Episode 798 transcript

FLOSS-798

Jonathan: Hey folks, this week Rob joins me and we talk with Carl Richel, the head honcho over at System76. We chat about their new desktop, Cosmic, we talk Rust, we talk System76 hardware, and even a little bit about Intel and AMD. You don't want to miss it, so stay tuned. This is Floss Weekly, episode 798, recorded Tuesday, August 27th.

It is time for Floss Weekly. That's the show about free Libre and open source software. I'm your host, Jonathan Bennett, and we have a treat today. I know I'm excited about our guests most of the time, but I'm extra excited today because we're talking with Carl Richel about System76, about Cosmic, about.

All kinds of stuff. I'm sure we're gonna talk Russ. We're gonna talk Wayland. We're going to talk the new cosmic desktop, which the alpha is out. It's great. Before we get him, let's talk with our co host. And today we have, we have Rob, we have the, the cereal desktop, the, the, the distro hopper, which is why Rob is on the show today, because I, I was, I was pretty confident that I could talk him into hopping, at least in a virtual machine over to the new cosmic alpha, Rob, welcome.

Hello, hello. Hey! We've been watching Cosmic for a long time now, haven't we?

Rob: We have. Been all curiously interested in all the places they've been going.

Jonathan: Yeah, I think so they they made it's oh, it's been like a year ago now carl will be able to tell us I'm sure exactly how long it's been but it's been like a year ago now They made their initial announcement and the things that that really caught my attention right away are one They're getting away from gnome, which i'm I I like the pop os desktop, but i'm not a huge gnome fan two they were going to rust base, which I think rust is just cool and three they're like And because we're doing Wayland and we're building it from the ground up, we're going to do HDR.

And I'm like, Oh, HDR is cool too. So like from the initial announcement, they had me hooked. And I've just been kind of eagerly awaiting the whole time. Like, Oh, when's it going to be ready? When's it going to be ready? Oh, now you, you gave it a

Rob: test drive, didn't you? I have I've given a test drive on a VM. I, I was really hoping to try it out on bare metal before the show, but you know, time flies and it gets away from me.

Jonathan: Yes, day job and family and all kinds of other stuff going on. And then there's that other show that we keep you busy with too. So we have a lot, but it looks really good. It looks really good. And even in a VM, it does. All right, let's, let's let's go ahead and bring, bring Carl on. Obviously the guy that knows what's really going on.

Carl, welcome to the show.

Carl: Thanks for having

Jonathan: me. Yeah, thanks for being here. This is, oh, this is going to be fun. I, we, we had Carl on back when we were still on the Twit Network. And Leo was so, so excited to be able to talk that he's like, Oh, I want to co host too. And so, you know, it was Doc and me and Leo all co hosting.

Well, I mean, Doc was the main host and then Leo was kind of the main host because it's Leo. But it's, it's great to have you back. And it's great. It's great to chat about, you know. First off, what's happened since then, but I think mainly what everybody's curious about what I'm curious about. So let's talk cosmic and let's

Carl: Yeah, I love to talk cosmic.

So Thanks for having me on to chat about it. I was thinking about the last show We were on together too with that with Leo and I remember distinctly he has he has a one of our darted pro laptops And he, he apologized for running Linux Mint on it, I think. Oh, that's But we're just happy, we're just happy when Linux works that way.

We, we're here to make Linux work on our products, whether it's Pop or Mint or, or Bootery or, or Arch.

Jonathan: Yeah. So I'm actually, I do the show on an HP Dev1, which is not quite your hardware, but it's also pretty much your hardware from what I understand, but running Pop! OS on it still. And I've had a Yeah. Yeah.

99 percent great experience with it. There's something weird with the power charge circuitry, but we work around that we make it work anyways

Carl: Yeah, that was a great product and a great collaboration of the folks at HP. You were really great to work with.

Jonathan: Yeah

Carl: They're very professional. I think Working with them on dev one Also helped us refine POP 2204, because that was coming out at that time.

And so we had just more people with different perspectives looking at what we were building and, and really helped, helped sharpen that, that operating system release.

Jonathan: Yeah. Okay. I didn't expect to get to this question so early, but it, it really leads into it. You did the collaboration with HP and there's, there's another company out there that I think might be really interesting to do a collaboration with.

And that's framework. And I'm, I'm just curious, like, do you, do you have opinions on framework? You consider them a competitor and have you ever thought about that? Like, I wonder if we could do something similar with the framework laptop.

Carl: Oh, I've been, having been in the hardware business for now.

Almost 19 years. I know I know how challenging it is and I, I think it's really admirable how well they've done getting into the market and doing a lot of creative work around right to repair. We have, we have that. But very much in common since day six and, and, and framework both care a lot about their about a user's right to repair, but their ownership over their products.

So so I think I congratulate them on, on doing something that I know very well, how hard it is to do, and they've done a fantastic job. So, so good for them.

Jonathan: Yeah. I, I, I kind of, I can envision a future where it's like either, you know, system 76 makes. One of the embedded main boards that people can put into a framework laptop or framework sells a system 76 version or maybe even makes cosmic, you know, maybe as simple as making cosmic a ship option that you can buy one of our laptops, you can put system 76 cosmic on it.

And everybody wins. Like it just, it, there, there seems like there's gotta be, there's gotta be some way that the cross pollination and that can happen.

Carl: I think that would be possible. There's, I'm very interested in modular mobile devices. So perhaps at a level, A little bit further than where framework currently goes, but very similar in that we have a main board that can be used across a number of different devices.

And you essentially connect daughter boards to that main board for your port layouts, and that might be used for a desktop or a laptop or any other kind of device. I think it takes a lot of creative engineering to kind of compete at a very high level and do that with really good thermals in a mobile platform.

But I mean, doing hard things is the kind of stuff we're interested in.

Jonathan: Yeah, absolutely. Absolutely. Okay. So let's, let's shift gears a little bit. Let's talk Cosmic. What was, what was the point where you guys realized, okay, we need to, you have, you have Pop! OS, which is a very, very heavily, heavily customized.

Ubuntu, and it's a very heavily customized GNOME. And at some point you guys made the decision that GNOME is just not cutting it anymore. I'm curious what, what that looked like. What was, what was the point where this, this harebrained scheme as it were, popped up where it's like, we need to, we need to start from scratch on this.

