Still Breathing: The Past, Present, and Future of Oxygen

So, my name I've been around that sounds weird. Um, I'm a designer. I work at KAB and I've been in Kada uh doing things for KD for a few years and I'll talk to about it in a few seconds. >> Hi, I'm Philip Fil. I work in academia as a posttock at the Institute for Social Research in Zagre in Croatia. Okay. So um see this Malaga 2005. Do you know what I was doing in Malaga in 2005? Presenting this thing called oxygen 21 years ago. Um this thing apparently refuses to die and I I assure you I tried to kill it. Um uh but um and um it was you know huge as some people would say uh these days. Um it was the KD uh for default theme. It was um kind of it redefined the way themes were made. Um because back then themes were uh not really a thing. everything in KDE itself and and most um uh desktops, open-source desktops back then were kind of collages of multiple efforts from a design perspective. Uh Oxygen wasn't that. Oxygen was the one of the first um efforts in open source as far as I know that try to do to do everything. Yes, we were kind of insane. um do the icon theme, do the cute style, and then some other idiots uh came up with plasma. Uh and that required us to do a complete set of weird things. Um and we did it all uh not knowing how hard it was. Um we did it and it has so today's uh a day of uh or today or yesterday uh um we were showing numbers. I also have numbers. Uh we had 5,823 commits and Oxygen repository add over 10,000 different files. The size of Oxygen on my git is 600 megabytes or something. Just the icons. Yeah, kind of insane. Um, and what am I most proud of it uh is that yes, as I said, it was a style of styles. It redefined how styles supposed to be made and it's um, uh, full encompassing and sort of thing. And it we did it. It was fun. And at the end, I was kind of exhausted. But anyone that knows me said that I would do the four series and this is four series alone. I wouldn't do anything beyond the four series. So KD4 we did oxygen and at the end of it I thought that my time was done. My language was done and someone else needed to come on and I was part of the kitty in the beginning. A kitty no the how do you call the um design group thingyama majingy. Uh but mostly trying to get away from it. Um so by uh by then I said that oxygen was dead. Um we we were done with oxygen and because um flatness was coming. Um and back then I saw the writing on the wall. Everything was becoming very flat. The and my design language as I express myself is tends to be a little bit over the top. Um and that's the way I like it. Uh-huh. Um and uh I I thought okay uh my voice uh I need to be replaced. Uh my voice needs to be replaced. And I stepped away and I thought uh that oxygen should be killed. And I tried to kill it. I said it on I think it was an AV meeting when I publicly said I'm done with this. I'm not going to carry on with this. And yes at at the end of KD4 I start um I stopped working on it. But but clearly oxygen isn't dead or we wouldn't [snorts] be having this presentation, right? Um so we'll we're jumping into 2026, but we'll take a little uh small jump back into the beginning of 2026. In my own personal timeline, I hadn't been active for a little while in KDE, but I was starting to get that old itch is when you uh becomes harder to ignore some issues that you see on the screen and you get so annoyed at some point. you say, "I'm going to fix it." And uh that's what happened uh with Oxygen. I just saw that things weren't looking right. In particular, I saw that the air theme was missing, which I thought was a bit unfair because it was the default theme for most of KD4. So, I said, "Let's bring it back." So, all in all, it's a pretty banal, not particularly interesting story, but what's the bigger story behind it, the context, what had been accumulating is what's more interesting. And that's something that we could call uh various names. We could say anti-minimalist backlash, minimalist fatigue, uh because it's been 12 years since we've had we've moved towards flat design and we've had these similar designs and we're not at a place where we have something new yet. We're shifting in some directions where it seems to be returning to more skeomorphic, more depth. Uh but we're still at a limbo. So uh how to explain this oxygen revival? Well, the simple explanation is you sometimes hear is people are just nostalgic about the good old times when we had less bills to pay, there was more we had more hair, there was less wars and so on. It's a pretty simple explanation, but we think we think there's a a bit of a bigger story behind it. It is a backdrop when we have users that we're in KD for and want to see it back, but we also have some new users. And so the part of the story we think is people uh not having this new thing and looking back at what was done and looking at these details and thinking aha it would be nice to have something like like this back. Uh and and one part is that Nun's going to talk about this the desktop being fun is that it really became more serious um like a more professional. We