LINUX Unplugged - 578: Young and the Rustless

Episode Date: September 2, 2024

Rust meets Linux in a clash of coding cultures. Why some developers are resisting, and where things go from here.Sponsored By:Core Contributor Membership: Take $1 a month of your membership for a lifetime!Tailscale: Tailscale is a programmable networking software that is private and secure by default - get it free on up to 100 devices! 1Password Extended Access Management: 1Password Extended Access Management is a device trust solution for companies with Okta, and they ensure that if a device isn't trusted and secure, it can't log into your cloud apps. Support LINUX UnpluggedLinks:💥 Gets Sats Quick and Easy with Strike📻 LINUX Unplugged on Fountain.FMBerlin with Brent: September Meetup @ Nextcloud Conference, Sat, Sep 14, 2024Berlin Buds on MatrixLinus Torvalds talks AI, Rust adoption, and why the Linux kernel is 'the only thing that matters'KubeCon + CloudNativeCon + Open Source Summit + AI_Dev China 2024Rust for filesystems — At the 2024 Linux Storage, Filesystem, Memory Management, and BPF Summit, Wedson Almeida Filho and Kent Overstreet led a combined storage and filesystem session on using Rust for Linux filesystems.Filesystem in Rust - Kent Overstreet and Wedson Almeida FilhoAsahi Lina: "I think people really don't appreciate just how incomplete Linux kernel API docs are, and how Rust solves part of the problem."One Of The Rust Linux Kernel Maintainers Steps Down - Cites "Nontechnical Nonsense" — One of the several Rust for Linux kernel maintainers has decided to step away from the project. The move is being driven at least in part due to having to deal with increased "nontechnical nonsense" raised around Rust programming language use within the Linux kernel.[PATCH 0/1] Retiring from the Rust for Linux project - Wedson Almeida FilhoOn Rust, Linux, developers, maintainers — There's been a couple of mentions of Rust4Linux in the past week or two, one from Linus on the speed of engagement and one about Wedson departing the project due to non-technical concerns. This got me thinking about project phases and developer types.The revival of the Linux C++ discussionMembership Summer Discount — Take $1 a month of your membership for a lifetime!Thank you Core ContributorsBrewHouse by the Lake – Amsterdam Brewery ShopBITCOIN WELL — The fastest and safest way to buy bitcoin in CanadaThe new JB server - KTZ systems — Join Alex, Chris, and Brent as we fly to Toronto to deploy our shiny new colo server in Canada. We'll be deploying the 45homelab HL15 server.Projectivy LauncherAerial Views ScreensaverOpenEBS — OpenEBS is a modern Block-Mode storage platform, a Hyper-Converged software Storage System and virtual NVMe-oF SAN (vSAN) Fabric that is natively integrates into the core of Kubernetes.oppy1984's Satoshi SurveyCoder Radio 582Keybase Filesystem Storage LimitsI Don't Care for GnomeDAVx5 — DAVx⁵ DAVx⁵ – CalDAV / CardDAV / WebDAV for AndroidFlock 2024 Universal Blue: Building the Future using Fedora AtomicPick: Shotcut - New Version 24.08Install Shotcut on Flathub — Shotcut supports many video, audio, and image formats via FFmpeg and screen, webcam, and audio capture. It uses a time-line for non-linear video editing of multiple tracks that may be composed of various file formats. Scrubbing and transport control are assisted by OpenGL GPU-based processing and a number of video and audio filters are available.Bonus Pick: Butler — Access your Home Assistant dashboard from a native companion UI, integrating better with your OS.Butler on GitHubRust for Linux revisitedRedox OS — Redox is a Unix-like general-purpose microkernel-based operating system written in Rust, aiming to bring the innovations of Rust to a modern microkernel, a full set of programs and be a complete alternative to Linux and BSD.

Every release, about half the people involved sent just one patch. And a lot of them never show up again. They may have something small they wanted to fix that they cared about, and they were not really colonel people. They found it for some other reason. And they sent their small patch to the colonel, and they were never interested in doing anything more. But then the other half keeps coming back. to the kernel and they were never interested in doing anything more, but then

The other half keeps coming back and and when it comes to rust I'm not going to be the one who manages the rust care because that's not my expertise as is true of so many other parts of the kernel I'm honestly I'm less of a programmer these days than I am, I call myself a technical lead because I'm not a manager. I don't manage people, I manage code. So I call myself a technical lead. I'm not, my day-to-day work is not programming, it is merging other people's codes and Rust will be one of those things.

Hello, friends, and welcome back to your weekly Linux talk show. My name is Chris. My name is Wes. And my name is Brent. Hello, gentlemen. Well, coming up on the show today, Rust in the Linux kernel. Was it like inviting a crab to a clam bake? Things got pretty messy this week.

We'll dive into all the debate, the burnt out maintainer, and try to figure out where this whole thing is heading. Is it going to be a feast or a fiasco? We'll talk about that. And then we're going to round the show out with some great boosts, some picks, and more. So before we go any further, let's say hello to our virtual lug. Time-appropriate greetings, Mumble Room. Hey, guys. Hello, guys. Hello. Nice to see you. We've got a tight crew in the on-air

and we've got a big crew up there in quiet listening. Thank you for joining us today in the Mumble Room. Yeah, they look angry about Rust, so it's perfect. And they all wore crab-themed costumes. That took some serious coordination. A big good morning to our friends over at Tailscale, tailscale.com slash unplugged. The easiest way to connect your devices and services directly to each other wherever they are.

It builds out a mesh network protected by... Waggled. I used it as an exit node service, so kind of like a more traditional VPN for this particular week while I was in the hotel. I just don't want them, you know, spying on me. I know, I know. You don't trust that Canadian internet.

And we've got Tailscale on some VPSs in exit node mode. So when I'm traveling, I just pick from where I want to come out. So I came out from Seattle so I could watch my local Seattle TV while I was in Canada. Beautiful. It's pretty neat. But you can really just build out any kind of network infrastructure you want over Tailscale. We're using it more and more all the time. It is the easiest way to

connect devices and services to each other. Get it for free on 100 devices at tailscale.com slash unplugged. Well, just a couple of quick things to note before we get into today's oversize show. Berlin with Brent is coming up

and that's pretty exciting. Hey, that's me. That is you. Yeah, I'll be in Berlin real soon. We have a meetup all organized that's in conjunction with the NextCloud conference. It's September 14th. Please come and join us if you're in town, if you're close by. We have

a great many people showing up, which I think will be really exciting. But the key here is if you'd like to join this or future events that are happening in Berlin, because there will be some, we have a Matrix room dedicated to the cause, Berlin Buds on Matrix. We have a link to that in the show notes as well. And I also want to point people to a special video that Alex has released of our fun up in Canada deploying our server. We'll have that in the show notes. We'll talk more about that in a little bit. And we'll also have a Canada meetup recap coming up very soon.

So stay tuned for that. But first, let's get into this Tempest in a Crab Pot situation. This week, there has been an eruption of debate around Rust in the Linux kernel. Also this week, a longtime Rust for Linux maintainer burned out and resigned. Boy, oh boy, are the crab pots out, boys. So I thought before we get into all of this, and there is a lot to get into, maybe we should zoom out and start with why there's all this work to put Rust in the Linux kernel in the first place, and then we'll get into the week's developments and talk about where things are going. Yeah, why are people even trying? Why do people care about this?

Right. And the reality is the C remains the primary language of the Linux kernel, and the incorporation of Rust into the Linux kernel is still pretty new. But the hope is that you get some security there, some reliability, maybe maintainability for, well, a certain type of developer. We'll get to that. There's a lot of reasons people love Rust. We don't need to really kind of sell anybody on it. It seems to be selling itself.

But one of the things that's come up in the conversation around Rust is long-term sustainability. And that's something that Linus addressed just a few months ago, about almost eight months ago during a Q&A with Dirk Hondale. And he talked about why rust in a context that wasn't exactly technical, but still absolutely critical. What you call at the top of the maintainer tree, we are certainly seeing a significant aging there. Obviously, as time goes by, that has to happen. there. Obviously, as time goes by, that has to happen. But if I look five years into the future,

and a lot of people will start hitting the 60s, and the first ones will approach the 70s. So where do you see this going? So I mean, to some degree, it's a good problem to have. I mean, so the Colonel Summit was, what, a month ago, something like that? Three weeks. Three weeks ago. And it was actually the first year when I personally reacted to the fact that, yes, a lot of us are going gray. But at the same time, part of the reason really is the Maintainer Summit is we try to limit it to about 30 people or so. Of those 30 people, at least three of them had been around for more than 30 years.

So they had been around since, like, the first year of Linux existing. And the fact that they are still around and still active and still end up coming to maintainer summits means that, yes, they are older and grayer, but it also means that we actually have a community where people do stick around.

But that's a double-edged sword. Absolutely. And it's, for example, one of the things I liked about the Rust side of the kernel was that there was one maintainer who was clearly much younger than most of the maintainers, and that was the Rust maintainer. We can clearly see that certain areas in the kernel bring in more young people. We had in the maintainer, at the maintainer summit, we had this clear division between the file system people

who were very careful and very staid, and they cared deeply about their code being 100% correct, because if you have a bug in a file system, the data on your disk may be gone. So these people take themselves very seriously and their code very seriously. And then you have the driver people who are a bit more okay.

Especially the GPU people seem to be like, anything goes, right? And you did notice that on the driver side, you have a much easier time finding young people. And that is traditionally how we've grown a lot of maintainers, including, I mean, Greg, with no hair. I apologize, Greg. You can beef me up later.