Yeah, why? Yes.

Carl: Yeah. Why? Okay. Well, it's a long, it's a long story, so I guess I have to decide where to even start, but at the best place is probably with why we decided to make the changes we wanted to make to gnome in the first place. We were. All of our engineers that were working on the Pop OS were using i3 at the time.

When they were using i3, I got became interested in it. And so I started using Sway. Then we were getting calls in our tech support department. I said, Hey, does I3 and Sway run on, on Pop OS or how do I set up I3 on Pop OS? And this was not just a one or two requests. This was extremely common. It was, it was coming in regularly.

Then I went to a, A firmware conference at Google for Open Firmware, which we do a lot of work in that area. And so we were giving a talk at the conference and everyone at the conference was using a tiling window. It became clear to us at that point that what we were providing at PopOS out of the box was not what our customers needed.

So in response to that, we decided to build the extensions we built into Gnome, which, which added auto tiling added a different UX in general with a launcher and application library and a doc. And we were adapting for some things we're writing new, new software for others. We were adapting extensions like dash to doc to provide the user experience that we wanted.

Then we found our users were craving and our customers were craving. So it all started with just the desire to. Respond to what our customers want and build something that they wanted. The next. Step was to try to make that more official. And we wanted to prove what we were doing and know, and then see if we can make more official.

We, we had a, a UX architects or a UX architect. Maria was on the design team. I had gone to a number of quad X and. And over time, it just became clear that our vision for the desktop, and we pitched the idea that you could take, we could kind of separate the idea that the UX is what we And in reality, it's a whole bunch of different pieces that you put together to create a holistic experience.

And we thought that Noam could perhaps take the, take the perspective that SimCity 6 could build our user experience out of it. Ubuntu could build theirs, Noam could build theirs. And the idea was considered interesting, but just not where Noam wanted to go.

Speaker 4: With

Carl: that being the case it became, Clear at that point that we would not be able to continue using gnome and continuously adapting it, you know, working against what was going on with upstream to build the experience our customers needed.

And so we decided to build a cosmic.

Jonathan: Do you, are you kind of of the opinion that maybe Gnome has shot themselves in the foot and some of the other downstream from Gnome are going to have to face similar decisions?

Carl: I don't know. Gnome has their vision for the desktop. I,

I don't share it. It's kind of that simple. What I, I kind of, I love this, what I love about Linux and open source is that it's, it is a collection of all of these different communities, interests and tools and different things that you could pull together to build something with. That's what's special to me about it.

And we don't have that at the desktop platform. That's what, that's the scratch we're trying to itch with Cosmic. So we have this thing where we have customers that clearly need something different because they're just telling us all the time we want to be able to build it, but we can't be the only ones.

I'm sure other distros have different target audiences and they should have a different UX because there are audiences different. We don't all have to be the same. In fact, the vibrance that comes from distros and their response to customers, I think is valuable. So we wanted to build a platform. In cosmic where you could take what we've made and it's explicitly designed to create your own user experience.

So you could build the Ubuntu UX out of it. Very, really, very simply, you can build Manjaro could have their own UX. You could build a UX more like sentiment. Out of cosmic. Those are the things that, you know, those are some of the projects that are, have kind of a unique approach to how they they adapt and customize custom environments, but cosmic allows you to do it at a very deeper level and at a professional level that's very high quality.

So so while it's going to be the flagship experience of Pop! OS, and I'm sure the first release is going to be very similar to what we Push out what we think is right for our users. What's most exciting to me is how people will. Brand cosmic, how they might experiment with different applets and user experiences and ways of launching applications or, or managing applications or moving windows, you know, all those things are up for grabs for, for experimenting with in the community.

Jonathan: And there's already, there's already been some, I know Rob's chomped at the bit, I'll get one more and then I'll hand it to Rob. There's already been other distros I know of. That have expressed interest in, but we could package cosmic. I'm pretty sure Fedora is planning on packaging and shipping a cosmic spin.

And that's gotta be exciting for you guys too.

Carl: It's exciting and encouraging and much, much earlier than I thought. I don't think I, we saw it coming. Not too long ago. And so we're learning how to be a great upstream for distributions. We consider distros our customer. We want to make a great, we want to make it easy to package easy to maintain.

We want to listen to their feedback about what we can do to you know, build a good experience for them. So I think they kind of heard that attitude coming from us that we're really care about you know, empowering distros and and yeah, it's in Fedora. OpenSUSE Serpent is packaged as well.

They're a new up and coming distro. I think that's pretty cool. I hope I'm pronouncing that right. Arch, Arch spin that has a focus on performance, their packaging as well. When they're, I think in the performance reboot, I think uses some different compiled flags or something along those lines, but yeah it's, it is encouraging and exciting and we hope we we hope we do well by everybody and all the distros that are putting their faith in us so early.

It

Rob: is exciting. So I mean, GNOME is a nice interface. If you want to do the GNOME way, but was some of your reason, was it just hard, more difficult or becoming difficult to keep up with their changes and the way their extensions work and having to keep updating yours to work with the newest version?

Carl: No, not really. It's that taking that route means that what we are doing every six months is adapting our extensions to work with the new version of Gnome, and maybe adding a little bit of icing or advancement or additional value. So that work isn't very fun. And we're just working Every six months working to get back to par, which is where we were and maybe adding a little more with Cosmic.

What we've done is built the entire foundation where it's it can be adapted to, as I said, other user experiences. But once we have that foundation in place, all of our engineering and all of our effort is adding features and capabilities, new experiences on top of Cosmic. So it means we'll be able to innovate much more quickly.

Just having this flexible base to build upon.

Rob: With that flexible base, The alpha you have today isn't doesn't have that flexibility yet or does it

Carl: it does yes it's the the alpha as It has essentially all the UI elements that you would need to use the operating system But all of those are also are stable Are interchangeable already.

So we have a panel on top and a dock on bottom, but you can remove the dock, move the panel to the bottom. You can make it more like a windows experience. You can move the dock to the left and then have a panel on top and you can have buttons in the top left, like unity had with Oh gosh, they called it lenses for a while in touch.