don't have any more bouncy balls and stuff like that. Maybe there's also some yearning to have it have more personality, be a little bit more fun. So, for the restoration, we don't want to go into the chronology uh because we think this presentation about oxygen is about so much more. So, there's blog posts uh and I'm only going to mention uh the big things that we had issues with. Uh, number one was these plasma themes that were pretty abandoned and they look bad. Like think of white text on a white background just um it it wasn't in a good shape. Uh so that that that was a chunk of the work. Uh of course um our Q style if you look at the code it's mostly from KD4 KD5 times. So, we had issues with fractional scaling, uh, controls looking choppy, things not rendering properly. In 6.8, this should be almost perfect, but there's still some issues around. And, uh, we had some missing icons. So, uh, Oxygen had issues with um deprecation of icon themes from the plasma themes because Oxygen doesn't just use symbolic icons at small sizes. So, we have a mixture of colorful and monochrome icons. And that means if in plasma code one thing is coded as just a simple I don't know Akenady icon name is Akanady if it's not Akanady symbolic it draws the colorful icon then we get a mix of colorful and monochrome icons and it doesn't look good all in all. Uh the hard part which is maybe that's also something that's interesting for the community is this part about getting contributors right getting a group of people that are going to maintain this uh because we know you're all busy you have a lot of stuff to do and maintaining an old theme is probably not something that we want to add to anyone's list of list of things to do. Uh so uh what was interesting here is when we actually we uh started talking about this we saw that there was a lot of fork work being done. So people on GitHub forking there was like two forks of the whole oxygen repo forks of oxygen icons. These people that were really interested in oxygen and could contribute upstream as well but we're just doing it somewhere on the side. So we try to reach out to them and now we have this nice situation where they join in and they started sending in patches and of course this contri this benefits more users than when you do a fork which is always the easy way because you can just do what you want but this is the more rewarding way. Uh one issue we faced was this uh of legacy. So Oxygen Nuna knows this better. Um it went through a lot of iterations in KD4. there was different designs. There was some more gradient, sometimes less. Then you have users that wants a specific version. They want this 4.4 version. They don't want the 4.8. And then you have to manage all of that and create some sort of synthetic version. But u what we have right now, it it seems to satisfy most. Yeah, >> you'll like it. Uh okay. numbers. Uh since the restoration we began, we're looking at two repos. So oxygen icons and the oxygen repo. We have around 82 uh uh merge requests merge. There's a little bit more commits. Uh we have 76 issues we closed on the buck tracker. That's looking at uh all of the like the whole Oxygen group and almost around 10 new contributors. So there it varies from people that have submitted a one-off patch and some that seem to be really interested to keep on contributing and this could be like a gateway drug for them. Of course they send a patch on oxygen, get experience with KD and start improving other parts of KD as well. Okay. So um what I've been up to with these guys uh so because I the I tried to kill things and people refused to kill them and kept on even even the old cute style people were maintaining it. I kept using it because okay it's my I I eat my own dog poop. Uh I'm not a dog. Uh let me say that. Um uh and because of that I I uh I started helping these guys out and okay let's let's try to see what we can do. Uh so what I do what I mainly do apart from you know thinking about styles and all this stuff I mainly waste time on these guys and they are monsters I tell you these are monsters. This doesn't do any justice to them because in order to see them properly you should see them in 124x 1024 pixels. That's that's actually the way you should see them. Yeah, I've been doing a couple of uh icons in the old style. I did 300 something icons for because the in plasma I I don't know if you know uh themes used to ship the icons themselves. So, plasma theme had to ship their own icons. Uh then some people uh removed that and oxygen became incredibly broken. Uh guess what you have to do? No, no, you have to do a bunch. Luckily, they are monochromatic and I can do them while I'm sleeping. I only did three like that. Uh so, so yeah, that's what I've been up to. >> And now the fun part, right? This is an old theme that's that's been restored, but really our whole point here is the future. Um, and the bigger story behind Oxygen is also a little something that it says about KD, which we think is something very positive is personally I don't know of an example of a desktop environment that kept a 20-year-old theme alive. Usually