I mean, my forehead is getting larger every year, so what can I say? So is mine, and my ears and my nose, but that is life. But that aspect, that angle of bringing somebody on that's a little bit younger that can work on the newer stuff, I think the joke in there about the GPU driver folks is probably the Asahi project. But again, as somebody who just spent the last week using Asahi Linux, I'm really grateful for that work that's being done. And they've got it done quick, too. And it doesn't mean that all of the Linux kernel is written in Rust.

Right. I mean, that's not going to happen on any foreseeable time frame. No. We're really kind of just figuring out where it fits. These kinds of things, they can take forever. Like, I like to think about the real-time Linux project. That thing's been going for like 20 years. And we're just kind of seeing some of that stuff properly upstreamed.

That's almost done now? 20 years. So one of the things we've talked a lot about over the last few kernel releases is the introduction of Rust into the kernel. And I think it has been relatively steady and quiet. What's your perception? Well, so Rust still is at this point where we don't, we have the initial infrastructure that merged last year.

It's been growing but we don't have any part of the kernel that really depends on Rust yet. To me Rust was one of those things that, A, it made technical sense, but to me personally, even more important was that we need to not stagnate as a kernel and as developers. And so I am always excited by trying something new and not getting too comfortable doing the same thing. I mean, I've been working on the kernel now for 32 years. Yeah, 32 years. And that's a long time to work on one single thing. But it's still interesting because it's not the same single thing.

I mean, Linux 32 years ago was very different from what Linux is today, obviously. And I actually often look for things where we can do new things and we can do things differently because it's so easy to get stuck in a rut and say, this is working just fine, right? And Rust has not really shown itself as the next great big thing, but I think during next year we'll actually be starting to integrate drivers and some even major subsystems that are starting to actively use it. So it's one of those things. It's going to take years before it's a big part of the kernel,

but it's certainly shaping up to be one of those. So you're writing Rust code yourself? You're reviewing Rust code? Oh, no, no. I have been reading Rust code a bit just so that I... You know, he actually is, he gives himself a hard time there, but he's been pretty good at at least trying to be able to read it and comment on it and give feedback. But that was eight months ago. So we had a more recent Q&A with Dirk at the Open Source Summit in Hong Kong just last week. And Linus updated us on his Rust thoughts. Yeah, he actually expressed some disappointment that things weren't going faster. Quote, I was expecting updates to be faster. But part of the problem is that old time kernel developers are used to C

and don't know Rust. They're not exactly excited about having to learn a new language that is, in some respects, very different. So there's been some pushback on Rust. Yeah, and he also noted that the Rust infrastructure itself hasn't really been as super stable, so it's not like a lot of people can jump on board. So we have a lot of daylight between the comments from Linus eight months ago and the comments from Linus last week and in between that time period there has been a lot of developments behind the scenes a lot of frustration that has been growing it's it is in part you, developers are used to see and they don't know Rust. But that doesn't actually paint the complete picture of what's going on here. There are technical concerns as well.

And I think that's what we should get into now. 1password.com slash unplugged. Number one, password.com slash unplugged. Quick question. Do your end users always, and I do mean always, without exception, work on company-owned devices and only use IT-approved apps? No, of course not. Even if they tell you they are, they're not really. So here's my next question. How do you keep your company's data safe when it's sitting on all those unmanaged apps

and devices? I walked into a situation once where the company didn't realize that half the staff had installed Dropbox. They had no idea that their data was being synced off-site, and they were a military contractor. 1Password has the answer to this question. Extended access management. Boom. By 1Password, their extended access management helps you secure every sign-in for every app on every device because it solves the problem that traditional IAMs and MDMs just can't touch. It's security for the way we actually work today, and it's generally available for companies with Okta and Microsoft Entra, and it's in beta for Google Workspace customers now. So go check it out and support the show. 1password.com slash unplugged. That's the number 1password.com slash unplugged. We thought we might dive in a little bit more into

the particulars by examining a talk given back at the Linux Storage File System Summit earlier this year. And there were a lot of players, including Rust for Linux maintainers, Kent Overstreet was part of the talk, and then a whole bunch of prominent Linux developers who work on storage and file system and BPF and other things were in attendance in the audience, and things

got a little heated. Realistically, this is what James was going to ask, is that this is kind of encoding the object model into the function prototype. What happens when we change the object model in the C code? We're going to have to change it here, too. So this is actually a problem.

Okay, yes, it will have to be changed. Who's going to do that? It will have to be done at the same time because we can't change the C code and have ROS code that behaves differently. And so there's a disconnect there that we need to

resolve. Yes, it's part of what we need to discuss. Yes, that's right. I agree with that. So are we looking at like a fundamental sort of problem of two different APIs in two different languages? that's right. I agree with that. So are we looking at like a fundamental sort of problem of two different APIs in two different languages? Yeah, right.

I mean, you're adding things here and where you have shared infrastructure, maybe we get to a point where eventually some stuff is, you know, more and more stuff is implemented in, say, Rust only. And if you want that feature

or subsystem, then you need Rust. But for a lot of the shared infrastructure, you're going to have a C version and you're going to have a Rust version. Yeah, and once it's in maintenance mode, if somebody makes a change in the C version, how do you know the Rust version is properly updated and working? And how do you even know you need to know? Right. And then like a lot of this work,

you know, a lot of what was being discussed at this talk is, you know, the Rust devs were basically trying to get information from the experts in the, you know, the people in those specific file systems or subsystems or whatever, and then encode that into the types in Rust so that you could, you know, the compiler could then check your work and take advantage of that. So they're basically trying to get like, hey, what is the specification or how does your thing work? And then they can check at build time. Right. all that. But then that means kind of what they're talking about here then, if devs want to go update or improve the

model, the way the C code works, the interface, then all of that encoding work on the Rust side needs to make sure that someone has to make sure that it keeps up. We're almost out of time here, and I suspect part of the problem here is you're trying to convince everyone to switch over to the

religion as promulgated by Rust. And the reality is that ain't going to happen because we have 50-plus file systems in Linux. They will not all be instantaneously converted over to Rust. Before that happens, we will continue to refactor C code because we want to make the C code better. If it breaks the Rust bindings, at least for the foreseeable future, the Rust bindings are a second-class citizen.

And those file systems that depend on the Rust bindings will break. And that is the Rust bindings problem, not the file system community at large problem. And that's going to be true for a long, long time. And I think we just simply need to accept that, right? Because the answer, you are not allowed to refactor to the C code because it would break five critical file systems that distros depend upon, is, like, not a starter. But that wasn't the answer.

Okay? So, like, we'll see. I suspect the best thing to do is you to continue maintaining your Rust bindings. Over time, there will be continued C code refactoring, right? Maybe we will start using, you know, K free RCU. If that breaks Rust, we will find out whether or not this concept of encoding huge amounts of semantics into the type system is a good thing or a bad thing. And instead of trying to convince us what is actually correct, let's see what happens in a year or two. And it will either work or it won't. And we will see, more likely, where does

the pain get allocated? Because with most of these sort of engineering things, it's almost always a pain allocation question. That's Ted Tissot, and you can hear the passion in his voice about this topic. He's worried about long-time maintainership here. And now the response to this clip, the individual you're going to hear responding to Ted is Winston Almeida, and they are the Microsoft engineer who publicly resigned from Rust for Linux maintainership this week. So keep that in mind if I'm right. I think I am. Keep that context in mind. This is before they resigned, when they respond to Ted. So Ted, let me try to respond to that. I'm not saying that C is going to be stuck in whatever

it has. That's not what I'm saying. Actually, what I'm trying to say is I want for you guys to tell me what the semantics are of these things that will reflect in Rust. That's all I'm saying. From my point of view, it's the same of any user, right? Users of these APIs, they have to know about the semantics

and they have to use them, right? And the rule is, if you make a change, a semantic change, you have to go and fix everybody else who you have broken, right? While Rust is a second-class citizen, I'm fine for you guys to say we don't care about this. You guys should look at it. So, here's the thing.

You're not going to force all of us to learn Rust. If I make a change, I will fix all of the C code because that's my responsibility. Because I don't know Rust, I'm not going to fix the Rust bindings. Sorry. Ted, let me try to say it more clearly. I'm not trying to force you or anybody else to do anything else, right? I'm sorry. Let me try to say it more clearly. I'm not trying to force you or anybody else to do anything else.

I'm just asking politely to see if you guys could give us semantic information. That's all I'm asking. You can hear Kent jumping in there a little bit, trying to just de-escalate a little bit. And you can hear Wedsen saying, I'm just trying to get the details so we can factor that in. Yeah, there was kind of a lot of that. And I think it's tricky to get the context right to, you know, when you're presenting a bunch of changes, like some of the talk ended up just trying to re-explain what the Rust folks were trying to do, as you said right

there. Like, I just, I'm trying to get examples of the context you all have so I can try encoding them in Rust and we can see what that looks like and if we like it. And then you end up with a bunch of side conversations around like the names of functions and how methods work in Rust. But as we heard there, there's the question of sort of who's responsible for doing what and how do we update these things. But there's also details in question around like, are you just copying kind of the way the C code does things? Do things have to match at like a low level? Or is it okay if they just perform the same high level function and the Rust version does something differently? Because, you know, when you do stuff in C, the way it often works with these kinds of type systems is if you work within them, you can,

you get to prove stuff, right? The compiler helps you prove stuff and show that your code doesn't have certain classes of bugs. But it means you have to buy into that system. And usually, depending on the implementation, there are classes of valid programs that won't type check correctly, but don't have bugs or don't have those bugs. You end up with some restrictions, like there's never a complete overlap and match. And so you, especially in places that didn't previously have those restrictions, there's some comparisons to maybe JavaScript and TypeScript, although TypeScript was designed to be a little more, you know, allow more stuff because it was on top versus Rust, which has a more strict compiler in many cases.