I'm starting to mix up some of the older Unix or unity user experience features. But in essence Together with panels and applets, applets are applications that run independently inside of the panel and on top of it. So the combination of panels and applets means you can build in a UX.

You can build a custom launcher that is different than our launcher and maybe is more like perhaps an elementary launcher or something along those lines and add a button to the panel and now you have an experience that's closer to that of elementary. So it does have that flexibility in it. And.

There's lots of heavy customization that can be done through config files. What we expose in settings is what we think makes most sense for users, but we would expect people making different user experiences to change the settings as well.

Rob: So, yeah, I guess I should have put a little more time into my experience that I had with it. But I have I have had some time and I explored around with it only on a virtual machine. I, I should have taken the time to put it on a bare metal. But, you know, before the show we were talking about on a virtual machine because of, I believe you said the graphics acceleration.

It's, it's not quite as Stable there because it did crash on me several times. So could you speak to that maybe? Yeah.

Jonathan: Is that, is that expected? I think that's expected, right? You said during the press call.

Carl: It's, it is expected and it's all because we don't have software rendering in our capacitor yet.

So Cosmic Comp doesn't have software rendering, and so you need hardware acceleration in a virtual machine, which is actually kind of fiddly to set up, and more fiddly than, than I'd like, so so we need to get software rendering done. It doesn't affect anyone on hardware. And of course, all of us developing Cosmic, we're all on hardware.

And so, so one of our focuses will be making sure that you have a good Cosmic experience for the second alpha inside of a VM. Because we understand, you know, the commitment just to install it on, on your hardware is high. It's not, it's not production software yet. So, We want to improve that experience for people in the second alpha.

Jonathan: Okay, I've got to jump in and ask. We talked about moving from GNOME. And this is like partially a troll question, but also partially like a real question. Did you consider KDE as your base? Or any of the other desktops? There, there had to be at least a thought of, I wonder if there's something else that's out there that could do this before you just jumped into the deep end of the pool and made your own.

Carl: There's very little that we didn't consider. We had a big pros and cons list for, for different directions. KDE is, I, they're kind of my spirit animal. I like their approach. I really, I really like like KDE and I like the community there. I think I think they're, they're great. Their response to what users need and want is admirable.

They're, but they're, they're technology stack is something I think there's two things with where, where we decided not to go with, you know, trying to adapt three that where we didn't didn't try to go with adapting K to E to our experience. One is we would probably not make K to E users happy. Not because of what we're doing, but because we might trim down things in a way that they'll say this is supposed to be KDE, but it doesn't have all the KDE stuff.

So, so I thought we might be a lose lose situation there where you know, people feel like it should be KDE, but we're building something different. It's not like KDE. And and I don't know how responsive you know, KDE folks would be you know, to, you know, to like more of a kind of middle ground between all of the options that they have and know.

But in any effect, I thought we couldn't make everyone happy. And we might get in the same kind of situation where upstream is not very happy with us. And, and then customers, I think it's something else or seconds. We're not, um, the technology stack C plus plus isn't a language that we work in.

It's not that. Our team couldn't learn C to go that route, but I just can't imagine being a brush shop and then going that way. Yeah, that's fair. Third Katie is as our vision from the start was that we wanted to build a platform for people to build experiences with. We wanted to be that, that user experience, that UI level that doesn't really exist today.

People are using Android to. You know, make car infotainment systems or, or exercise bikes or, you know, specialized devices. Well, this is a, this is a Linux and open source UI project that enables companies and, and distros to do that on their own. I can't, you could I mean, this would be, this would be something that'd be very easy for, for valve to build a steam deck experience with, for instance.

Oh, so, and it's intended for that. So that's also just a distinction or a difference with what our core intent was in building Cosmic.

Jonathan: Yeah, excellent.

Rob: So being so involved in open source for so many years, Any thoughts on open source hardware, like getting Pop! OS working or in Cosmic working with a RISC

Speaker 4: V?

Carl: I think RISC V is very exciting. It's it always turns up in our channels when something new is happening in the RISC V hardware world. We're being the size that we are. We can really do like one really big project at a time. And so cosmic is that one really big project that we, that we're wanting to knock out and then we'll see what's next.

Jonathan: Yeah. Interesting. With it, with it being built on. Well, with Rust, which compiles basically everywhere now, and on top of existing technologies like like Wayland and all those things. I, I kind of suspect that when you get to the point that, okay, we're ready to do our main release, like, the bulk of the development is done with Cosmic.

You know, for that initial release. I kind of suspect that it's going to be a fairly easy lift to go to another platform. If they have like reasonably well working graphics drivers. I, in fact, you, you might find that somebody from the community beats you to it and makes it work there. We've, we've seen stuff like that over the years, time after time,

Carl: I would be elated to see that.

I think it would be awesome. And that's one of the things that's so encouraging about what's happening with the cosmic community to we have hundreds of people contributing patches, building features. It's the energy around cosmic is very, very exciting. So so, yeah, I think the next extension of that is Cosmic's running on this or, Hey, I built a, I put these, you know, five applets that when you combine them together you have this really unique way of using a computer or using this device that doesn't have a keyboard or, you know, Yeah, that's the stuff I'm really excited for.

And I think if we've done well, then we're going to see lots of that over the next, next few years.

Jonathan: Yeah. Okay. We do have a question from the chat room. I've got a bunch of stuff that I want to get to too, but mashed potato asks also, he says, why Ubuntu and not Debian? Well,

Carl: Debian Ubuntu does a fantastic job of pulling together a modern or, Very recent, but stable base, I think.

So combined with our long experience with the Ubuntu I think we can build a pop OS as a better product with Ubuntu snaps are making that more challenging and I'm not I'm not like anti snap either. I, the only thing I don't like is about snap is I kind of like the technology. I just. I don't like a proprietary store.

That's, I would prefer it to be like repositories that we all are used to in Linux and Flathub, which is, you know, an open store. That's my only beef with it, but I would love to have Snap working in you know, Cosmic Cosmic App Store, for instance, and have it as an option for people to turn on or off, you know, depending on their preferences.