they die or they just keep staying the default. Um, and and this I think is an interesting story with my social science background. I've read so many critiques about what it means when every single core of our world is dominated by the principles of utility, efficiency and the costbenefit perspective. Uh and I agree with those critiques and what the example of oxygen clearly shows here is something else because you have an old theme. If you're just looking at efficiency, you would have deleted it in 2015 because what's the point? It's more efficient to just develop one thing and and work on it. Uh but it in KD it recognized this historic value of oxygen. It recognized the artistic value behind it. And what we see is also it recognized user choice. So it doesn't just give you one theme that you can use but you have an option to choose this style u that's retro in general. Uh and uh oxygen really benefits from KD. I mentioned this with with this fork work um from all of the infrastructure and we're really grateful for also everyone that's helped us out with patches and um we'd like to keep it this way. Yeah. Uh now what does the future hold? Again, this is a story about Oxygen, but also it's a story about every single future theme, new theme that's going to be created uh and theme creators. So Oxygen is a theme. It's not the most complex theme, but it's uh more complex than breeze. Uh and when developing such a theme, we get challenges. Like for all of the theme creators, we would also welcome more efficient theming because right now anyone that is going to try to do a theme, it just takes a lot of work, a lot of work and a lot of dev power. So that's also that's something that everyone's going to welcome more efficient but at the same time we have an oxygen as as example of something that needs um needs more things supported so that it can be uh so it can work right. Uh so we we would argue for preserving the this power either through these old ways or uh having um the new ways support all of these properties. So you can do something that isn't just flat or minimal, but also sometimes even wild. Why not? And uh we're excited about KD's future where it can offer also multiple uh first party themes from users to choose on. So not just like a retro theme like Oxygen, but maybe like Oxygen Squared, something new. and and okay so uh do you like the wallpapers do you like these wallpapers I I made them no I didn't I didn't make these wallpapers I asked this thing to AI generate some wallpapers in the style of nun from kitty >> we don't like it actually That's the that's my point and that's my point to kill oxygen and all oxygen and all this stuff that I'm doing. Anything that has been done is going to be repetitive. It's going to be boring and it's going to be easy to make. Only the new stuff, only the new ideas will actually shine in a sea of sameness, easy to reproduce, easy to ask a AI agent to make it. So that's uh why I really um want to create something new. And I'm going to show you something new in a second. Well, I think it's new. It's relatively new. Well, at least it's in my brain. It's new. I haven't seen it anywhere. And I'll show it in a second. Um Marco, did he went to the plasma meeting in Porto? Okay. So, you remember the the the video that we were shown there? That that's the thing that kept that kept on a polisher polisher brilliant designer. I recommend everybody to go and watch this video on YouTube. Um it's easy to find uh and you just search for um solemn design versus serious design. Um and I want us to in oxygen especially to do serious design. What is serious design? Serious design is playful, is silly, it's stupid, it's all of the things that are hard. Solemn design is all of the things that an enterprise distills out of a a design language. Serious design is done on the back of a napkin. Solemn design is done through iterations of version control, multiple design systems, and all that stuff. That's solemn. I'm not really interested in solemn design. Solemn design needs to exist. It's if in in the end how we produce things. I was doing solemn design by the end of Oxygen. Um especially on the four in the four series um I'm doing serious design or trying to do serious design now. So um let's see something new. Okay. So, I want to make idiotic ideas come alive. Can you see this? I need to get anyway. You can see it. Let's live demos. What can possibly go wrong? Hey. Okay, there we go. So, uh ideas. I want to see crazy things. I want I want to I want I want to control, you know, the radius of stupid windows. I want I want the transparency to have like little corners that show things and the corners are inverted and shown inside the the thing. Um, and I want to like only draw the shadow on the inside of something, not on the outside. But how hard could it be? Hey, I've done this. This is QML. This is working QML. It even scales because all of the thing is, you know, this is actual vector stuff. It's not it's not like pix maps I stole. And I want I want buttons that I bounce. Why? Why? I want b I want bouncing buttons. I give me bouncing. And I