Down the Linux rabbit hole

Welcome to Changelog and Friends, your weekly talk show about the Linux rabbit hole. Big thank you to our friends and our partners at fly.io. Check them out, fly.io. Okay, let's enter the home lab. Well friends, Agentic Postgres is here, and it's from our friends over at Tiger Data. This is the very first database built for agents and is built to let you build faster. You know, a fun side note is 80% of Claude was built with AI. Over a year ago, 25% of Google's code was AI generated. It's safe to say that now it's probably close to 100%. Most people I talk to, most developers I talk to right now, almost all their code is being generated. That's a different world. Here's the deal. Agents are the new developers. They don't click, they don't scroll. They call, they retrieve, they parallelize. They plug in your infrastructure to places you need it to perform, but your database is probably still thinking about humans only because that's kind of where Postgres is at. Tiger Data's philosophy is that when your agents need to spin up sandboxes, run migrations, query huge volumes, a vector and text data, well, normal Postgres, it might choke, and so they fix that. Here's where we're at right now. Agentic Postgres delivers these three big leaps, native search and retrieval, instant zero copy forks, and MCP server plus your CLI plus a cool free tier. Now, if this is intriguing at all, head over to tigeredata.com, install the CLI, just three commands, spin up an Agentic Postgres service, and let your agents work at the speed they expect, not the speed of the old way. The new way, Agentic Postgres, it's built for agents, is designed to elevate your developer experience and build the next big thing. Again, go to tigeredata.com to learn more. Don't mind me, I got a little cough drop in my mouth there. I got this like nasal drip, so if I sound a little nasally, that's why. But I do have hot water, a tablespoon-ish of honey, and a little lemon, so.

Well, that's what you need. A spoonful of sugar helps the medicine go down, so they tell me.

Gosh, that's so good, so good. Alex, how are you, man? How's life? You good?

Mostly good.

Mostly good.

I mean, you see me here in my attic studio in Raleigh, and we've had the house on the market for three, four months now. We're trying to move back to England.

That's right, yeah.

But nobody's buying houses right now, so we're kind of living in this perfect show home of a house.

And you got it all ready to sell. It's always clean, right? It's always clean, so you can show it.

It's nice.

You're in the middle of a recording, you got to show me in an hour, got to end the recording and go, right? Is that how it works?

Yeah, that's it, yep, that's it.

London's calling, you got, is this on air material? A little bit of it or no?

Yeah, it's not as public knowledge. You cool with that? Yeah, yeah.

Okay, just in case.

Yeah, well, if we don't end up selling, if we don't end up selling, we'll stay. We'll stay, we'll stay. What choice do I have? If the place doesn't sell.

It is what it is, right? I hate that phrase, Alex. It is what it is.

It is what it is.

I feel like it's just such a give up moment, you know what I mean?

You know, many years ago when I was working at the Apple store, we used to have these little bets on the Genius Bar about how a conversation would go from the opening line. And if someone used the phrase, what it is is, we'd be like, uh oh, this is going to be a good one.

Tell me your story about how your photos don't sink to grandma's device or there's offloading happening or where are my apps?

I tell you, there are only so many times you can reset someone's iCloud password within the same 10 minute appointment without losing your sanity. Oh, yeah.

That's, how long ago was that for you? About a decade or more?

Yeah, a long time. 15 years? iPhone 4 era. So when was that? A long time ago.

2008?

Yeah, maybe.

I think so.

10, 11, 12 maybe. I don't know, something like that. A while.

Oh, I guess 2007 was the first one.

Yeah. I was there when the iPhone 4 launched. I remember we had, because it wasn't available. I think so. It was certainly a really good one.

I mean, the current version is an iteration of the iPhone 4, right? It's the same blocky. Really?

But I remember it wasn't available in many countries at that point. So what we ended up happening was, we'd get a bunch of what we called resellers queuing up outside the store first thing in the morning with thousands of pounds in cash in their pockets buying as many iPhones as we could sell them to ship them off to Dubai or Abu Dhabi or just, you know, it was all Arabic customers that we had purchasing these things for resale. It was kind of crazy, actually.

Yeah, and that was back in London?

Yeah, yeah, well, Manchester, but yeah.

Forgive my lack of geography understanding of the United Kingdom. I just, I don't know where things are. I've never been there to know where things are at, so.

I say London to Manchester like it's a big deal, but it's probably no further than Houston to Dallas in reality.

So like four or five hours kind of thing?

Yeah, it's pretty close really.

Well, if you said Dallas and I lived in Austin, I would say Austin, because it's Austin or Houston. So I get the correction, I get the correction. Interesting, interesting. Well, friends, you know, if any of that made it into the show, which I imagine some of it might, maybe all of it, who knows? This is my good friend, Alex. We've become friends over the years.

Hello.

Big fan, as you know, Alex, of your podcast that is no longer in place. Man, self-hosted is such a wild world these days, bro. Obviously, as you know, I was the last, I don't think the last time, but like one of the last times we talked, I was surprised by your creation of, what was it, Linux Server IO, is that right?