But because we don't want to have a proprietary store in Pop! OS by default, we have to do quite a bit of additional packaging work. Like Thunderbird in 2404 moved to Snap. So we're packaging Thunderbird as a Debian for Pop! OS. So

Speaker 4: it does,

Carl: It does add that work, but. There's a lot of other work that they do that is a layer on top of Debian that makes our jobs easier.

Jonathan: Yeah, interesting. Okay, so Let's talk, let's talk Wayland for a minute. Something that's not been entirely clear to me. I think I know the answer to this. With Cosmic, is there an x11 option at all or are you guys Wayland only?

Carl: Wayland only. But there is X11 X Wayland support in Cosmic Comp, so it doesn't really feel, at this point, the Wayland protocols, and this is just a community wide thing, the Wayland protocols are getting really, really good, really, really close to being, like, I think their NVIDIA is finally playing nice, I know it's been a long road for them, but that was our biggest, our biggest question mark wasn't, you know, For releasing Cosmic was NVIDIA because if if NVIDIA drivers were not working, we would just have to delay until that was until it was working too many of our customers use NVIDIA and care about the, the hardware.

Speaker 4: Yeah.

Carl: So but besides that, essentially what we're, we're at the point where moving to Wayland, I don't think. There'll be some corner cases, but there will be very far few between. Otherwise for most folks it's just going to be a more secure higher quality experience.

Jonathan: Yeah. So it, I think that's one of the things that's actually really compelling about cosmic is that you guys were able to, with the exception of ex Wayland, which I don't think counts in this case you were able to just sort of push aside all of the X 11 cruft and start from a fresh slate.

I think that's really compelling because like, X11 has a lot of cruft around it.

Carl: Yeah, it's, it is interesting that that's something we haven't, X raying it is still a pain. It really is difficult because the, the knowledge to implement the things that are there are just. Buried in some guy's head that wrote it 40 years ago.

Yeah. It couldn't be 40 years. That sounds like way too long. Maybe 30, but that's where it's at. And the documentation is the code is it is definitely challenging. But so we spent a lot of time just working on getting XWayland working so that that experience can, can feel seamless.

Rob: So I assume all of your.

All of your apps, applets, tools are all direct Weyland and not going through the ex Weylandshim.

Carl: Great.

Jonathan: Yeah. That makes sense. That makes sense. Okay. I, I, I sort of watched the Weyland development process and sometimes that's painful. I, and I've got to ask do, do you guys get frustrated with the slow pace that Weyland takes and the, it's, from the outside, it seems like.

Every decision gets bike shedded for like a year before anything gets added. Is it, has that been a source of frustration or are we kind of past where that's a problem?

Carl: No, I think the way we approach it we have two engineers that do most of that work quite a bit in the Wayland protocols, Ian and Victoria, and their approach is in essence, This is something that we have to do to deliver a product that's going to work for our customers and our users.

So if we have to release you know, pre V1 protocol even if we have to change it later to match what's, you know, the eventual the eventual protocol, then that's something that we'll do. And we don't feel you know, we'll, you know, we don't feel like that's you know That's bad because Wayland didn't move fast enough.

It's just you know, it's just part of the process of working in a community with a lot of different interests.

Jonathan: And I think that's probably the advantage. One, one of the advantages with Wayland is because Wayland itself is pretty much just a specification. And the, the implementers are very, very free to add their own specifications, to add their own interfaces on top of that.

And so you have things like what KDE has done, where the KDE guys got tired of waiting for. HDR to materialize in Weyland upstream. And so they finally just said, okay, fine. You know what? We're gonna build our own and because we're gonna have fun with it We're gonna call it frog and nobody knows why but that's what we're gonna call it And hey, look with a couple of days, you know a couple More than days, a couple of months of hacking on it.

You guys can use HDR on a couple of programs in KDE because we got tired of waiting for Wayland. And I think that's secretly one of the superpowers that Wayland has.

Carl: Yeah. And that's an okay approach. I mean, we, we do have to get the things done for, for users and that's just a given and that's, you know, part of building software and building products.

If the consensus protocols take a little bit longer and we have to, you know, adapt to them. Afterwards, that's all right.

Jonathan: Yeah. Speaking of HDR, this is actually one of the, for me, not for everybody, but for me, it's one of the killer the killer features back when, when y'all first announced is that a cosmic is going to have HDR.

And I know, I know that's not ready with version one or the release one, the. Candidate one, whatever we're calling it. But I know last time I asked you about this You said that you you anticipated it coming with the next pre release what what is the what does the roadmap and timeline look like on that now?

Carl: I hope I didn't say that I could I am so optimistic terrible timelines Optimism and bad timelines just don't go well together. But the thing is Well, when I get into a tirade about really big projects, it's not actually possible to predict how long it would take, but with with HDR it is a focus of ours and we're Victoria is the, the you know, lead developer for Cosmic Comp, and she's going to all the HDR hack fest and that are going on for Linux.

And so this is a, it's a broad community wide effort to get HDR working really well

Speaker 4: in Windows.

Carl: I expect, so I don't think the first release of Cosmic will have HDR, but I believe we'll add it sometime in the months afterwards. I just don't want it to be a blocker. Because I think we've got, we nailed so much here that it would be okay to release and then add HDR

Jonathan: afterwards.

Carl: And HDR is also really power hungry. And so there are a lot of considerations with like how often it should be used, particularly on a, like a laptop.

Jonathan: Yeah.

Carl: So there's things to think through there a lot of times. HDR content might be, um, you know it might be a choice whether you want to see something in HDR content all the time.

Jonathan: That's fair. I, it, it comes to mind that I, I asked you the sort of troll question about KDE earlier. But it actually comes to mind that. approach to doing this would just be to re implement and use the, the KDE HDR protocol, because at that point, you've already got a couple of applications that are wired up to use it.

Whereas as far as I know, there's nothing that's wired up to use the, the kind of in process HDR stuff that upstream Wayland is doing. And so if you wanted to get it out the door, Earlier, you know, with the next release just re implement what the, the KDE frog protocol and, and let in player and a couple of others use it.

Carl: I have, I have a strong suspicion. It's not that easy. Okay. That's fair. I don't, I don't, I, it's just a, just a suspicion.

Jonathan: All right, so let's let's see, where do we want to go next? We asked about HDR, we asked about Wayland Rob, yes?

