discussion will end anytime soon right my concern there is

we need to find something some solution for these apps. Sorry, I keep trying to speak a iPad, sorry. My concern there is that if that, like, discussion goes through with the browser side, then it becomes a thing where there actually will be apps that, like, there will be web apps that kind of expect it. And then it's like, what do you do there? It's like, at that point...

I feel... Yeah. I feel like I don't have to... worry a lot about it because, well, like I said, web browsers block just normal pop-up windows. I do not think browsers will just make, like let apps do that just all the time. It will be at least something that applications will have to support not doing that, even if it's just for, hey, I'm using a browser on a tablet. There ain't no multiple windows here.

So it needs to do something sensible. Right, right. And the majority of software currently is also like Windows oriented, right, where they can do all the positioning stuff. But the vast majority doesn't do anything crazy with Windows because generally it's a much easier user experience to do one window where or like maybe with some dialogues mixed in instead of a crazy interface with five windows that the user needs to drag around and like be mindful of which windows above which other window and there's a reason gimp does

one singular window by default nowadays instead of the three split windows yeah yeah um it so it's sounds like to me you're relatively apprehensive about supporting this protocol yeah so i think we need to find some solution for these like niche edge case apps i think it would be really good to support more scientific applications obviously it only has advantages but we need to find a way that works without bringing along all the problems we had on X-11. Like, my favorite example is K-KAT. K-KAT is a great app. I use it regularly.

But the one thing that really annoys me about it is that it tries to be clever about window management. I imagine on a lot of setups that is totally fine, but I have sometimes use my laptop on the go and sometimes I connect my external monitor. When I start Kiket, it remembers its last position. So I started on the external monitor. It remembers it was on the laptop the last time. So it starts on the laptop display. I move the project window to the external monitor. Open the schematics window. it remembers I had the schematics window open on the laptop display the last time so I go and move that up I add a new component that's a new window which remembers its position separately it starts on the laptop display I see I changed the property of a component same thing it has like 10 different windows each position remember separately and it always starts on the wrong screen so like that's

That's the kind of thing I want to avoid, because with session restoration, I can just, in the compositor settings, say, don't restore their positions, and then you're done. We can also be smarter about it and say, don't restore the screen it was on. If the app goes and says, restore the session, it should just move the window to the screen I'm actually currently working on. are a lot of things we can do and be smart about if it's better than just we throw a position over the fence and the compositor has to decide whether or not to break the app or accept its position. So that granularity is possible with the Session Restoration then I actually wasn't aware of that.

Yeah, it's all completely in compositor control, right? So if you want your positions to be remembered, you can make the composite to do that. If you don't want them to be remembered, or if you want nothing to re-remembered from the last session, you can tell the compositor to forget everything. Like, obviously, it depends on the compositor to add that option. On the positioning side, at least, I will make sure that option is added because, like I said, it annoys me a lot when apps get it wrong. Okay, I think you've given me like a relatively new perspective. on this

protocol um I yeah no I I at the end of the day I hope that what we have is use cases that make sense

obviously like there there are there are there are dumb things you can do um are you aware what is it called um there's a game called window kill like there are there are dumb things you can for anyone who's unaware

this is a game which will take advantage of the fact that you can position windows on windows Windows and yeah, that makes sense and you actually shoot at different windows and it will like, it fires projectiles between, it's obviously

just abusing the fact that you can do this. It's super cool but like that's, whilst it's cool, it's not necessarily like a real legitimate, you don't want most apps to be able to do this, right? Like this, it's cool that you can, but like it's more of an

oversight rather than an intention. Yeah. As far as I know, I don't know if it was that game or a similar one, they can also just open a normal window that's like transparent in the background and just render a bunch of windows inside of like the actual real window. And just like you don't like for a game, that's fine. Sure. It might look a bit less fancy, but it's okay.

It's a fun game, so. So, KDE and the BSDs, because BSD is a, the BSD world is like, how do I say it? Unix nowadays basically is Linux, right? I think for most,

that's pretty much where we're at. There is still some corporate, old school systems and there's, oh, can you, are you gone? Do we lose you? You back? Yes. Where did you stop hearing me?