A lot of the patterns that you found in existing C code might not be, you know, you might need to do more complicated things on the Rust side to sort of fit that in. Or if you were designing it from scratch, you know, the Rust way would guide you to design it a different way. And so there's this question of like, do you copy the C way to what level? Or are we allowed to improve things as part of this work? This kind of comes down to, do we do refactorings or not? Are we going to do refactorings and cleanup and come up with better names? We have to do cleanups and refactoring. Otherwise...

That's Kent Overstreet there. And he's saying we have to do it. Otherwise, it's just an unmaintainable mess. Yeah, so you've got that, which that's more changes to be added. And then I think you kind of got a large, like Linus earlier we heard talking about, he seems to think in the long term, in the high level, we've got to accommodate for these things, whether it's Rust, maybe a better, you know,

an advanced version of C++, or another language comes along that has, you know, different affordances, different UX, but offers the same benefits of Rust. I think Linus is trying to look out in the big picture, and I'm not sure if that message has filtered into all of the kernel community in the same way, right? Like, there's going to be all of this low-level fighting. I think what we've also seen, like, that's just going to be, you know, this is new stuff trying to be pushed across a diverse set of maintainers and viewpoints. That's expected. My question is, are we going to see sort of buy-in on the high-level general idea of,

you know, these types of, are these tools useful, right? So, like, the Rust folks see it. Kent sees it, right? He thinks, like is this is helping me make more robust programs making programming easier for me a more solid file system yeah like there's some examples in that talk of you know like oh in trying to encode this stuff in the rust type system i learned stuff and he has some thanks for other developers out there like who taught him stuff about like file system internals that he didn't even know and he's

done a bunch of file system you know and so and it's like and then once you've encoded that that that lives there or another example might be something like reference counting reference counting is a basic version of garbage collection so maybe you've got some data structure in the kernel and you have multiple places in consumer code that wants to access that so for each place that has it they increment a little counter the the reference counter. And then when they're done with it, they decrement it, they subtract one, right? And that lets you say, you can see how many people are using it, but it also lets you say, oh, if the count is zero, I can clean up that memory. No one's using it anymore. That requires your code to follow that protocol, right? So you got to, when you take it, you need

to add one. And when you're done with it, you need to subtract one. Well, what happens if you forget to subtract one? That means the code that's going to try to clean that up is never going to be able to clean up that object. And okay, one data structure in the kernel is not a big deal, but you can see over time for a stable kernel, like those adds up and you get a memory leak. Or what about the reverse where you actually accidentally call decrement twice? So now you've subtracted two. So you end up with a situation where the garbage collector says, oh, there's zero references, but that was a math error. And there is still one reference where some piece of code out there thinks it can still use that data. And then you get something like use after free vulnerability.

And now you have a serious security issue. Right. a serious security issue. Right. And if you're willing to put in the work and you can make it work for you and buy into the Rust type system and the compiler and those tools, it can check that for you

and make sure that if you do increment, you always call decrement. If you forget, it'll tell you not to. And if you try to do it twice, it'll yell at you too. That just helps you catch that kind of stuff. So on one side, you have people that are seeing those things

and like, look, I want these tools in my work. It's going to make a better end product, we believe. Yeah. And then I think just in like in the TED quote there, still some open question of like, are these techniques that will work in the problem domains I'm working on not being sold on it? And then in some of the mailing list discussions, I think we also saw just various folks not sort of like, are we doing Rust? Is Rust happening or not? And it feels like maybe Linus is saying, and yeah, and some folks are still in the like, is it happening or not? So you don't, maybe you don't quite have as broad of a general high level sort of like, yes, we will support this effort, even if we have to disagree a bunch in the trenches. Yeah, yeah.

Well, you and I did a lot of digging around, a lot of sleuthing, and one name kept coming up in a lot of these threads, and that is our very own Neil, and he's there in the mumble room, so I want to bring him in for a moment. And Neil, I'm just curious what your thoughts are on all of this stuff as it's going down. Yeah, so I was at LSF, actually, while this whole thing was going on. You know, personally, I think Ted was a little bit out of line for being so terribly reactive. He was very passionate. Well, there was passion, sure,

but I think he was also being overly reactive to it. But I also kind of get it. So there's some missing context. So for those who don't know, a huge chunk of LSF is intentionally not recorded. It's intended for maintainers of the storage, memory management, and BPF groups to be able to have frank conversations with each other to figure out what they need to be doing, what they are doing, and what they need to be doing. Some of the unrecorded lightning talk type things and discussions that happened prior to that particular talk probably drove him and several others a little bit more sensitive than they should have been. Like there was at least a couple of sessions where,

like I brought up the, for example, I brought up a topic of like, can we have consistent dashboards for each of the file systems that do their own CICD? Because every file system does it differently. But the problem is that nobody knows what everyone is doing and what the state of things are to integrate them. So I propose this idea of having an FS Devel Next, where everybody merges into and merges from for integration purposes, and also having some kind of dashboard, or at least a reporting list or something where all of the stuff gets, you know, dumped into so people can see holistically what's going on in file systems. Because, you know, this

is one of those things that I also have to worry about because I deal with file system stuff all the time. It sort of took a wrong turn when a couple of folks in there started to just that instead of, you know, coming up with interfaces to make all the data presentable and useful to everyone, that everyone should migrate to a single CI system designed in a specific way from a specific developer, and that would solve all the problems. I'm intentionally being a little bit vague here, but this set off pretty much everyone and it turned a productive conversation into a shouting match. Uh, this happened a couple

of times, um, in topics that kind of were circling around the usage, uh, the, the, the file systems and the developers that were considering, you know, rust integration. There was also some conversations about rust, um, you know, in, in other some conversations about rust um you know in in other unrecorded sessions where concerns were brought up about the fact that the um kernel compatibility was so uh the the sorry compiler compatibility was so tight that effectively nobody could really develop for it there is a very important requirement that anything that is developed in the Linux kernel has to be not only useful and reliable, but it also has to be portable. And portable goes in multiple dimensions. And the important dimension that kind of goes

unsaid a lot of times is that it has to be able to work on older operating system platforms because Linux kernel has to be able to build and run on everything and everywhere. Especially Greg Cage has this mantra of you should be running the latest stable or long-term kernels. Well, that principle gets violated if you can't compile it.

There's lots of these little things that happened leading up to this particular session that I think set TED off a little bit more than probably should have. But also, I also think Ted is wrong because the principle in the Linux kernel is that when you break something, it is your responsibility to fix it everywhere. If you're not willing to do that part, then you should then don't make the change. So as soon as something starts depending on Rust infrastructure in storage, that's going to have to be a thing. Like, it's no different than, you know, whenever the documentation subsystems change, like they've switched tools over the years. And that means

everybody has to learn what the new tool is and all the new syntax and all that stuff. It doesn't really matter what type of thing it is. Everyone has to deal with it. That's like how the Linux kernel development even works at all. But, you know, they're not, a lot of them are just not prepared for that because they're used to not having to work with each other. because they're used to not having to work with each other. Also kind of more upfront and center is, isn't a very like the next or maybe this is 42, right?

Is shipping with the blue screen QR code system to feature that requires rust. Yeah. And I'm trying to actually get that feature turned on in the Linux kernel in Fedora. Like I'm trying to get rust enablement turned on in Fedora. And that's been, I've been working on this, I think since, I don't know, six, since the first patches that finally landed

to widen the compiler range, because the real problem up until, I think literally 6.11 was you were, we were using so many, we were using a, you know, vendored alloc from the standard library, and so we were bound to nightly features that were not stabilizing because the standard library doesn't stabilize ever.

And so the compiler ranges were so tight that it was difficult to build, and I experienced this over and over again as the Fedora Sahi kernel maintainer. And the change to making it so now that we have a supportable version range that is going up all the time, it's widening with each release, in theory, means that this is now supportable from a distribution perspective, or at least from a Fedora distribution perspective, it is supportable. I can't speak for other distributions. I actually suspect it's not ready yet for most distributions because it's still too tight for them but for fedora it's fine right the problem ironically right now is that it's still buggy and not working on all the

different compiler uh things all the different compiling modes that the fedora kernel is tested on uh and so far and this kind of comes back to the circles back to the other problem, which is people are really not interested in helping to make this stuff work. So there's a lot of negative energy around any sort of enablement that supports this infrastructure, any sort of work to make this infrastructure work better. It's what has made the Rust stuff drag on for so long. Working with the Asahi Linux team, I know you talk about it and other influencers talk about it a lot where Asahi Linux is like a big project,

but the Linux kernel maintainers think it's a toy. They think it's like at least one of the major kernel developers who's like a maintainer of like three, five ish subsystems or something like that called it a non reputable toy because it doesn't matter. It didn't matter, which is ridiculous because it has conformance. It is performant. It is operating in thousands of machines like I have the I have access to the data to prove that. And the only reason it's out of tree is nobody wants to review anything. And there's nothing any of us

coming into this project can do about it. Yeah, you do have to watch out that you don't, on the other side, you don't get to say like, well, Rust hasn't proven itself if you're also not looking at what Rust is doing or for any new technology. If you don't get to say like well rust hasn't proven itself if you're also not looking at what rust is doing or for any new technology you know like if you don't if you don't pay attention then you don't get to claim that it not you know your ignorance is a sign of it not being suitable

and then you you have people where their priority is is to just keep what they've got working and they don't really want to be bothered with it and they get emotional about it too i mean this is this is across it feels like this is a human problem, Neil. It feels like we see this... Of course it is. Yeah. Yeah, you've got people with different priorities that are contributing to the kernel.