Rob: Yeah, so, I believe, when is the final version of Cosmic expected? Is that later this year, or is that a little optimistic?

Oh yeah, so now the

Carl: optimists bad timeline guy has to make a guess. I really want to make a you'll have a release this year and I think there's a good chance the problem, the problem might be it ends, it lands in like December, which is a hard time to release a big project. It's hard on your, your employees that would like to see their families and those types of things.

So I think if we can't make it by. early to mid November, then we would probably do first quarter of 25.

Jonathan: That makes sense. What is the, what has the reception been like though for the, the, the pre release that's out? Like, do you, do you have a, do you have an idea of how many people have downloaded it and tried it?

And then what's the feedback been?

Carl: Gosh, I don't have. I don't have a good idea. I know that we're in for downloads for the alpha. I know it's into like tens of thousands or something along those lines, but that's just like the Popeye says. I don't know about any other. We don't have telemetry telling us how many people have installed cosmic.

Speaker 4: Right.

Carl: I've almost gotten through most of the press coverage. Cosmic, that's We were, we've been so busy after the release, we had all hands company events and engineering meetings. And, and so we're just all, I'm still recovering this week from all that. But everything I've seen is I think very promising and it's a good response.

I think one of my takeaways from what I've seen is the people that the The time that it takes to build a foundation and the time that it takes to build things on top of it is very, very different. So what people are seeing today, and I think if this took two years, then it might be like two years before it's really there.

But what we're going to show over just the next couple of months is that things will happen very, very quickly. So that's that's one thing I noticed that there's kind of an impression that it's, you know, a lot is there, but. There's a lot to go. They'll see that the things that there are to go aren't really really big projects.

Jonathan: Yeah. That's the, the, the 80, 20 rule. You get the, yeah, yeah. Yeah. That's, that's a, that's a real thing. It really is. Yeah.

Carl: Yeah. And then bug fixes to We're focused on putting features out and getting all the features out and then sharpening everything up. But I don't worry about bugs very much either because every time we go to nail down bugs with few exceptions, those are kind of the quick things.

That's just the quality control and and, you know, turning a focus to all the little paper cuts.

Jonathan: Yeah. Okay. So I've got to ask, this comes to mind because. Another project that I'm a part of, we're dealing with this this morning. Have you gotten to the point, and I'm sure you must have, you've, you've been around, you've been doing, you know, Papa West long enough, if you've gotten to the point to where with Cosmic, you've got users complaining, you made all of these changes and you didn't ask us first.

Carl: I feel surprisingly, no, um, and I read almost every bug report And I mean, I'm almost through all of them that have come in since the alpha started. And there are, there are feature requests. There's not a whole lot of you change this and I hate you yet.

I, I fully expected. So this happened to us when we moved from, you know, Pop! Without our cosmic UX and no. To our cosmic UX and no, and they said, I want my overview back and I want this and this and these other things. And so we were prepared for it. This transition is going to be different because the UX from POP 2204 to 2404 is going to be very, very similar.

You're going to have the same launcher, app library, doc, panel. It's going to feel the same, but everything under the hood is brand new. The apps are new. And so you know, fortunately, I don't think the bar has Too terribly high to reach the app functionality that we had before. With the exception of files, file browsers are just really complex.

And so there's a lot of a lot of work there. I think we're going to, we're going to make it there too, but point being. The UX is going to change. It's going to be what we're going to hear. I suspect is I can't install my clipboard extension or this caffeine extension, or, you know, these other things that we're replacing with applets.

All of that, I think is, is, is absolutely necessary because you know, a desktop that has independent running applications in the panel instead of monkey patch, JavaScript, it's just. Going to be better. So more extensive extendable. So we'll have some pain with that. But I, I think from what we've seen already, a lot of enthusiasm people already building applets to replace those things.

So, you know, hopefully, hopefully it's a very small number of people that we.

Jonathan: That's, that's actually an interesting question. Is there a simple, like, scripting tool for doing extensions? Is, does that exist in Cosmic at all?

Carl: Yeah, we have a, we have an applet template, and essentially you're writing a small application that is embedded in the panel, and it can have its own settings.

So, That's and it provides essentially the same functionality. You can spawn windows. You can modify how workspaces are, you know, presented. There's a lot of different things you can do with applets. So So that is, that's our replacement for the idea of extensions.

Rob: So those applets would all be, the developers would all be making those in Rust?

Carl: Right. And they are all independent applications as well. So if you open System Monitor, Cosmic and you can just search for cosmic and system monitor. You will see an applet for a user applet, a Bluetooth applet notifications applet, every app tray, everything that you see in the panel is an independent application, meaning that it can't step on the other applications.

So if there's something wrong with it, it doesn't crash your desktop. Or if there's you could, and you could add or remove them at will. That also means that we can sandbox them. So a community that's building applets doesn't have to be explicit, you know trusted with access to your home directory.

We can we can sandbox them you know, away from that and have the same kind of permissions and portals access that you have in applications in applets that you have in extensions.

Jonathan: So, so when someone installs one of these applets, do they literally get a security pop up? You know, these are the permissions that this applet wants to have.

Like, have you, have you gotten into that, that fine grain sort of of handling?

Carl: We haven't gotten there yet because everything's first party so far, but there are community applets being built and that's something that we're going to have to think about and work on as well.

Jonathan: Yeah. I, I, I hate to say it, but you know, it's coming.

Someone's going to write an applet and it's going to be down to borrow something from the Android world. This applet turns your flashlight on. And Oh, by the way, secretly, it also reads your clipboard and sends it to us. Like that's the price of success. It's

Carl: true. Yeah. And so it's, we'll have to work, we'll want to work on it early.

We are going to have, we'll have a repository for applets and Every every poll, we're going to do engineering and and design review for applets that go into cosmic, and that will be available in the cosmic store. So to I mean, there's going to be more effort for us up front, but because our toolkit is very young, because the community is very new.

And the project is new, it's just going to mean there'll be a little bit longer review process, but we are going to actually do full code reviews. And in this case, even design reviews and provide design feedback so that so that applets can fit into the cosmic UX. And that's largely because the our templates and our widget library is just.

It's just young. So it's it's, it might be as a developer today, hard to turn to know exactly like how, and our documentation is, is young too. So how, you know, how should I, where should these buttons go to fit the UX and some things like that, that we just you need to, Tighten up for the community.