You just started saying Unix is Linux. Oh, yes, yes. Most of the Unix usage nowadays is Linux. There's still some of that corporate older systems. There's definitely some usage of the VSDs, but there's obviously concern from some people that this move into Wayland

obviously causes issues there. Now, my understanding is things are in a better state than people make them out to be, but do you happen to know much about where things are with Wayland on that side? So, as far as I know, there aren't that many people working on it,

which is definitely an issue if you want to use it. But freebies, has a working plasma wayland session. From what I heard, there might still be some issues to figure out, but it generally works fine. And, well, there is more than a year for them to figure out the remaining details. In general, from upstream, there is a limited amount of things we can do to support it. if there aren't a lot of people actively caring for it.

But generally, we have it in CI, we make sure things at least compile and we don't add anything, any dependencies or anything that won't work on the BSDs. And if someone reports a bug on it and can show a back trace or some specific steps, how to reproduce issues, we would be able to take. can look just the same as on Linux. Yeah, because I know I know this is a thing people I think

most of the discussion of it is from Linux people who don't actually use BSD because every time I've every time someone mentions this like oh there's no Wayland support on BSD like I can literally point you to a compositor made by free BSD people

for free BSD. Like it's that's clearly not the case um all people will point to like a five-year-old blog post from the NetBSD team where it's just like, yeah, at the time they didn't want to support it, but I actually am not

entirely sure where things are. I know two or so years ago there was some open BSD people working on some sway stuff, but as you said, there's just not that many people working in this space. Like most people in the Unix world now are on Linux.

So there's just less work there being done, there's less discussion of it being done. And, yeah, like, things are going to just naturally take more time there. Yeah. But if there aren't any issues, like that we can help with upstream, just people should reach out. We can't fix what we don't know is broken.

Yeah. But, like, from the core wayland's side, it's been supported for a while now I think and like I said you can run a plasmawareline session now so if anyone really wants to use it go make sure it works fine

and if not file bug reports please now we talked a bit about like screen capture automation before but what about like general keyboard automation and all of the stuff that you can do on X11 through tools like XDO tool and various other things

that exist. Where do things stand right now with plasma? So you can use X due to on Wayland, like on plasma Wayland just fine. There is an option in system settings to remove

the permission prompt for it and for like input emulation that's all plumbed through X Y land. It will do your key presses and that stuff just fine you can also use the remote desktop portal for all the remote things like doing screencasts and injecting inputs there are some things that you will need kd specific tools for like if you want to change output settings you can use case

screen doctor. And if you want to manipulate Windows, there's KDO tool, which uses Quinn scripting to implement a lot of additional functionality that you can combine with the other tools to do. So all sorts of things. There have also been some discussions about like integrating some automation tools into plasma natively. Nothing has. really happened yet about it, but I hope at some point we'll expand on that, because that obviously has the potential to make life a lot easier for a lot of people. Yeah, I, like, screen automation is one of these things, which I don't do that much of, but people that have these, those people that have these very established workflows, and sort of, I

understand why trying to move that off of what you're using right now, even if it is possible, is difficult, right? Like, if you have, if you've been using something for 20 years, you have 20 years of X-11 scripts to do all of the automation

that you want to do on your system, like, any sort of change to that is going to be, it's going to be a big endeavor to move your system over. Yeah. So these people, really need to take that year of time and make things happen because where either they need to

adapt their scripts to plasma wayland or they need to adapt their scripts to another desktop environment they're going to switch to what they choose is of course up to everyone personally but especially with some of the x-11 tools still working on the plasma wayland through the compatibility things it should hopefully for the vast amount of people that have scripts not be too difficult to port things over so one area where I do see people

have issues is every so often I'll see someone in my comments say it'll be something along the lines of K-win doesn't work well on an Intel GPU and I ask them what they have and then they say oh I have a 3,700K I'm like

yes that is a what 15 year old CPU at this point there is there is certainly a argument to be had that Wayland doesn't work anywhere near as well on this like you know really old hardware and for

people like that, I can understand why, you know, a move to Whalen is not really viable, especially if you're in a position where you legitimately cannot afford, you know, you can't afford new hardware. Yeah, we, at some point, we got a bug report from someone using an Intel, I think it was GMA 500 GPU, yeah, from 2004. So, 21 years old by this point. So, and we did fix some issues with that, like, fixing means it now properly falls back to a software renderer. Right, right.