want Hey, I want my because Okay, this is an exaggeration ex exaggeration because this is s this is silly. But when I use the button, you notice that the shadow follows my thing. that it's actually that's why the bouncy button uh the bouncy thing is important because it allows me to do that and because I can control the bounciness I can control the shadow and I do the shadow slightly different and I bounce it a little bit more and where I press it it you know it it does the flex and the shadow beneath the thing is a little bit less important. Um, so I would like us to to be able to do these things more um on both sides. And then this is the playfulness, the the serious design. At least I think it's serious even though it's completely crazy. Um, why and in a second I'll show you a few a few things more maybe even more crazy than this if you want to. Um, but then it it um I was thinking like um why is this important? Seriously, why is this important? Um why do I care about this? I care about this because I love this and I love you guys and I love everything around it. And this is the way I express love. And I think this is an easy way for users to understand that we love them. This takes care. The care that we put in a design is the same care that we put on code and it's easy and it's not easy to for a user regular user to understand that we love them because they are not going to check the well some of some of them that's how I usually got more developers in they would look at the code and fall in love with the code and then start contributing but for regular users they don't get that they they will maybe see the pixels and they will understand the love that goes into the pixels and they'll feel the love that we have to them. So that's why I think this is actually more important than it looks. Um just a final thing. So maybe because it's fun. What else can I do? So I did a thing. There's actually a lot of other things that I'm not showing here, but I'm just going to show just the multi effect where But this doesn't make sense. Okay. Anyway, that's just because I have access to all of those things, I can do even crazier. Oh, come on. Live demos. What can possibly go wrong? Can I even press a play button? Because I can do these things. I can even do crazier things. And I can make See, now it's following my mouse. And this is cool, right? Is it stupid? Probably. But it's cool and it makes me feel something. And we should I made made a blog the other day like why everything's so boring. Everything's so meh. Can I feel something these days? And this at least makes me feel what an idiot. >> [snorts] >> Okay. So, that was that >> uh I don't know that that that is a a you problem, not a me. [laughter] Not a me problem. I'm just I'm just having fun. Uh how do I get back to my presentation? No idea. Like I'm not a Uh-huh. Uh-huh. Uhhuh. Uh so if you want to get involved uh you there's things there's a chat thingy machingy where we you know chat um you're not a coder we need designers actually need a lot more designers too a lot we really need more designers in KD this is seriously historically my biggest failure in KDE was and uh by the end of Kitty4 um I didn't attract uh a designer community. We had a few a couple but it's always that it's only just a couple of designers. It's really hard to mentor and to foster a space that is conducive to designers. Um we on the developer side of things we tend to lose track of our conversations an alienate like we will spend 3 hours debating the proper way of drawing anything and then the and the designer on that conversation that just asked for something just goes away like what my brain just and it feels like it doesn't belong. So we keep on losing designers in that way. Unless they are have some sort of masochism handy. Is it massochism? Um so unless you have some sort of masochism who keeps sticking around because you kind of like the guys or something. Um you we lose the design. So we need to improve that. That's something I plan to improve. uh or create a space where they feel safe and they and they see peers and they listen to peers and they have people that can translate um their frustrations and understand their frustrations, explain them. Okay, the thing you want to do, which is crazy the way I like it, can't be done unless we change this and this and that. It's actually very interesting. You can ask Marco. Um, designers think differently and present sideways solutions for things that usually developers don't think of in the same way. Um, we didn't have anti-housing on the corners on KD 4.1 4. We didn't have anti-housing on the corners. Uh, and then we fixed it and we managed to come up with multiple fixes for that thing. Someone some of those fixes were really hacky. I'm super proud that we actually had partial fixa and we were the first ones to have antiologist corners minus one pixel, right? It was one pixel that wasn't [laughter] [snorts] uh but yeah um and so join us. Um it's the it's the it's cool. Um uh because it's you know the gold desktop. Um what is this? I guess just uh we could add that just on this thread of oxygen