Well, you know, the funny thing is that, like, this also sparked another conversation about, like, even for the C style, C side of the Linux kernel, should it actually stay C? Like, there was hallway track conversations about, you know, and there was... I think y'all saw it at the beginning of the year, there was a revival of the discussion of like, well, what about C++ for the C style stuff? Because with C++ 20 and newer, I mean, yes, it's not mandatory in the same way that Rust is, you can do some of these things, and it would be an ergonomic

improvement over some of the really goofy things that the Linux kernel code has to do. Linux kernel, so I should back up and say, operating system kernel development is not that special. What is special is that the Linux kernel, its flavor of C uses a lot of custom stuff that really looks like they're beating C++ features into C and not doing it that well. And that makes it really baroque and difficult for people to understand how to consume it. And like, I suspect this is actually the crux of the problem for being

able to have Rust interop inside of the kernel internals, because that Rust guy was asking for, hey, what are your semantics of your APIs? And like, give me a heads up so I can just refine the APIs based on it. I suspect given how bad C is as a programming language, no one actually knows. And no one that that can actually and no one wants to say it yeah and i suspect that's a bigger part of the problem and like i mean me personally right like i'm a polyglot programmer i've worked with lots of different programming languages but my go-tos are actually c++ and python i personally and again this is me speaking personally while i'm not super comfortable with rust i i have not successfully written a Rust program yet, though I am trying to learn it as are even more important, like drivers and stuff like this, moving things to a combination of C++ and Rust would be a stronger benefit than any of the other things that people are, you know, sort of proposing in what I feel sometimes is either in mixed faith or bad faith to try to stall modernization of the Linux kernel.

Yeah, boy. It feels like it's been, I guess, digging through this, I understand the technical arguments, but it does feel like overarching. There is definitely a bit of a culture pinch happening here. And what you've just made me realize is that everything's really up in the air. Maybe C won't always be the predominant language in the kernel one day,

but those things, if that's going to happen, it's going to be a long time. I will link in the show notes an article that I think is – well, actually, there's going to be a lot of good stuff to people go read on all of this. I feel, you know, really sad for the longtime maintainer that kind of burned out this week and resigned. You know, they've been at it for four years. This kind of stuff has just got to be great. And they said, you know, it was ultimately the non-technical discussions that burned them out there's just they always do yeah it's not the technical discussions that that usually make

people flame out it's the interpersonal conflicts like i can tell you that like it is there are days where i question why i'm doing any of this because I've been like verbally abused in some meaningful way for something I'm working on. Or I've been like some edge of cyber bullied or whatever, because someone doesn't you know, someone just wants to get their chops off on me or something like that. And it's just it sucks. Our friends at Fountain FM have just released version 1.1. That lightning-powered, value-for-value podcasting 2.0 app is now also a Nostrad client. What's cool about this is your identity can move across apps, including Fountain, and your social graph moves with you. So when you boost or comment on a podcast in Fountain, you're now sharing it to your Nostrad network.

And if you link to our show, it'll publish on the Nostra network for everyone else to see. Audio posts from Nostra clients now appear in Fountain as part of a curated feed of really good audio across the Nostra network. And zaps are now supported right there in Fountain for zapping posts. And you can now mention across Fountain and Noster networks, which also works across all the other Noster apps. It is truly the web of trust now being brought to one of the best

podcasting apps out there. So go grab Fountain at fountain.fm. Start listening to the show. Get more features for the podcast you do listen to. And you can boost

the creators you think deserve a little support. So go check it out. It's a great app. And version 1.1 just came out at fountain.fm. I was very fortunate this week to have a few of you gents join me here in Canada. We had our little Canada meetup, which happened in Toronto this Thursday at the Brewhouse by the lake, the Amsterdam Brewery, which was an amazing venue,

wasn't it? Oh, so good. And they were so accommodating. They had a beautiful, beautiful look, great lighting, great decorations, just a great vibe. And then they had upstairs, they had this balcony area that they completely gave to us. And so we were able to just fill it out and they dedicated three waitstaff to us. So nobody was waiting for anything. They were moving and bringing stuff constantly. And then if you wanted to step outside and get a little cold air, you know, get a little breeze, it was right there on the waterfront. Wow. And so just a beautiful view. You could see what their space needle knockoff, you know, it was great. You have to say a huge shout out to the venue as well

because two days before we kind of realized that our numbers had doubled within the last couple days and so instead of having 20 some people uh we had to book it for like something like 50 so they were great at saying oh yeah that's not a problem we'll just reorganize a couple things and you guys will have a nice time and uh sure enough we did yeah we did they were really i mean out of all the venues we've done i've never worked with a more like laid back happy to accommodate staff and you know brent while we were racking the server brent was calling up being like uh the number's gone up again i should have just gave them a link to our meetup page and have them just self-serve the numbers. Wire up like an ESP counter kind of thing.

Right. And as Brent does, he got to work getting some audio. We can't always fit all of the interviews in, but there's a couple of special ones I want to play. And this first one is Kiro, and it's spelled K-I-R-O. And he's special to us. K-I-R-O-F-M- us k-i-r-o fm 97.3 tacoma seattle your world hi my name is kiro i've been listening like a gb fan for probably over a decade now all started when i wanted to switch away from horrible Windows 8 to something a little bit more useful. And I found this YouTube video about an Arch Linux challenge.

And I'm like, this is exactly what I need. And I ended up just sitting in a terminal for a couple of days, installing Arch, and then doing it all over again the day after. Are you still running Arch? No, actually I am on Fedora right now because life gets in the way and sometimes you need something a little bit simpler. It's so true. And Kira, we've spent some time together through the website.

You helped quite a lot in getting the website up and running when we were first kind of asking the community to be a part of it. How was that experience for you? Well, that was kind of like a wild thing for me because like I had suddenly I learned about like an opportunity that I can kind of give back to the JB community. And one night I just kind of opened some script that I think Alex wrote originally to kind of scrape the JB notes. And I was like, let me take a look at this. And I started kind of modifying it to making it work and scrape the WordPress site and grab everything that needs to happen.

And I think it created a monster. Because I just kind of keep adding stuff without cleaning things and the code was horrible but it worked it did the job that was pretty fun and then I think I got to know you through the

community a little bit more and we kind of chatted I was helping her out doing extra additional things to the website keep a little bit of planning stuff and another thing that I was helping her out, doing extra additional things to the website, keep a little bit of planning stuff. And another thing that I was proud of, I think I was the one who put the Podverse embedded player into the site. That's pretty cool. And actually, I think it's Mitch.

Mitch or Mike? Mitch, I think it's Mitch. Yeah, I actually, after I reached out to him and helped out, did a little bit extra open source contribution there and did some kind of a drag and drop little feature thing on Podware's website as well. Nice. Well, I could tell you're kind of really proud of those. But also I want to say thanks on behalf of everyone for your contributions to JB, really, for giving back. That's huge.

And how was that experience of working with the community to get things done? It was a lot of fun. It was very rewarding. I was telling all my family members and friends, like, oh, look at me, I'm an open source developer now. Because it's always something that I wanted to do and get involved with people that you don't know

across all over the world. And you somehow find the common ground to do something you love. I don't know. It was a lot of fun. I don't even have a lot of words how to explain how much fun it was, but it was great. Amazing. Well, Kira, thanks for being here.

And I guess we'll see you at another meetup soon. Definitely. That, I think, captures the entire energy of the website crew who have helped for the last couple of years. Just built us a website that, A, matched so many of the hopes that who have helped for the last couple of years, just built us a website that a like match so many of the hopes that we had for that website, especially when it was just an idea. And we thought, Oh, how can we do this? And a bunch of people jumped in with ideas and even implementations of so some of those ideas. So just a huge shout out to everyone who's helped

with the website. I mean, that's chronicled quite a bit in office hours if you want to jump in there and see some of the back catalog. But yeah, thank you to Kira and everyone who's helped out. Given how it goes, they're probably making improvements right now. We're busy doing this. It's so great. And if we could, we'd thank them every single episode. You know, there's folks in there that are making sure

every episode gets published on the main site and that things build and things scrape properly. We're really, really grateful. The next conversation that Brent had, I also had a chance to talk with these two, and this one hit me right in the feels. This is Sean and Alex. We've got some folks here, and I would like to know where you came from, first of all, and introduce yourself. Excellent. My name is Sean Perkins. I'm a huge fan. I've been listening as long as my son here is old, which is he's 15.

Nice. And so where did you travel from? Upstate New York, Albany area. Very nice. And introduce yourself as well. My name's Alex Perkins. I'm Sean Perkins' son. Came from the same place, obviously. Cool. And do you dabble with computers at all? obviously. Cool. And do you dabble with computers at all? No, to be honest, I don't know as much about computers as I was just excited to meet the men who are an inspiration to the man who's an inspiration to me. Oh, that's sweet. So how does that make you feel? Gosh, when I heard him first say that, I was blown away. Still am. I am speechless. Thanks, son. So how's the meetup? You guys having a good time?

Great time. I actually got to get out of here to interview myself, all these beautiful people that are here. Yeah, it's such a great collection of folks, right? And we're what, like 50-some-plus people here? How is it hanging around all these kind of people? You having a good time? I'm having a great time. I'm with nerds. I'm with my people. Very nice.