Because the GPU is so incapable that it can't do more than a few shader instructions. It's like OpenGL2.1. Yeah, the limit wasn't even performance. but it literally can't fit our shader with just a few color management operations into the hardware. If you have that kind of hardware, well, I'm sorry.

Like, there are limits to what we can do. And it will work. Like, your system will not completely break, but performance will definitely not be optimal. If you have anything that's like 15 years old, that should generally work fine uh it might be that with some particularly weak integrated chip use you will have increased latency or that you can't handle like as crazy high resolutions because to some degree they still had 2d accelerators to handle those cases

and modern graphics APIs at least at the moment don't have any abstractions to make use of such things. There have been some discussions at XTC about potentially introducing new APIs. I doubt that old hardware would get support for new drivers like that, though. So yeah, at some point, if you can't use XRop with compositing, like if your only choice is to use it without compositing,

then that's the way. hardware is just too old for us to really properly support. On that note, I've had people sort of ask this many times. Why is compositing this like core part of Wayland? So if you've played games on X-Hork, on plasma X-11, you might have noticed that once you start a game, all the graphics turns to shit. So you open your menu and it's completely black.

Like the background is completely opaque. There's no transparency. There's no animations. Plasma shell even has lots of special code to like change textures. So it doesn't attempt to do shadows and things like that. That is what turning off compositing means. Ex-Ox still does compositing. It just does a much, much worse job.

Like with the limited 2D acceleration, or nowadays with the OpenGL abstraction that does the kind of same things. It's just way more limited. And so you have these two separate layers that are, well, for the user, very annoying to deal with. We got some bug report about it recently, even about, hey, something goes wrong when I start a game.

This looks all completely broken. Yeah, that's exorke. And Wayland just does away with that. Like, if a compositor wants to, it can also implement a more limited compositing level like that. It's generally a bad idea because, well, even if the user experience wasn't terrible, if you change the visuals of applications,

some applications will not handle that well. Like Plasma Shell specifically had to do extra steps to work well with that on Xorg. and most apps don't really go out of their way to do it. I think GTK just gets big black bars around its window nowadays. If you disable compositing, it's not great. It also means that the compositor actually can do all the nice and efficient things. If there's no additional layer below it that it needs to somehow

interact with and negotiate who gets to paint a window we can do things like take a window take its buffers and put them on special hardware on the graphics card to reduce latency and skip the GPU compositing step right I think a lot of people on x-11 when they hear the word compositing they just think window blur so people kind of they they don't like because You know, nowadays when people are running a window manager, they're running like PICOM with it. And that's the reason why they're running it. So it's a lot more to it than just that.

Yeah. And basic transparency is like the first thing you should think of. And also rounded corners, like all the nice visual things that nowadays, well, I say nowadays, that for the last 20 years have been the standard everywhere else. are optional on X-Rog and that's just another great thing. I

might like I how words sorry we'll try that try that one again so my first experience with Plysmah was using Plasma 6

and the launch was from my experience mostly good. I don't remember there being some early teething problems. One problem I made a bit of noise about when it happened was the the shader caching to the drive and then that very much crushing mechanical hard drives, because understandably no one in the team probably test on mechanical hard drives, and it wasn't a problem on SSDs. And it's the same problem that exists with Ex-Sorg, right? like if no one's testing it

no one's going to know if there's a problem there yeah some people in the plasma team keep like really ancient laptops around just to like if there's a performance issue detected like that people

find on old laptops or old desktops they can go and wait three days until it compiles plasma and then test with a a system that has the same kind of specifications because, yeah, on my, like, two-year-old laptop,

I will not find the performance issues that you will find on a 15-year-old laptop. Yeah. Well, especially one, you know, has a good GPU in it. It's not like a bottom-of-the-barrel budget laptop. Yeah. I mean, I don't notice when the software cursor has turned on. on. That's, like, we have automatic tests for that now, but like in a somewhat recent refactor

for legacy drivers specifically, for legacy mode setting drivers, that broke and it didn't use the hardware cause anymore. Even when I went and, like, explicitly used the legacy API on my laptop, I will not notice that. Like, it's rendered on the GPU, sure, instead of the lower latency hardware plane, but the difference isn't that huge if your GPU is really, really fast. So yeah, sometimes we do need to explicitly go out of our way and test on really old machines.