Jonathan: Yeah. Okay. So to get, to get an application into the store, I assume it's got to be open source and use an OSI approved license.

Carl: Not necessarily. I don't think that would be a requirement, but we haven't seen anything that isn't.

Jonathan: I just, in thinking through this, it's like, well, how do you, how do you do code review if you can't get to the source?

Carl: Oh, that's a great point. No, you're absolutely right.

Jonathan: And then I guess the question that, well, so, I mean, it's interesting. This is all very new stuff to you guys. You're, you're, you're sort of thinking through all this as you go along. I think the question that kind of naturally follows after that is, is there even a mechanism for someone to install an applet that does not come through the store?

Like you, you would probably want to discourage that, but at the same time, you might want to still allow it. It is, is that even an option? Is that a thing?

Carl: Oh, yeah, absolutely. You can download code from a repo. You can do that today. That's where most of these applets reside. It's in a repo spread out on GitHub.

You can download those, install them and add them to the panel.

Rob: Yeah. Okay. So, so with each applet being its own program. Would other developers be able to make like a C program or anything and actually run as an applet?

Carl: That's, so it is possible with a different toolkit. So our ICE toolkit and I hope I'm, I'm hoping I'm getting this right, but with our, with our toolkits, I believe it requires that it has to be written in Rust.

If you're using Slint, which is another toolkit that has as libcosmic widgets or, or matching widget widgets, you'd be able to use C or C or you know, the other languages that Slint supports.

Jonathan: Yeah. Very cool. Is, is. Is Cosmic built on top of one of the existing compositors? Like, is it, does it use WL roots or, or one of those?

And I can't remember. I know I've looked this up before and I can't remember off the top of my head.

Carl: No, it's it's a Bruss compositor. That's it's based on a community project called Smithing. Which and so we, we hired Victoria. She was the lead of the Smithy project, that was her project.

And so Victoria was developing Cosmic Comp and Smithy is a big part of, we have a Smithy toolkits and a lot of other things we've built around it over the last couple of years. So that's all all brushed and all all new.

Jonathan: That's okay. That's, that's really cool. I did not realize that you'd gone out and you, you hired the developer that was building the thing that was really similar to what you wanted to build.

That's, that's actually really neat. One of the things that we talk about here is developer open source developers have rent to pay and they need to eat too. I, oh, that, that is really cool that you guys, you guys found a developer and hired him, brought him on board. And so you're, you're kind of bringing the existing project on board.

Making it work with what you need to do. I, I think that is an absolute win for everybody.

Carl: Yeah. Yeah. It's, we wanted to build a the compositor. We thought that was a, the compositor has a huge impact on the field of the operating system. It is kind of central to, to the OS experience. And so we knew that was going to be a key part.

And Victoria is absolutely incredible. An amazing person, amazing engineer. So we're very, very excited to have her on the team.

Jonathan: Do you see the possibility of other, like, desktop environments that, so, someone else out there that doesn't want to use Cosmic but wants to use Smithy? It's that, like, I assume that's possible.

Is there anybody doing it? Do you anticipate that happening?

Carl: I think what we're seeing happen is that people that want to work with compositors but don't want to work in, C or C they want to work in Rust they're going to SmithA and using SmithA as their target for things like accessibility, for instance, and other, you know, unique areas where we want to, you know, improve the Linux stack.

Jonathan: How long have you guys been a Rust shop? System76?

Carl: Probably six years, I think. A lot has changed in that time, hasn't it? It certainly has, but when we started Cosmic, there wasn't, there's still the, the breast windowing environment and ecosystem is very, very nascent. And when we started Cosmic, I know more than two years ago, it was very, very young.

Yeah, there was, you didn't build applications in Rust and today, well, you could. And and I might've seen the bindings, the bindings for GTK are actually quite good. And we were used to using those as well. So you can build apps that way, but not the pure stack that we wanted to build. Where the toolkits, the widgets, everything about the application was written in Rust, including like other things, like there was no text rendering.

So. We had to write text rendering, but the really cool part about all of this is it's a it's a growing community of really passionate and incredible engineers and we all, we're getting, we get to be a part of providing the tools that they're using to build things with now. The Cosmic Text is becoming the standard.

Text render for for Rust Rust applications. You know, that's, that's the awesome thing to be able to contribute to the world.

Jonathan: Yeah, it's, it's, it's really neat. I know something that we've seen like as part of Rust making it into the Linux kernel, if for one thing, it's, it's kind of a, a. A shift for the kernel guys, because now they've been reading and writing in C so long, and they have to add this other language they sort of had to be familiar with.

But one of the other things that's interesting to watch for that is, as Rust is being added to the kernel, Rust itself is changing as a result. They have had to make You know, a lot of changes. They've fixed things. They've made things better in Rust. Have, have y'all seen sort of something similar where as you're building a compositor out of Rust and as you're building these different pieces, that some of the things that you had to work with have now made it back into the Rust language?

Carl: Yeah, I don't know details. So, you know, about that, that's, but I, but I know that everything evolves together if it's going to move forward. And that's the, and the, the base language and language itself is, is no different than any other component. You know, evolves for its use if they're, if you're doing things well.

Jonathan: Yeah, yeah, it makes sense. Are, are we at the point to where we, we are up and coming programmers? We probably need to just start teaching them Rust definitely in addition to C, but you know, is, is there a, is there a point whether it's now or in the future that we teach them Rust instead of C?

Carl: I would think so.

I mean, I went to a memory safety. By default I know I hear the borrow checker is a pain in the ass, but but I think once once you get it, it just, it's a, it makes a lot of sense. I, I don't like language wars. So I don't know how far I

Speaker 4: would

Carl: go into it, but I know it's the, I think building a the bout of memory, safe code and coverage that we're going to have, you know, obviously we have to use, you know, you know, non safe for us for lower level integration, you know, C libraries and other things like that, but Cosmic has so much coverage you know, safe for us code that I think it's.

By the time we're done, I think it's going to be a far more secure operating system and have more coverage than anything else that's out there.