Would you come back if we do this again? Absolutely. Absolutely. And if you get any further out east, I think it's going to be way more than 50 people. Sounds amazing. Well, thanks for being here, you two. Thanks for having us. Yes, definitely.

Thanks for having us. We might just have to do that. You know? New York meetup or something like that. Oh, yeah. Thanks for having us. We might just have to do that. New York meetup or something like that. That could be pretty great. We should just take a year and go from place to place. Go on the road. Do the show on the road. That'd be amazing.

We've already got a bus, or you do. True. Really great conversations. Almost everyone there talked to me about boosting or wanting to boost, but they had a common question. That was, how do I get SATats in Canada? Like, where's a fast, safe place to get sats? And the recommendation I got over and over again is BitcoinWell.com. BitcoinWell.com is a Canadian company, and I think they're even on Lightning.

And they support integrated self-custody features and all of that stuff that I think is really solid looking. So if you'd like to boost it and you need a source for Sats, it looks like BitcoinWell.com could be the way to go. Now, I got to take a moment and pause the show because what made all of this possible is the yeoman's hard work by Alex from Self Hosted. You know, he built this system for us. He did the software setup.

He did all of the hardware burn and testing. He shipped the thing up to Toronto to the data center. And then, of course, he got on a plane and made sure we got it all set up and working. So huge shout out to our buddy Alex. Really, really, really, really just super helpful. Not possible without him. And there's another individual we have to give a shout out to. And that's listener Steven. He was just a massive help. He's the one that owns the data center and helped us deploy the new rig. And he made it go so smooth on day one. Alex on his KTZ systems channel put out a video of us upgrading the fans and putting in the storage and racking and stacking. And it's just

a meticulous room. Absolute modern gear. He pre-provisioned the power, you know, the PDUs. He pre-provisioned the network for us, so we have our own VLANs for management and production. Killer. So we really just showed up, did the upgrades, and racked and stacked that sucker. So I grabbed Steven just so we can get a little background

during the meetup. Well, the MVP of this trip is here with me right now. It's Steven. And let's be real, Alex. This went smooth because of Stephen, right? 99% preparation. Yes. 1% actual effort on the day. Yeah. Thank you, Stephen, for everything you did.

You're very welcome. So background for the folks listening. Where is this data center? And why did you create it? So I suppose we're like north of Toronto. Don't want to get too specific. But effectively, I own an MSP, and I needed a place to put backups for my clients. Effectively, I had a little bit of extra room in the servers for a server rack. We got lucky. Yeah, you guys got lucky. I don't do this really for anyone else because I'm kind of power limited, but you guys are welcome as long as you want. Well, that's nice, and we appreciate it. And it went super smooth. We're thrilled.

It was one of the quickest things we did while we were here. We left the entire day for the install because you never know what's going to happen in these situations. And we were done by lunchtime. Yeah, that's a good job. Well done. Yeah. Now the work kind of begins, you know, now we've got to set up all the systems and migrate the data. I'm sure we'll be talking more about some of that in the future and definitely covering some of that on the self-hosted podcast. So be sure you're subscribed over there. Thank you to everybody who made it out to the meetup that traveled.

We had people that, you know, came from hours and miles away to people that were local that was there. It's so great. I'm so glad we had a chance to do it. And thank you, everybody. Booster Graham. Now we have some boost to get into, gentlemen. And our first one comes from Rotted Mood. 100,000 sats. Hey, rich lobster.

Thank you, Rotted Mood. Isis, here's a little bit to the vacation fund. Suggestions on what to air. How about a best of the members feed since not everyone gets to hear those? That's solid. I bet you we could make a full episode out of just clipping some stuff from the members feed that the public never gets. But the members heard it a while ago. Little trickier to go through because they're long and a lot of chapters and all of that.

You know, like some of those are like three hour suckers. Yeah, so much nonsense. So, but I do like that idea a lot. Because there's some stuff that's never aired to the general public. Extra content. So thank you, Rotted Mood. We're going to start stirring that one around. And thank you very much for being our baller.

Colorado coder Colin comes in with 80,902 satoshis. I hoard that which your kind covet. Long time listener, first time booster. Hey-o! Thank you. I managed to convince the day job to let me use Ubuntu, but it must be SOC 2 compliant, so I would love to hear any Linux security or hardening tips.

P.S. This is a zip code boost. Uh-oh, Wes. Uh-oh, you got your map? Oh, good, you do. I do. Okay, and under the hood is a zip code boost. Uh-oh, Wes. Uh-oh. You got your map? Oh, good. You do. I do. Okay.

And under the hood, this was two boosts, actually. With identical messages, one 500 sets, and then one, which I'm going to take as the, I'm going to hope that was a test boost. Yeah, like a test signal. Yeah. So I'm going with 80402. Raj. Okay.

Coordinating. Which would seem to be a postal code in Golden, Colorado. Ah, that makes sense. That would make sense based on the fact that he's a Colorado coder. Thank you, Colin. I really appreciate that boost. And tips for SOC 2 compliance. I don't know if I have SOC 2 specific compliance.

Yeah, I feel like, I mean, I haven't had to do that for a long time. But back in the day, you know, I mean, a lot of it ended up being sort of like you needed to come up with plans for how you were going to handle, you know, who has access to what and change control and updates. And then a lot of it is kind of just going back and forth with your auditor proving that you did the thing that you said you were going to do. And I think, I mean, you know, Ubuntu specifically has, you know, updates, support plans, extended support. You can integrate it with things like LDAP and Active Directory. So I think the pieces are there. It's probably going to be more of a question of like, you know, do all the players in

your org, are they willing to sign off on? Yep, that meets our standards or whatever. Right. But there's probably also some really good you know, use AppArmor or whatever like nitty gritty security tips for folks in the enterprise. I'd be curious if any of our boosters out there have ideas. Absolutely. Well, our buddy bear boosted in 25 000 sets put some macaroni and cheese on there too uh you're getting a little spicy in the pre-show lately chris

definitely time to take some time off here's a bit to support that vacation what did i say what did i say do you remember did i get spicy last week you didn't bring your soapbox but there's always something. Maybe we can look at the, maybe does the boost data have a timecode we can look at maybe for this one? I don't know if it was included. He says, P.S., it would be fun to post a pre-show or some of the clips as an episode. Show the non-members what they're missing. Ah, two for two.

Isn't that interesting? Yeah, see the members, they love the feed. All right, thank you, members, they love the feed. All right, thank you, Bear. And you're right. So we are making plans informally for September 22nd. I plan to actually announce it, but this episode got really big. So we're going to have like a best of, probably of the members stuff, or maybe something else, something in there on September 22nd.

Ooh, for Colorado Colin, Neil has a quick tip. Main thing for SOC 2 on Ubuntu is make sure you turn on security logging, which is suppressed by default. So there you go. Curious, he comes in with 24,690 cents. Pew, pew, pew! For the Nix question,

you won't please everyone. I was one of those that say, hey, can we talk about not Nix for once? And it seems like you moved the bar too far the other way now. I'll also put out my two cents. While I used Knicks, I also hated Knicks because the docs, the whole feeling of complexity for complexity's sake. We do hear that a lot.

I don't 100% agree with the sentiment, but we absolutely do hear that a lot. And it's a lot to take in when you first give it a look. That's for sure. Yeah. When I first started, I felt like there was context missing. Like I was missing the before and the after. Why are we going to? Why are we?

Maybe that's where the complexity for complexity. Yeah. You know, you're just like, why are we going through all these hoops? Why do I have to do it this weird way? Yeah. Yeah. Hey, did you see Nick's? Oh, wait, I thought we weren't doing it during the boost anymore.

Oh, was that a thing? All right. One last time. One last time. Yeah doing it during the boost anymore. Oh, was that a thing? All right, one last time. No, no, one last time. Yeah, this is the last one. Okay. And after that, we're not doing it during the boost anymore. All right, gentlemen.

Cheers. Enjoy your last one. He also adds up, for FidoO 2 keys, I have three. One is on my car keys, and one is in a safe in the house, and one is in a safe at an undisclosed location. Those keys are how I unlock my password manager. And with how secure that is, I don't think twice about storing pass keys and OTP keys in my password manager.

Because if you got into my password manager, I've got bigger problems. I like the undisclosed location. It's basically at my place. Oh, well, now it's disclosed. Oh, sorry. Whoa, Boosin with a granddaddy duck. Things are looking up for old McDuck. Dash capital D, NixOS, a.k.a. DistroFlag equals NixOS.

All right. All right. I love a distro that is tastefully themed and enjoy getting new ideas on software and setups, but I actually mostly just stick with NixOS these days. I was gifted a 2012 MacBook Air and initially set it up with Endeavor OS. Within

a day, I set things up with NixOS with Pantheon. My NixConfig just makes it simple to quickly set up a new machine. Oh, well, you know what? If you ever want to... Boah! If you ever want to send in a screenshot of your Pantheon setup on NixOS and, you know, just boost us a link to, like, an imager or something like that, I'd love to see that.

Good for you, too. I don't hear a lot of people using Pantheon outside of elementary OS, and I love it. Yeah, on a Mac, too. It could be a slick little setup. Yeah. It might feel right at home. Now, Chris, this one here from Kyle Potts might make you feel a little old,

but it's for 20,000 sets. I am programmed in multiple techniques. I've been listening since the Lunduk days when I was in middle school, and now I'm a 30-something-year-old developer.

Happy to see my favorite podcast grow with me, and happy to see I can contribute back with value to values. Cheers. What? Chris is having a reaction over here.