being more than oxygen is this that it can create a space for creativity for fun that it the stuff that we're showing you we're not saying this has to be the default we understand what the default should generally be but we think this this oxygen can be some sort of space where people can experiment and we can try to work on making Katie if someone wants to have it more fun more playful um Yeah. >> Yeah. Create a space for people to be serious to do serious design that they can share and then it can be shared to Kitty. Uh so that actually then becomes solemn. Not too solemn. Please don't never too solemn but that can and and it it brings new types of people into KD that people bring hopefully it will bring people with different mindsets and into KD and and start spreading uh more love into into design type of things in KD. But we we really need to do things that they find fun. It's not just, you know, the things that we usually put out. It's not those are not super fun for a designer these days. Yeah. So, that's about it. Uh, thank you. Blah blah blah blah blah blah. You know me. And there's a guy. You have questions? [applause] >> [applause] >> Yeah, thank you very much Nuno and and Phil and we actually have quite a enough time for questions. So yeah, let's have a catch on among the audiences first. Okay. >> Hello. That looked really nice. >> Um, now I'm going to give you [clears throat] a engineering question. >> Okay, >> how will we get those two GTK applications? [laughter] >> You all guys know the story, right? I'm not going to repeat that story, right? We tried. >> Yeah. >> Yeah. I I guess it's going to be a quite a problem to get it to cute widget applications also. Didn't you saw that conversation last year? That just the one that we had. Like that was literally what I was showing them. [laughter] There they are. >> Yeah. Yeah. >> Yeah. No, we we do have an issue nowadays. We want a cohesive look, but we just have a Q style. Then it just gets like a proxy with QQC desktop style. We don't have a proper cute quick theme and the GDK app. Well, the better not >> talk about that. >> The GDK. Yeah. >> Yeah. So currently the effort in oxygen which there is a reason for oxygen this current oxygen for me personally at least my main reason to be helping up with this is at least I get to learn how things are working right now because I was kind of off the the thing for a few years. Um so I need to understand okay how am I how am I going to fix all cute how I'm going to fix giraami so it shows uh oxygen but now I've looked into kami I want to fix other things there sorry um uh but yeah uh how the styling works you have the union people trying to do uh things in that area maybe we can do some things hopefully if you support at least old auction I think it should be doable it's just some gradients crazy stuff like I No. [laughter] >> Yeah. >> Let's talk about that. [laughter] >> Um uh so yeah, that's uh that's what we're trying to do. >> Okay. Uh next questions. We still have enough time. >> Okay. No. Ah, >> more demos. >> Oh, you want to see more demos? Okay. >> What's your professional designer opinion on Fluffy Bunny? >> Fluffy Bunny. I have nothing to do with that. I'm very proud to have nothing to do with that. The people that did that are heroes. Seriously, that thing is fun. Uh that is seriously fun. I was I was appalled and proud at the same time. And it's also something that's in need of restoration. So >> pretty sure it works. >> It's kind of broken. >> No, no. Most of the things are broken. Yeah, cuz all the plasma changes. >> No, no, no, no, no. I do I do serious work. [laughter] >> Yeah. [laughter] Um so I have a question about uh on the side of testing. So I think it's clear that the industry is moving towards or it's has an appetite for like you know coming back to depth and perception and you know like special understandings of what they see on the screen and it's because they got lost in the flatness right like you know everything looks like a control therefore it must be a control and it's hard to extrapolate that I think sometimes too like finding the middle ground of like you were saying you know like you know uh people accepting uh changes but not deviating too far from what seems familiar because you know we'll have an aversion to change. Um, I'm thinking I think it would be awesome if stuff like what you're building, what we're building, uh, as on the side of theming could be tested in a way that users could give us feedback. I don't know if you're open to that. I think we don't have a good framework I will admit you know today in our systems to sort of expect that that kind of stuff but I think it would be awesome because I think one other big missing piece of our work is that tele not telemetry but like let's say the the feedback from users when you're asking a very specific designoriented question right uh what is your take on on testing some of our UIs or experimenting and getting some feedback from users. >> Um, this is not going to be very popular. Um, I I don't like testing like the way usually people talk about testing, which is um showing