Just how long it takes to compile is a big annoyance. Another area where I find, I always find problems anytime I go use it. Whenever I use a new desktop with a vertical monitor, It's always very funny when I run into really obvious vertical monitor bugs because I know that no one's using a vertical monitor. I remember this wasn't in plasma... There were plasma issues at the time, at one point, I think early plasma 6.

But I'll never forget in some of the early cosmic testing where it was very obvious literally no one had used a vertical monitor because I pressed screenshot my vertical monitor and the entire screen blacks out. Oh. But yes, that's a very, very common thing that nobody tests for. Like one of my colleagues has a vertical monitor and his setup. And yeah, from time to time, I just do some larger refactor or something like the recent mirroring changes, for example, where... Oh, right. Yeah, that was... Yeah, that was initially pretty broken.

And then you need to go out of your way and test all the weird configurations. Otherwise, you will never find such problems. It's also like one of the cases where X-11 apps positioning their stuff often went horribly wrong, at least in plasma five times. We had so many bug reports about context menus being misplaced. if you have a vertical monitor to the right of another monitor so that there's no screen at the zero position lots of X-11 apps do not like that at all like it's it's rarely tested

like yeah another area is like the accessibility stuff because that's like an area which is difficult to test because a lot of people, a lot of the people they're going to be writing the code don't necessarily rely on the tools. So you may brush over areas which you don't even

realize are a problem. Yeah, so it is sometimes a challenge because where accessibility isn't just one thing would be so nice if it was. But like,

I somewhat regularly use the zoom effect in Quinn because well if you're working on like graphical changes if you do rounded corners or something you can zoom so far in that you see the individual pixels and then you can count them and see

yes this is definitely wrong or correct and that's super useful outside of accessibility cases So, well, if anything breaks there, I will immediately notice, like, within a few days at most. But if a screen reader breaks, not a lot of plasma developers, like, test that often. Like, I'm confident that at least one person always tested before a release, but nobody leaves that on all the time. well it's not even a matter of most developers it's not even a matter of testing it's

it's also the combination of environments you might be running at once because it might work by itself but it might be oh if I combine three or four other settings some weird esoteric problem happens yeah it is a different mode in the screen reader like where it reads aloud all the keys that you are pressing recently there was a bug about that I think Like, there are so many things you also need to know about how people use these tools. Because if you enable a screen reader, you just tap through some application quickly and see it reads aloud all the things you select. But then, I don't know, you switch between apps and it doesn't say anything.

you probably wouldn't notice, but if you're blind, it should definitely read the window titles or I don't even know what exactly is expected there. Window title, application name. If that's missing or broken, you may not realize. So we need as much feedback there as we can get. we do have a plasma accessibility engineer that's hired by the KDEV I've had on the show a couple of times yeah Ritchie has like uses the zoom effect and the screen reader

so he will notice things and issues that we don't and can improve them The feedback that we've gotten about the Zoom effect, at least, is that our Zoom effect is really great, and that's amazing. But some feedback I got with the, we are dropping the X-11 session was that some people rather use the magnification app that we have instead of the effects. so they can have a magnified like the window on a different monitor while like targeting with a mouse an area on a different screen which is not also just not just useful for accessibility but also for like graphic design again if you want to look at pixels

without zooming in your whole screen so there's a hundred thousand edge cases that we should ideally support and if there are problems please bug reports or at least show up in a chat somewhere and tell us about

the problems you have because we just don't know right right yeah that's that's such an important thing where like when it's an even with an issue that might seem