Oh, my God. I just had a reaction over here. Oh my God. I just had a heart attack. Oh my God. I'm old. Wow, Kyle. That's great. Thank you for hanging in there and jumping on pod verse and trying out the podcasting

to do the revolution. You're riding it right there on the leading edge with us, buddy. Nice to hear from you, man. I'm going to process that for a moment. Just wow. That's a lot. You for a moment just wow that's a lot you know i mean that's awesome that's awesome and they're still not sick of computers i know i know run while you

can you know thank you kyle so much i think is that your first boost too is that there you go thank you for setting it up well neil has a little note here for you. He says, I started listening to Lass when I was in middle school as well. It was the very first episode. Oh, my God. Brent, f*** Brent. Jeez. God, I'm trying to recover from the last one still. Oh, sorry.

Mount Bread came in live right now as we're going. I hoard that which your kind covet. With 37,217 sats. Zip code boost. If the crew ever finds their way close, I'd love to help organize a meetup. OS, that map's going to get some work. So it's 37217.

He's right to it. He also, while you're looking that up, he sent us a row of ducks. Just a live boost. He says, I'm patiently waiting for my cosmic pop OS over here. Patiently waiting in Nashville, Tennessee. Hey, Nashville would be great. Yeah, let's do a Nashville meetup.

No kidding. I would love to do it. Oh man, I'd love to do that. Perhaps, perhaps one day. Perhaps one day. All right. Also, Torped came in with 5,150 sats. That's not possible.

Nothing can do that. Torped did. Would anyone use an OS that does hand-smacking whenever you're trying to do something known to be insecure for your computer? For instance, it'd prevent you from passing text from the internet into your firewall config. What's the implementation mechanism? Like, is it an electric shock? Is there a mechanical hand? Oh, I like that.

Is it just, like, cursing me on the command line? You know, what about if it's just, like, vibrating haptic feedback in the keyboard? Like, it doesn't have to physically hurt you, but it could like buzz you. Oh, like a lane guidance kind of thing. Yeah, exactly. Yeah, like lane guidance. Like playing Operation. I like that you went to implementation of physical.

Mine went to like the software side. Like, how are you monitoring and determining what I'm doing? It's got to be like some sort of crazy LLM or something. I don't know. It just waits for you to turn off SELinux. I'd never do that. Our buddy listener, Jeff Boosin with 17 100 cents tough little shit little okay first up because jeff is a sweetheart here is a well-earned happy birthday drew happy birthday drew happy birthday happy true Indeed. Yeah, editor Drew went around the sun one more time.

And we're glad he did. Mm-hmm. Mm-hmm. Okay, and then on this next boost, Jeff points out something Apple TV reminder, which our buddy Hybrid Sarcasm also brought up, as did VT52, who says, hey, by the way, Chris, you've got some splaining to do about your Apple TV switch. Ah, yes. So on the live stream, I asked the audience to remind me. Because I've switched away from the Apple TV after being a very big proponent as the best set-top box out there for Plex, Jellyfin, YouTube, Netflix, you know, just the app experience on a TV.

Flex, Jellyfin, YouTube, Netflix, you know, just the app experience on a TV. And I still maintain it's absolutely one of the best, especially if you are in the iOS ecosystem already. But as I kind of transitioned, I felt kind of left out. I couldn't AirPlay. I couldn't use the remote features without getting some sort of hacky app. And I just don't like the inability for me to sideload my own applications. Like I've started thinking about maybe I want something that has sponsor block baked in. Maybe I do, maybe I don't, but I don't have an option on the Apple TV.

But Android TV stock, especially on the Nvidia Shield, has been perverted over the years by Google. And they've done things like put advertisements right there on the home screen. They're sometimes almost sexually explicit. You know, I've got young kids. They've also, you know, NVIDIA has done things like they've added Netflix buttons to the remote. I don't want the Netflix app to launch and, you know, people grab the remote and accidentally press that button all the time.

I want to be able to re-sign that key. I should be able to re-assign that key. Yeah, no kidding. I want that, you know, I want those features. Just a bad tasteless thing. Yeah, it's just no good. So I ended up stacking a few apps on an NVIDIA Shield that you can pick up on eBay for a pretty good price, and it's still a very, very good set-top box.

And the first thing I did is I got a file manager that's in the Play Store. You just want a file manager that will... Wait, you can have a file manager on a mobile device? Even a set-top TV box. Crazy, right? And this one actually has a built-in web browser. Some of them do, some of them don't. So you can download the APKs and install them once you allow unknown sources right there in the file manager. So not only is sideloading possible, it's actually kind of easy. Yeah. And so I use, especially with the NVIDIA app on the phone, you get keyboard on your phone, so you can do all the typing on your mobile device. Productivity Launcher is the first thing, or Projectivity Launcher, P-R-O-J-E-C-T-I-V-Y.

Projectivity Launcher by Spocky is the, I think, the absolutely necessary piece to make this actually doable for me because it replaces the launcher with something that is much cleaner, much simpler, much simpler, much more focused on just getting to your apps and playing back your media. You're doing a good job. And so Productivity Launcher is a big recommendation for me. And then I also installed the Aerial Views screensaver, which essentially rips off all the Apple TV screensavers and puts them on your Android. Let's, you know, choose a few options. Stacking all this together, I'm pretty happy with this

by installing Projectivity, by installing the File Manager, by installing Arial. I'm getting there. Sort of like a bespoke little Apple TV. Yeah. Only I could also sideload, say, the Home Assistant app, which I think I'm going to do this week, and then see if I can pull data and metrics

and do even more control over the device and do more automations. So thank you, everybody, who helped me. Remember that listener, Jeff, boosted with that hybrid sarcasm, like Wes said also, and VT52. So thank you very much. And I'm going to keep experimenting with it for a while. The family isn't loving it, but they're not hating it either.

And the Jellyfin app is way, way, way better on Android TV. So you get some benefits too. Not just a sideways move. Shout out to Yakhip who came in with a Spaceballs boost. That's Yakchip. So the combination is one, two, three, four, five. I like Yakhip, but Yakchip also works. They just said boost.

Gene Bean comes in with double ducks. This old duck still got it. Responding to last week's episode, the reality is Kent needs to follow the same rules as everybody else and submit things at the appropriate time. It does seem like with Linus's reminder, you know, a modified merge set was pulled last time

and this stuff just happened for this most recent RC. It does seem like they sorted things out since last episode and the stuff that needed to get upstream got upstreamed by Linus and things are moving forward. So that actually resolved pretty well. And, okay, here we go. We were talking about distributed storage in the pic last

week and you were kind of bemoaning like, ah, what should I do? I have a dream about an ideal storage system. So Gene Bean responds, hey Chris, how about OpenEBS with MinIO atop it for when you need block storage? Yeah, I could see that. Since OpenEBS is a Kubernetes thing, maybe you could deploy it atop

an immutable base such as Nix, or even better, Fedora Core OS. The latter is the default base for OKD, for what it's worth. Also calls me out, side note, this is a boost at Wes's request, which is true. I just figured you might miss it in Matrix. So boost is the way to get it in front of you.

This is the way. Especially this week when I was traveling, I am so far behind on Matrix. So thank you, Gene Bean. Open EBS and then Minnow gives you the S3 bucket like access on top of that. I think later in the boost too, we'll have another suggestion around distributed storage and Kubernetes. So

maybe there's K8s in your future. So Ham g came in with a row of ducks and a row of sticks fun will now commence hey i disagree with the 1.5 factor authentication even if you use another app like too fast you're still going to have your password database in some phone i think you have a separate db for two FA codes with a different password. It's secure enough. The advantage of TOTP is they are temporary. So even if the password is sniffed or key logged, the account is safe. Now they continue here saying, hey, a little follow up in the something you have, something you know model, you would have your password DBs and know two different secure

passphrases, both useless without the other. I want two keys I have to turn simultaneously. And then my password manager unlocks. Like a big old school dashboard you got in school. Yeah, and I can't, you can't, the same person can't turn the keys, so I always have to enlist the help of somebody else, and we have to turn at the same time.

Chris, you hate your current password management strategy. I'm sure this one would not... You don't think so? Well, I don't want to get a call at midnight saying, hey, you just want to check some messages. Brent, I need you to come turn the key. Brent! What am I going to do, Brent?

I need you to turn the key. No, you just teach Levi. Yeah, that's true. That might work. Oppie 1984 comes in with 4,000 sats. Analysis mode, password 80085. And he writes, so your discussion about a listener survey and possibly having listeners boost in, well, that got me thinking.

I had an idea, a self-hosted sat-powered survey. I set the details up at oppie1984.com slash satoshi-survey. I hope somebody can do something with this idea. I read through it. He's got a nice little write-up there. So, Oppie, we will link to that in the show notes, so the good folks can click on that and check it out at oppie1984.com and his Satoshi survey guide.

Sir Alex Gates boosts in with 10,000 sats. It's over 9,000! Podcasting 2.0 consultant. Via Breeze, if you were curious, I'm using NixOS with K3s for server apps. However, to consider it for work, I would need federal government vendor solutions.

That is something that Determinant Systems is very much focused on, is essentially making Nix work for folks that have compliance requirements. So there could be something to sniff around there. A little boost here from DD Smith, 2000 Satoshis. A suggestion for a horrible drink, if you can find it.