Jonathan: Yeah, it's it's definitely been interesting to watch. And you know, there's there's been things that have happened like Rust developers Every once in a while, I get reminded that like, there are, there are bugs out there other than memory bugs.

We're just, we're so used to those being the norm though, with languages like C that we sometimes forget that, oh yeah, logic bugs exist too.

Carl: Yeah, I think that, I think the I mean, why, Well, we so often hammer memory safety homes because of so many critical vulnerabilities are safety vulnerabilities and you know, things that that we can solve at the language level.

Jonathan: Yeah, absolutely. Okay. We've, we have talked about cosmic. We've talked about Russ. We've talked about Wayland. I want to ask more about system 76 itself. And. I guess first off, is there anything, is there anything upcoming with system 76 as a company that you, you want to plug or let people know about?

Like, is there, is there some secret new hardware that you want to announce on the show? What's what's coming down the

Carl: pipe that you're

Jonathan: excited about?

Carl: There's going to be a new workstation. From system 76 called Thaleo Astra.

Speaker 4: And

Carl: Thaleo Astra will be unlike any other desktop that we have in the, in the line.

And that's all I have to say so far. That's fair. That's fair. It will, it is. Extremely performant and very unique desktop. And we consider it a workstation because of its level of performance and its uniqueness, but I think there will be a lot of people excited about about this product.

Jonathan: Yeah.

Now, I, I honestly cannot remember. I think your. Intel shop. Do you offer AMD on any of your hardware? I just, I can't remember off the top of my head.

Carl: We do. AMD is killing it on desktop. They really are. Intel's fourth generation is quite good. 14th generation. So we have Ryzen 9, 000 products. We have Intel well, this is desktop Intel 14th Gen 15th Gen, I think it's coming in October and Threadripper products on on some workstations.

We also have a a value desktop that is 12th Gen Intel. It's because a lot of people schools and and other, you know, environments don't need You know, the latest and greatest. They need, you know, value. So that's what that price for on the mobile side. We have rising laptops into Intel laptops.

And and NVIDIA graphics on the higher end laptops. So, so really, it's pretty broad spectrum. I think that's one of the interesting things about Pop! West 2 is when we were, when we were building pop, we thought, okay, we're just going to make sure it works really well on our hardware. It turns out that our hardware, we cover so many different it's a pretty good representation.

Yeah. Yeah. And it just ended up that, well, Pop! West works great on hardware because our lab is full and it needs to work well. So so yeah, we have we have a lot of, depending on what the, what the user's needs are, customer's needs are, Intel or AMD or AMD graphics or Nvidia graphics, kind of the full spectrum.

Jonathan: Yeah.

Carl: Which I think is a pretty, it's an amazing thing.

Jonathan: That is really cool.

Carl: That Linux Ships with on everything that's new today. Yeah, that's that's amazing.

Jonathan: Yeah, okay You you may not want to answer this one and that's fair I'm curious though. Have you guys gotten bit by the the Intel 13th and 14th generation problem where where the chips try to eat themselves?

Carl: So yeah, we're it's a familiar with it But at its base, I believe that this Well, that's our, our take is that this was a weatherboard manufacturers, redlining their

Speaker 4: firmware.

Carl: And it was just it wasn't within specifications. I don't know why it was so different than previous generations with Intel specifications and how they were implemented.

But what we found, we spent, we, we did have some customers find that, okay. I've got instability when I'm compiling. And that's what our customers are obviously there, you know, the common use. Heads up. Yeah, so we did so we identified the problem. We worked through it with Intel. We worked through it with our upstream board manufacturers and, and within a few weeks we had a stable firmware shipped out to customers.

So I don't know. I don't know if it was as big as it's, like, we've got a As it's, you know, came out to be I know I've heard oxidization things and other you know, stuff. I don't know. What is, what is your take?

Jonathan: Well, let me, let me put it this way. I have a customer right now that bought two new desktops from another vendor, a very, very large vendor that I, I guess I won't name a very large vendor though.

And they have had problems with those desktops being unstable the entire time that they've had them. And they are now running on the most, on the latest board firmware, and they're still having problems with them being unstable. And so we are probably today going to, we finally figured out that this is.

almost certainly what it is because they're 13th gen Intel. And so we are today going to start the fight to get that vendor to RMA the chips. And I don't know how that's going to go, but I'm reasonably certain that what has happened is that those chips have, have eaten themselves to some extent. I

Carl: have the perfect solution for you.

It's nice being a you know, small, medium sized, hardware manufacturer because When our customers contacted us, it was the guy that's 15 feet away from the QA guy that's doing the testing, reproducing problems, was getting calls and tech support about a problem and walked over and talked to him and we said, yes, this is a problem.

And we're able to work through it pretty quickly. I mean,

Jonathan: so that's, that's like been the biggest problem with working with the bigger vendors is you call in and you get You get their first tier of customer service and those guys are programmed to say, oh, no, no. There's nothing wrong with our hardware.

It must be, why don't you go and reinstall windows? That'll surely help. That'll fix it. You know? Did your reboot? Yeah. Have you rebo, have you rebooted the computer? Have you rebooted your router just to make sure right. Oh, okay. Have ha I've got, I've gotta ask, have you guys seen have you had to RMA any hardware as a result of this?

Is that, has that been a thing?

Carl: We haven't had to army hardware. We did have one customer who just because of speed decided to move to thread ripper.

So, you know, and they, they had a vote of confidence for pressing, Hey, we know you're going to get this, but we've got, you know, we had we need a fix now.

So we just swapped their desktops out for them. But but no, not something like the not, not really like, it wasn't a return problem.

Speaker 4: Sure.

Carl: That's, that's something you're always watching out for. Yeah. Because if there's, if there's a pattern and it means returns, then you've got, You've got a big problem.

You've got to get something fixed. You've got to get, you have something to fix.

Jonathan: Yep. Absolutely. Rob, do you have anything else you want to get in before we, we get towards wrapping? No, I don't have anything else. Okay, okay. It has been great. I've got to ask, Carl, is there anything that you wanted to tell folks about that we didn't ask about?

I feel like we've covered a lot of things, but is there anything that's just burning?

Carl: Oh, I don't know. I would just urge everyone Play with play with cosmic. It's an early alpha, but

Speaker 4: it's

Carl: tech and it's fun. It's just a fun desktop. And that was you know, our intent is not just to, you know, build something because we're a company and delivered to our customers.