really obvious there's a reason why it's fixed. Like in a lot of cases, it might be obvious, but it might not be, it might just be one or two settings away from what someone else has tested. It might not be a crazy weird configuration where you're using, you know, some weird hardware that no one has, but it's just plasma has so many different knobs to tweak that it can be very easy to get to a point where people, you just, you just can't test every single case. Like you try to test as much you can, you hope that things are working, but like, users

really do help out in regards to that. Yeah, especially when it comes to users that aren't actually using the plasma options, but are using some third-party application. or additional quinn effects or additional add-ons somewhere in plasma that aren't upstream. If they break or if there are problems, it's even less likely that anyone from us will know. We need to be told first, which is broken and more importantly, which ones are actually important, how many people have this issue. Because if one person has an issue with a third-party app, we probably say, well, sorry, but we can't fix everything. But if there's 10 people saying, hey, in this app, I can't type into this text field, and it's specific to plasma, then that's a, okay, this is clearly important, let's go fix it.

Or work around it or work with the developers of that app, right there. Always options. So what do you find yourself getting, you're going to wrap up in a bit, what do you find yourself working on right now? What has got your attention? Right now, screen mirroring because it's pretty terrible.

Like it works completely fine if you have screens with the same aspect ratio. If you don't, yeah. Like you're going from a 16 by 10 laptop to an external monitor, that's usually going to be 16 by now. Yeah. Or like a 3x2 laptop to a 32 by 9 monitor, like in my case. I'm testing the absolute extreme case.

Yeah. So, yes. Like on X-11 and Whalen both, so far that was like faked by, oh, we just positioned the output to be on the same spot and then scale it. so it's the same size. But if the aspect ratio doesn't match, then you have parts that just stand out or overlap wrongly. It's super chanky and terrible, especially on this very special setup.

You get, like, on the laptop, you see a tiny slice of the external monitor and the panel just goes through the bottom of the laptop, but you don't see any app icons because they're all way further to the left on the monitor. So what I've been working on is to just properly take that into account, render with black bars to the sides, and make sure that apps also don't need to check, oh, is the user mirroring their screen? Otherwise, they will just show you the option, show on this screen or show on that screen, but it's on the same position, right? Plasma Shale has a lot of special code to avoid putting.

two panels on the same mirrored screen because it's actually two screens underneath, right? So there will be a whole lot of bucks fixed with this. And then I can hopefully get back to some exciting feature work in the future. Well, made you want to jump onto mirroring. Was it something you specifically needed or were you just like that's broken? I want to do it. Well, it's mostly that's broken.

I need to fix that. It's my job to fix all the things in Quinn. So that's one. We have a bug report with probably 50 duplicates by now, if that's even enough. Like a lot of issues that are just around the topic. Like some are, yes, there's no black bars, but others are our overview effects only shows windows on one of the two screens if you mirror them because at some point we fix that if a window overlaps two screens it's shown on both so now it's shown on only one of the mirrored screens

which can cause quite the confusion in presentations when you just open the effect and then the audience sees nothing and you see all the windows right right okay that has happened at an event this So, yeah, it's just a motivation. This is terrible. Let's make it work nicely. And I think, and I think it's really nice. This will end in plasma 6.6, so I hope the users will agree with me there. When is that version scheduled for, is?

February, I think. Okay. so what is it now three months yeah it's December isn't it two months two months that's how that's how that works well at the start depends on one of three two to three ish months whatever doesn't matter two and a half we'll settle in two and a half yeah and the other things i i hope to still push in before plasma six and six uh the other items on that screen section and the significant issues Like, I think custom modes is also a pretty common ask because some people like to overclock their screens. I did that maybe six years ago or something.

Right. And some people just have like CRTs or weird screens that have broken resolutions and where you need to fall something. so yeah this actually I've read that one I wasn't sure what that one actually was that makes sense I this is a problem

I've actually had a lot of people ask me about like hey is there a way I can set some custom nonsense to my screen I guess yeah KDE is going to be your answer hopefully

by February hopefully yeah I mean that's surprisingly been a small feature to add. It wasn't that complicated because we already generate some modes automatically in Quinn because people annoyed us about like resolutions not being supported by

their screen, which wasn't our fault. Exorch just generated resolution that the screen didn't actually support and added them to the list. Yes, it did. Yep. So we just had to copy it. So So all the code was already there to generate modes. We just needed to add a user-facing toggle for it. That's one of those slightly concerning ones, you want to make sure there's a recovery state for that because otherwise, you know,

you could put your monitor into a situation where you just doesn't do anything. Yeah, our user interface has a revert toggle that, like, if you don't confirm, it will just revert the setting. So hopefully people don't use the command line to switch to a newly generated mode because if it's broken, you're screwed.