Beverly by Coke. I had it in a flight at the Coke store in Vegas. Amazingly bad. Oh, I've never even heard of this. So this is the evolution, for those of you following along at home, of our next drinking game. We're going to phase out the alcohol and we're going to just start drinking punishing drinks i suggested v8 as just a start it's not horrible but it's definitely not enjoyable while you're doing a show but this beverly cola well the first thing that comes up when i search for it is like old-timey ads from uh a hundred years ago

practically like yeah i bet that tastes not great yeah Yeah, it was marketed for the Italian market as a non-alcoholic aperitif. Introduced in 1969. Okay. All right. So not 100 years ago. Not quite. It was a good year, though. Yeah.

Moon Knight comes in with 2,500 sats. Coming in hot with the boost. I'm testing out Fountain's Noster integration, boys. Not sure what's supposed to happen, but I'm doing it. This is big. We'll talk about that at another time, but Fountain 1.1 is out. It's massive.

If you go back and listen to all your Silos are Busted episode, and now you'll understand why it's such a big deal that Fountain has shipped Nostra integration. Hybrid sarcasm boosts in with 12,345 stats. Make it so. The hell was that? Spaceball 1. They've gone to plaid.

Uh, Chris? Yep? Whatever happened to using IPFS for your distributed storage needs? Solid question, Hybrid. So, I view IPFS as a good public distribution mechanism for some things. And IPFS podcasting in particular is really compelling. We tested it pretty extensively with Office Hours and we distributed, oh gosh, I don't know, five or ten episodes on IPFS. And people that didn't use IPFS didn't even realize it because they have those bridges that just did an HTTP route.

So I think it's great for public-facing stuff. I don't think it's what you want for your internal LAN storage, your high-speed LAN data stuff, maybe stuff that your database needs. I just don't think it's right for that. I think it's more for external access calls,

in my opinion. You just don't want to put your DVD collection on IPFS. Well, somebody better pin it. That's all I'm saying. But thank you, Hybrid. Great question. Dude boosted in 5005 sets. Is that, you know, I think that's technically a Jar Jar boost.

Mr. Bustin with happiness seeing you second. For two factor, I use Vault Warden in Docker exposed to the web using VPS, Nginx, and Tailscale. And about every six months or so, I do a full export key pass to archive. Smart. Yeah, there you go. So, dude, I'm curious. When you say exposed to the web, do you mean, yeah, VPS? Why?

Why? Why not just have it all on your tailnet? Why have it available to the general web through a VPS and Nginx? Yeah, maybe dude needs to access it from places that you don't want to install the client because it seems like the security goes up by a noticeable factor if you just kept it contained to the tail net so i'm sure he has a reason i'd love to know dude and thank you for the boost hastep comes in with 2000 sats just a boost to suggest that brent should do a short imported american maple syrup for the nixo has drinking game. Oh!

He has to drink our bad maple syrup. You know, Vermont has some okay stuff. But that ain't Jemima's. No, I'm not doing that. Oh, come on. Not even for the next challenge? You tricked me once already. I don't trust you. We'll drink the good stuff and you can cross-cultural exchange. You know, it occurs to me maybe Jude's got

a situation where, you know, to log in to Tailscale, they're relying on the on that password manager. Oh, like, yeah, to get your tailnet off. Fair enough. Bun comes in with 5,000 cents. You're supposed! I've actually not really noticed

Secure Boot being much of an issue lately, at least personally. I used to have it always disabled so I could install Linux distros, but I guess Bun doesn't feel the need to do that anymore. Yeah, I actually haven't noticed it being a big issue myself. This was one of the questions we asked, and not a lot of people gave us answers, but I appreciate you taking the time, Bun, because I, too, you know, have not noticed this being a huge issue these days, but, of course, last week. I think the audience is just secure boots savvy. Mm-hmm. It could be.

Red5D sends a row of ducks. For a distributed file system, check out SeaweedFS. It's been around for a little while and has great documentation and many options for replication. Yes, that is absolutely on my list because my co-host on Coder Radio, Mr. Dominic, talks quite fondly about it. He's got it running in production for some of his clients, and I think he really likes Seaweed FS. So absolutely, I have not played with it yet, but Mike's got me thinking about it, and Red, I appreciate the cosign there. You can check out Coder Radio 582 if you want to hear more. We'll have it linked.

Thank you. Anonymous comes in with a Spaceballs boost, 12,345 sats. Yes, that's amazing. I've got the same combination on my luggage. No message, though, so we appreciate it. And moving right along, we got 3,333 sats from user 22. B-O-O-S-T. And he's just letting us know that he sent us a nice long email about using Nebula and

Hetzner for his VPS setup. So we'll dig that out once we get back to the mailbox. And Jin from Matik comes in with 2,000 sats right below the on-air cutoff. Or right above, I should say. I don't remember which show you were talking about storage solutions in Git, so I'm going to boost here. Did you know Keybase.io is enabling

a free encrypted Git repository, and they're also giving everyone 250 gigabytes of encrypted storage. Whoa. Just be patient when syncing a large chunk of data and files because, well, it takes more time with the encryption. You can read more in their documentation

at Keybase.io. Jen, I had no idea they were giving away 250 gigabytes. I think Keybase.io has been slept on by a lot of us. We've been sleeping on Keybase and I know they've changed hands now. I was a very early adopter.

I am still a big fan. They've kind of gone in a direction I'm not really interested in as far as their Slack stuff, but man, Keybase, I really wish it could have gotten a little bit bigger. I wish they could have been sustainable on their own. I want to live in the parallel world where everyone was just on Keybase, I really wish it could have gotten a little bit bigger. I wish they could have been sustainable on their own. I want to live in the parallel world where everyone was just on Keybase.

Yeah, yeah. That's a really nice deal, though. Thank you, Jim. Appreciate that. Distro stew boosts in with 5,000 cents. You're supposed! Wes touched on different attack vectors on password databases. Well, here's an idea.

If you don't want to be hosed when your vault gets hacked, consider peppering your passwords. Peppering my password? Yeah, you basically have a single word or phrase that you keep only in your head, in addition to your vault password. For critical passwords in your vault, like your bank, when you log in, autofill the stored password like normal, but before submitting it, manually append your other word. Ah, your pepper. Your little seasoning. That's right. In KeePass, I take these important accounts so I know I need to pepper when logging in. DistroStu. Solid idea. Dang it, I don't think we can hack DistroStu.

No, DistroStu is safe, that's for sure. Thank you. Appreciate that. VMAX boosted in 5000 sets. You supposed! I have the perfect distributed storage system that matches every requirement you mentioned. Wes even brought it up.

Just home install Longhorn. Honestly, it's nice once Cube is running. When I try out a Homelab app you guys recommend, I tweak a compose file to use PVCs and my cluster's tailscale operator as the ingress. Then Longhorn allocates

the redundant storage across my mismatched four thin clients across my house, my HL15 and my rack server. Okay, Vamax, I want to see your setup. I want to see your house. That's awesome. I love that you have an HL15. All of this is just my style.

Maybe Vamax will just let you run on their cluster. Yeah, really. Then you won't have to solve the distributed storage thing. Yeah, and I don't mind using a thing client. It takes me back. Wow, that's cool, Vamax. Any more details you want to share, please do. Spectrus came in with a row of ducks, and he says,

people have said it more elegantly, but keep the next content going. He also sent us a few other notes, which we appreciate. And GeekDude came in with 16,000 sets. How about that? That sounds fun. Fun will now commence. And Geek Dude writes, newish listener and a member. Right on. Thank you very much. We appreciate that.

The podcast was recommended to me on a call with people from Red Hat. And I have greatly enjoyed the content across all of the Jupiter Broadcasting shows. I want to send back some extra value for the Too Fast app recommendation. I'm happy to switch off of Authy. I originally used Authy for the desktop app, which they recently dropped support for.

I know, geek dude. He says, I'd like to find a way to authenticate without needing a mobile device, but the Too Fast browser extension is a good alternative. Well put. I agree, the desktop app was the nice thing about Authy, and then having it sync, of course, was really nice. but the too-fast browser extension is a good alternative. Well put. I agree, the desktop app was the nice thing about Authy, and then having it sync, of course, was really nice.

Thank you for taking the time to get the boosting going and getting all that figured out. I know that takes a few extra steps, and depending on the path you take, it can be extra challenging. I see you used Fountain, so hopefully it was pretty smooth for you. Appreciate that boost. Fomato boosts in with $14,118. Heck yeah! booth for you appreciate that boost famado busin with 14 118 cents heck yeah i'm still catching up from not listening to podcasts whilst on vacation but i was in your neck of the woods for several weeks there literally in a u.s national forest oh nice

this is a postal code boost that will get you to the right county at least just multiply the amount by seven oh man we have some beautiful places encoded yeah yeah you gotta do that should have a really the computer do that for you that's you know you don't have to put all the way he's got a whole pen and paper out well it goes with the map yeah i know but geez you don't have to go so analog and then my analog zip code database oh this zip code includes uh includes such areas as Leavenworth, Washington. Oh, I hope it was Leavenworth. That is such a unique experience in Washington State. They picked up a little Bavarian town and dropped it in the mountains.

And it's really something special. I hope that's where it was. Thank you for the boost. Our dear VT sent in a total of 6,666. I think that's three rows of ducks. Oh, you must be... You know what? There you go.

I think so. Hey, what do you think about a Seattle area meetup? East side, maybe? Feeling a bit jealous of my Spokane brethren. All right, we could consider it. We better do it before it gets too cold. Maybe... All right, we'll confab on that. I continue here. It's been a long time

since I liked GNOME. No hate for folks that like it, but it just doesn't feel very intuitive to me. Anyways, I came across the treatise that puts some of my GNOME discomfort into words. It perhaps is a bit harsh, but fair, I think. Yeah, He linked us to quite the takedown of Gnome. It was really something I read through it this morning. I was like, wow, this person started at the tour screen and then just went from there. Like it is a comprehensive thing.