something to a person asking what do you think about it? Um, turns out when people spew out words about how they think about something they've just experienced, they will spew out the words as they understand what they've just experienced. Usually that's just a postprofessionalization of what they just experienced. I don't like asking and that's how usually we see test user testing. That's not proper user testing. proper user testing is you shows things, you give him exercises and you actually see how it struggles. You have eye tracking and you see where the user is looking and you see the struggles that he's doing. So that's the sort of thing I love. Not easy to do. But so I don't like the the last ask last ask this which I do. I do a lot of show stuff. I was talking to you, right? Um I I do a lot of show stuff. Uh, I put it on blogs and I show to random people and I just show it and see the reaction. But I'm actually not listening is if you know you would believe that to be true. [laughter] I'm actually not I'm I'm getting vibes. This is a word now. I'm I'm creating my tokens uh to understand what's actually going on. what's what's triggering people to not like something or like like so I'm actually kind of ignoring all the words they say and trying to get the feeling of what they are experiencing. So I see ah okay I have to change this and I'm trying to rationalize and thinking what did actually trigger you or B and or C uh for this or that uh reaction cuz trusting the the what they tell me uh yeah um just to add on to this I am somewhat biased cuz I do opinion research for for a living. [laughter] Yeah. But um no, I I think there's space for it. Uh definitely, especially like on user surveys cuz they answer different questions than telemetry. Just saying for instance, someone is using this theme doesn't explain what this person thinks about that theme. So I I definitely see the space and potential there, but it has a lot of limitations because we can never reach the representative user. It's always going to be biased. So I would encourage this more research but I would be careful with >> yeah with the conclusions. Yeah. >> Okay. Thank you. So we still have more time for questions and looks like there is no online questions so far but I am still open for Yes. Do you know of any plans to bring back the KD 3.5 theme? [laughter] >> Um, which is plastic. Plastic was fine. That's good. >> Keramic. Oh, Keramic was Kamic was wild. Kamic was serious. Kamic was serious. I want to see more things like I didn't like it personally but it was serious. It was serious design. Someone will actually care to do some crazy thing. Yeah. Back then we had a lot of things like that. Serious experimentation. Nowadays we have flat rectangles. Well they they bring they they figured out they needed drop shadows. At least that they took them like a full week to figure that [laughter] out. some more. And uh regarding the GTK thing, perhaps yeah, just perhaps the question we should ask is how GTK could follow us in this regard because we are setting the path. We are making people dream. It's okay. Not yet. >> Not quite. >> Right. >> So, uh uh it's their problems. Yeah. >> Right. >> And maybe in the future we could work with them helping them catching us. Yeah. >> Um now, seriously, I'm using this word a lot. Um I I like what Tiggate does. Seriously, I do. Uh they are faithful to their design principles. They care deeply about their design principles and they uphold them and they force everyone to use them. That's fantastic. Um but that's uh which created a sort of response like you know the story like Ky we in oxygen we completely ported oxygen over to livita. Was it already lia back then? No, it wasn't called. Yeah. So it was fully working. It was like completely amazing. The GTK apps look more KD than KD apps. [laughter] Um, it worked for a couple of months and then Yep. >> Okay. This will be possibly one last question. >> Sorry. Okay. Last question. >> Yeah. >> I'm going to be pretty quick, I think. Um, I already spoke to you in private about this idea. Um, and I'm just wondering if you can say to the group like I am if we're already bringing up like old styles that were really cool. I have specific fontness of the um Quinn window decorations Fahrenheit which the the specific feature of it was that the buttons were not on top >> but were hanging off the side. Is this I'm if I understood you correctly this is still very doable. It's just nobody does it for some reason. >> So just maybe a keep something to do something fun with window decorations. >> Yeah design something. We need we need better ways to to talk the Uh so there is nothing that stops you from implementing uh such decorations. It's just as you said that nobody implements it and because of it uh we started simplifying some coin uh internal details. So for like for example for constraining window inside the work area. So we now just assume that the title bar in the uh at the top. So um but yeah there is nothing that stops you from implementing that. Once there is such a decoration we can then start looking into um fixing wind.