Or like you need to get a second screen to change it again. You know it's going to happen. Do it through SSH. Yeah, I have wondered if the command line tool also needs like a timeout, like reverting. But I can imagine if you're using it a script that would be super inconvenient.

I guess you could have like a dash-dash-no-confirm or something like that. Yeah, but people that are currently using scripts with it will then have it broken. Yeah, maybe I could limit it to custom modes only that the user generated. it. Maybe I should also just print a warning. Please, please use the UI. But in general, you can use SSH to just run the command and go back to a supported mode if it's really broken. Most people do have phones, so it's not that hard as long as SSAH is set up. If not, I think switching to a different virtual terminus switches modes

so that could work yeah um it should I don't see why it wouldn't whatever yeah um there's solutions I'm sure you can find a monitor

you need to find I'm sure you know someone with a monitor I'm sure you know someone with a monitor I I just hope that not too many displays get broken in response to this, because with the CRT, at least with some CRTs, as far as I know, you can't break

the display if you configure them wrong, so. Yeah. Be careful with what you're doing. Don't run anything stupid. I already put in the protocol for the output setting that it's the user's responsibility to ensure the mode is correct. So I think we'll wrap up there It has been a pleasure talking to you

It's the pleasure to finally properly meet you And yeah Do you have anything you want to say about the migration That we didn't mention or Are we basically good? I don't think I have anything more to say I hope it goes well

we'll see you in a year I will definitely be there to watch any of the fallout that happens and I'm sure that people like Nate will be replying to far too many comments regarding what's happening as well because I guess I don't know he does it for the love of the game

I don't know he seems to love getting involved he's he is basically as good of a spokesperson as you can have for a project. He does a good job, clearly. Yeah. He cares a lot about helping people. That's part of the reason why we're all working on KD software, right? But he's especially good at it.

Yeah, yeah. I guess if you have problems, go report bugs. Try out modern versions of plasma. If there's things that are missing, now is the time. Now is the time to report them. Not in a year and a half from now. You know what's going to happen though. The second, the second the version comes out, this is broken, that's broken. It's been broken for a year. I didn't report it. So, um, if people want to go, you've got a blog. Um, people want to check that out. Where can they find that? Semunda.gitup.io. That is with three A's. I'll leave a link down below.

Check out the fundraiser. It's well past its goal, but the goal does not stop until, well, you can donate whenever. But the fundraiser goes to the end of December. Anything else you want to direct people to? Uh, are we? Okay.

Just, like, donating is always great. Let's break more records, please. Yeah, yeah. Well, next year, there's going to, it's going to be basically everyone that's going to have the notification, so maybe. Yeah, next year, Kubuntu LTS will have it as well. So assuming most people migrate over, well, maybe there will be another record broken. I'm excited to see it. I'm so happy to see what's been happening.

Anyway, I'll do my outro and then we'll sign off. Okay, so my main channel is Brodie Robertson. I do Linux videos there six each day a week. Sometimes I stream as well. I'll probably do a stream at some point for the Cosmic 1.0 release. So if you want to see me breaking Cosmic again and looking for more bugs, check that out. It's always fun to do some bug hoarding for that. I've got the gaming channel that is at Brodie on Games.

Right now I'll be playing through Silk Song. And maybe devil may cry. Check it out. And if you're watching the video version, this, you can find the audio version. I've done like 300 of these. I still can't do my outro. The audio version, Tech Over Tea, on basically every audio podcast platform. There is an RSS feed as well.

Check that out. And the video is on YouTube at Tech Over Tea. I'll give you the final word. How do you want to sign us off? KDI is awesome. Goodbye. Perfect.