Let me tell you what. Um, yeah. Fascinating. Thank you. VT 52. Appreciate that boost. I appreciate that link.

Montessorix comes in with 14,345 sets, and you know there's some space balls in there. Yes, that's amazing. I've got the same combination on my luggage. Now he writes, I quit my job two months ago and I'm starting my own IT business. Wow, congratulations. I couldn't resign myself to work for another employer forcing me to daily drive Windows. Fair enough. Yeah. Gosh. Besides desktop Linux, would love to hear also more about Linux content in the enterprise context. I'm still unclear. Is NixOS used at scale in production?

Yeah, absolutely. You just don't hear people going and shouting about it, right? You don't usually hear these companies. In fact, there are now businesses around NixOS that are trying to deliver that. We mentioned Determinate Systems. There are others as well. Obviously, it depends on what you set as your scale, but

there are folks depending on NixOS fleets and container workloads and all kinds of stuff. And government agencies that are quite large that are using it in different ways. I've been going through Graphene OS episodes. A quick tip, the key to flawless and fast seed vault backups is to configure NextCloud through the DAVX5 app, not the native

NextCloud integration. You know what? That's actually how we do it. That is how we do it. That DAVX5 app not the native next cloud integration you know what that's actually how we do it that is how we do it is that dat d-a-v-x-5 app killer killer and so i don't have to use like any google services or at all to sync my calendar my address book or any of that stuff it all syncs to my next cloud with dav x5 i really like it all right we got000 sats from user 22. It's over 9,000. I recently returned from Latvia and I was absolutely blown away with this new protocol. Truly believe we are seeing a shift to something new and amazing. Incredibly optimistic about the future. Yeah. Big sat adoption out there. It's interesting. Some of the countries out there, they're so far ahead of us because they have

the financial need. In the West, we have a lot of financial privilege. Our existing banking system works. It all kind of just works for us. We're born into the system. But when I traveled to El Salvador, what I learned is something like damn near 50% or higher, maybe it was even, I think it was even higher than that, of the folks there are just completely unbanked. They have no bank accounts at all. And so they're able to go from zero to 100 with the Lightning Network and Sats. It's really awesome. Autobrain came in with Spaceballs Boost.

So the culmination is 1, 2, 3, 4, 5. I had a great online co-op Steam session on Bazite last night. Smooth and slick experience. I'm still trying to understand what Bazite is. I listened to the episode on the, quote, distro type a few times. I'm still not quite getting it. Is there an easy way to define, quote, cloud native for the non-expert? And it's a, you know, a word used in many contexts that could probably mean many things, but,

you know, it grew up in the cloud era and it deploys tools that you would see there, things like CICD pipelines, things like OCI images and containers. And so for, you know, some of theBLU stuff, they're not using traditional tools that other distros might have used in the past or still use. They're using the same tools you'd use to deploy like a modern SaaS application with continuous delivery and API driven. Sorry to interrupt you, but if you think about it, Wes, would you agree with the statement that it's really talking about the tooling to compose the distribution, and it's a new type of user that can now compose those because they're using

tooling that is native to building cloud applications, which they already stand probably for their DevOps-type job or whatever job they have where they're deploying stuff inside containers and building systems up. They can now take that skill set of understanding how to layer containers and understanding things like that and use it to actually build a distribution. And you can involve more people. You can automate things. You can check things. It's just kind of a new approach. And it's taking what people have already kind of developed around containers and applying it to how you put together a holistic distribution. Yeah. If you're curious, you might just go check out the repos on GitHub because they've got,

you know, GitHub actions files to find. They've got container files. So you really can poke in a lot of areas. And I'm glad it worked well for you. Even if you don't know how it fully works, it's a good sign that it's working smoothly. But right before we go, I'm going to pull up just one from the thanks section because the guy Swan boosted in with 210 sats just to say being able to boost and comment on Fountain

as a post on Noster is just so cool. It is. It's going to lead to a whole new generation of discovery. Very excited about it. And the podcasts that aren't participating are not going to reach those new users. It's a wild time out there, man. Mount Bread rounds us up. Live boost.

Oh, we already read that one. But he says live boost, which I love. I just love the live boost. So we had a banger of a support and boost section. And I heard so many positive things from everybody at the Canadian meetup. And I heard so many positive things from everybody at the Canadian meetup. Boy, are we just ringing with appreciation for our audience and really feeling motivated to just keep doing this thing for years, keep making the best show possible for all of you, and really be held responsible only to our audience, which is massive in this era of content. When you look at all of the mainstream media out there, all of the social media platforms, the way the algorithm perverts YouTube,

the way none of the news can give it to you straight, and you can't trust any of their motivations. What we're building here is a truly open content delivery system over RSS where the community actively supports, interacts, and maintains it. It's open source content. It's not source code, but it's completely different than any other kind of media out there. And we had 72 unique participants this episode. Out of tens and tens and tens and tens of thousands of listeners, 72 people stepped up

and really helped us out. We had 37 of those that were just streaming us sats. We stacked 73,965 sats just by folks streaming while they listen. Thank you to our streamers. When you combine that with the folks that sent a boost in, it's absolutely bonkers. With all of their support, we raised 551,839 sats.

Fountain FM 1.1 just came out. The ecosystem's getting even better as Podverse is doing a rewrite to make the application more responsive and have better battery life. Castomatic has just been winning iOS users over one after another. There's so many apps out there. You can now participate at podcastapps.com. There's more features when you start using those apps, including the live stream in the app. Podping means you get the episode notification within 90 seconds of us releasing it.

There's transcripts, there's chapters, and a lot more, and it's all based on open standards. Thank you, everybody who participates. You have no idea how much it means, but I try the best I can to express it. I got two picks for us. I'm going to give Shotcut a shout out here because we haven't mentioned it and they just had a new release this week. Shotcut is my go-to video editor on Linux. There's a lot to choose from. Katie and Live's fantastic.

I don't think I knew this. I know. I realized it when they had the new release. I'm like, I don't think I ever talk about this. AvidMux is actually my go-to editor. Yeah. But Shotcut, like if I'm, say, going to put together like a little family video, Shotcut, and they just added subtitle support and a bunch of other stuff. Version 24.08 just came out and it was almost immediately updated on FlatHub. So if you have the Flatpak, you get it right away.

Many video and audio formats, essentially anything FFmpeg can do. It is OpenGL GPU accelerated. So the UI is nice and slick. You guys know I got a high standard for this kind of stuff. I love Shotcut. And I also wanted to give a mention to Butler. From our buddy, Cassidy James Blade, he's created Butler, which is a dedicated app that lets you access your home assistant dashboard from a native Linux UI. I think this is really slick. I generally just have this pinned in a tab in my browser, but he's brought it into its own app. And you get a native header bar. You save, when you open it, it saves and restores its position and its size. It has light and dark mode support.

And additionally, I could see, now he doesn't have this yet, but God, I'd love to see this. I'd love to see this. I'd love to see this. It looks minimal too. I mean, it's like a meg. Here's what I want. I want this so bad. I want love to see this. I'd love to see this. I'd love to see this. It looks minimal too. I mean, it's like a meg. Here's what I want. I want this so bad.

I want somebody like this app. I want this Butler app, but I also want it looking at any kind of sensor data. Like what are my disk size and how full are they? What's my CPU at? What's my temperatures at? I want all this data fed back

to my home assistant. The Mac users get that. I want it on Linux so bad. And so maybe Butler gets us there one day. Cassidy, I don't know if he's planning to take it there. I doubt it. I didn't ask him. But if anybody knows how to do that, please let me know.

Please let me know. I want to integrate my Linux box with Home Assistant as much as possible, especially like my ones at home. So if you have like, oh, my servers. Could you imagine if I could reboot my servers from the home assistant dashboard if i could see what the temperatures was and i could chart that and the usage and the load i would love that so if you know of anything that does that please boost in and tell me in the meantime butler at least gives us a nice dashboard companion to use the home assistant ui in linux now i don't know if this is maintained but apparently there's a net data sensor integration

for Home Assistant. Oh, geez, of course. That's one way I could go if I had to. I'd love another suggestion if anybody has it. Please do boost it and tell us that. Also, we'd love to hear your thoughts on any of the Rust coverage. And if we got too technical for you this episode.

We asked you before if you liked the technical stuff. You said yes. So we got technical this episode. I'd love to know what you think. Yeah, if you answer positively, next episode will just be a Rust tutorial. See you next week. Same bad time, same bad station.

We will be live for at least a couple more weeks until we take a break. You can join us at noon Pacific, 3pm Eastern over at jblive.tv. It's in your local time at jupiterbroadcasting.com slash calendar. Links to everything we talked about today, yep, that's at linuxunplugged.com slash 578.

Go check out Self Hosted. You'll hear a lot more about our new Colo server and the setup there and get to hear from Alex's perspective, who's really the expert on it. And then, of course, Coda Radio. You know, we talked about Seaweed FS in there. There's a lot of great episodes. Those are all listed at jupiterbroadcasting.com.

It's a whole network of shows. Big shout out to everybody who joined us at the Toronto meetup. We absolutely had a fantastic time and cannot wait to go back. Our server's there, so we very likely will, hopefully with a GPU in hand and lots of upgrades. Thanks so much for joining us on this week's episode of the Unplugged program. We'll see you back here next Sunday. Thank you. you