We really just love technology and want to make it fun. And so we hope it's something fun for you to tinker with and play, build stuff. So let's get out there and make stuff.

Jonathan: I do. I do know what I want to ask. That is what, what is the process going to look like for people that have existing Like system 76 hardware running Pop OS.

At what point is Cosmic going to roll out to them? How does that going to work?

Carl: Once our release is final, then there will be a button there. You'll get a notification that says. PopOS 24. 04 is available when you click it, it'll you know, there'll be plenty of disclosures, like you're upgrading to a new this isn't just your typical, you know, upgrade to a new version, here's a new desktop.

Yes. So we want to be very transparent. That's okay. You're not going to have gnome anymore. When you do this, you don't have to do it. You can stay with where you're at and it's going to be supported for, you know, a few more years. So you can feel comfortable there. But, we're, we're we're moving to the future and a new desktop.

Rob: Yeah. So it's good. I was going to say, so you're not going to have a pop up every, every few hours. It pops up and said, time to update, time to update. No, like, you know, some other companies have done.

Carl: Yeah, we'll have to be we, we want to be gentle. Because you do want, you do want to keep people up to date, you know, and there's, there's a balance between being a nag and doing that well.

So I think we're okay with it.

Jonathan: Yeah. So it's going to be, it's going to be the default. And is it going to be the only choice in 20, in 24. 04? Or can people still run Gnome if they want to? Is it, is it even out there still?

Carl: They can run gnome just like running like installing gnome desktop. Gosh, I think that's the the right meta package.

Ours is cosmic sessions or cosmic sessions. I'm thinking, I think it's gnome desktop, but yeah, they'll they'll be able to install gnome and run it if you like.

Jonathan: And it'll

Carl: be it'll be vanilla. You know, so yeah, that makes sense. Comparison like

Jonathan: What you guys don't want to have to do is continue to maintain the Pop!

OS desktop experience on two different desktop environments, right? That's just, that's, that would be ridiculous.

Carl: It would be a lot of Yes, providing the same experience. Yeah, no, we're just we're going to pull the bandaid off and run for the future.

Jonathan: Yep. Yep. Well, I, I've got to say we, we have the Linux show.

And so we talk about this from time to time and I've got to say it's become the thing we will tell people don't install some weird niche distro for your first distro. Don't install some weird niche distro for your grandma or your friend on their first time on Linux. Stick with a distro that's. Solid and that everybody knows something about and so we have this kind of list It's like you use Fedora use Ubuntu or use pop OS And so I think it's it's really saying something that you guys are making that list And in such a short time too.

Yes. Yes And What's really fascinating me is that I think we're getting really close to the point to where it's going to be and you should really set them up with one of the established desktop environments, you know, like KDE or gnome. And I think we're about to add cosmic to that list, which again is really saying something.

But it's also, it's also really exciting, like for the community to have a third real choice there. And so I think, I think that's neat. Yeah. That's great.

Carl: I do too. I, I am of the belief that there can be no, there's no such thing as too many distros, no such thing as too many des I think the whole, it's fun to create things.

Go out and make stuff. That's where the vibrance and the that's what's awesome about Linux.

Jonathan: Yeah, I would agree. I would agree. Alright, so we've gotta ask, what is your favorite text editor and scripting language?

Carl: Text editor? Well, right now I'm using Cosmic Text. So I don't know It's the new default scripting language.

I have no idea. Python.

Jonathan: There you go.

Carl: Yeah. Oh,

Jonathan: I reminded the old Dilbert cartoon. You have drank from the cup of management and so you don't write script anymore.

Carl: I don't know. Well, and the real reason I said Python is because that's what I used to write.

Jonathan: That's fair. It's a thing. It happens. All right.

Hey, thank you, man, for agreeing to come on. It's been a blast. The hour has just flown by and we'll have to, we'll have to have you back. You know, once, maybe after, after the 20, 24. 04 drops and everything has, has gone solid, we have you know, full releases and We'll have you back and ask about how it went.

It'd be fun.

Carl: I'd love to be back. Thanks for having me.

Jonathan: Awesome. Thank you. Thank you so much. All right, Rob,

Rob: what do you think? I'm going to be trying this out on bare metal one of these days and I'm going to be daily driving it way before I should be.

Jonathan: Yeah. I, I feel like I probably will too. I, I, I kind of want to, I kind of want to jump to a jump to it on the laptop here.

I need to wait because this is my production machine, but it's, it's, it's fun.

Rob: I'll have other machines to work on. It won't be my, my only one available. If it, if it, you know, if it's not quite ready yet, but yeah, I think it's ready enough that I could use it for most things, most of the time.

Jonathan: Yup. Yup. That's a lot of fun.

Oh, all right. Well, hey, do you have anything you want to plug?

Rob: You know, just come check me out on the Untitled Linux show, or if you want to Find me directly. Go to robertpcampbell. com.

Jonathan: Yeah, very cool. All right. I want to let folks know that next week we are planning to talk with Laurie LaRusso about Percona.

And that is certainly going to be a lot of fun. Percona is open source database software. And then the week after that, we're talking Ladybird, which Ladybird isn't from the ground up. Browser because there's basically only two browsers right now. And so these guys think there really should be a third one.

And so they're working on Lady Bird. That's gonna be a lot of fun too. So some neat stuff coming. As far as finding me, of course there is Hacka Day. Keep an eye on the Hacka day. We've got the security column goes live there on Fridays. And then there's also the Untitled Linux Show, which I do with Rob and the rest of the guys over at TWIT TV on the TWIT network.

And that is a blast and you should definitely check it out. There's also my YouTube channel, if you're really, if you really want more and that is mostly MeshTastic content these days, but if that's something that you, you want to get into it's youtube. com slash at J. P. Bennett, I think. I don't have that pulled up, I believe that's what it is, or you can just search for me on the, on the YouTubes and you'll find it.

Some really exciting stuff in MeshTastic coming with the 2. 5 release coming up, so check that out if you're interested. Other than that We appreciate everybody being here. Those that catch us live and get us on the download. Thank you so much for watching or listening, and we will see you next week on Floss Weekly.