Yeah.

Is that the, is that what it was?

Yep.

Gosh, and we found that out, and you were the very first entry in there, because obviously you started it.

Yeah, it was my personal blog, and then morphed into something else at a later date.

And now it's like where the trusted images of all your favorite things is a home lab, or that's usually your source. Maybe if you're running Plex in a container, you're running that Plex container. I'm personally not. I'm using the official Plex Media Server.

Well, fun story with that one. Plex actually approached us back in the day to help them write that Docker container. So technically, you're just running a fork of the Linux Server container, technically speaking.

Technically, okay, a derivative.

Yes, a derivative is more accurate. Yeah, it was a fun project, because 10, 15 years ago, whenever it was now, getting old, there was just no standardization in the containerization space. People weren't writing, if they existed, they weren't writing coherent documentation at all, and there was no kind of standardized base image pattern or anything. And there is some weirdness in how the Linux Server images ended up architecting around S6 to have this init system inside the container, because obviously if PID1 dies, then the container dies with it. And some of the services need multiple things running inside the same container. So it's a bit of a weird thing, like trying to containerize some of these apps that were written before containers were really the native deployment format for server software in general now.

Yeah. How do you feel about the Podman Docker war? Do you feel like it's a war that's won? Are you a Podman person? What's your feels?

I'm not. You're not. You know, I'm ex-Red Hat, so I really should be pro-Podman, shouldn't I? But the thing that Podman misses for me is it's a very purist implementation. So it's extremely technically sound. What it misses for me, though, is some of the user spice on top, like that last 10% that Docker closed in terms of usability, because all the primitives for Docker pretty much were there before Docker came along. It was just the packaging of Docker with the standardized image format that really made it take off. The trouble with Podman, though, is when you want to do basic stuff, like even just mounting a volume, you've got to do this user ID shuffle, because I'm talking specifically about rootless Podman here. You've got to do this user ID shuffle, because the user IDs inside the container don't map to the IDs on the host properly, unless you do this shift. And so it ends up being this world of complicated UID script shuffling nonsense that I just haven't got time for.

I guess I'm not that much of a Docker user to know exactly what you're talking, I would definitely map UIDs to UIDs, but I never really hit the issue. If you mismatch on ownership, there's an issue, but I sort of leave it to the AI or the Stack Overflows or something else that's slightly above my pay grade.

Well, I'll break it down as best I can, just really simply. So in a rootless Podman container, you end up shifting the user IDs by roughly 100,000 or so, it depends on how you've got it configured. So on your host, you would probably have a user ID of something like 1,000.

Right, 1,000, yeah.

And inside the container, it would be the user ID 100,000 and 1,000. So like 101,000. So you've got to find a way to map those IDs from what the container sees to what the host can speak so that the file permissions for those bind mount volumes actually work. And that whole mess, the fact that I've just had to explain that to you is exactly my point.

Yes, I'm feeling your point.

With the daemon that Docker has running as root, yes, there are some security implications with that, but also comes simplicity for the user experience.

Right, this is the world of like potentially SE Linux, I think like you maybe you mentioned, I think does that play into it at all? Like this whole purist security Linux thing that I'm just sort of being exposed to. I've just now moved to Fedora 43, first time I've ever ran Fedora as a desktop even, I'm loving it, honestly, it's really cool. Traditionally, I've been an Ubuntu server, I've only really ever ran Linux as a server, never as a desktop. And so only in the last few months have I actually taken the plunge and began to run a system that's based on Linux. And honestly, Silicon Valley back there is running on Fedora 43, just so you know.

All right, good. Yeah, there's some other interesting stuff that Podman does as well, like they have this thing called the quadlets, don't know if you've heard of Podman quadlets, it's a bit of a silly name. Not yet. But essentially, they're systemd units that let you run your Podman containers as systemd units, instead of using something like Docker Compose to define the state of the world that that container would see.

Yeah, so you have a .service file for it instead.

Effectively, yeah. I mean, so, look, I know I give Podman a hard time, but it does do some stuff really well, so.

Well, I didn't ask you to talk about Podman at length, but I do, I'm really curious about this, so I think the one thing you mentioned and just glossed over, which may be super important, which is rootless versus root, right? Like that seems to be the divider there, which is the big reason why people think they want to use Podman, it's more secure because it's rootless versus root. That seems to cause problems based on what you just said.

Well, I come at this from the home lab angle, the pragmatist angle of, I just need my plexum to work.

It's me, I'm secure, right?

Exactly. Right. Whereas I think a lot of the motivation for rootless Podman and that whole movement in terms of boot C and a lot of the containerization changes that are happening is because of the enterprise, where that stuff does matter, where SELinux does matter, AppArmor does matter. But in my home lab, like, I don't care.

I don't need that stuff, yeah. Yeah, I mean, like, even if you had an API locally that you were doing something with, no auth, right? Why do you have auth? No auth, tiered auth systems. Like if I'm an external- I just get annoyed. Yeah, like there's no auth for an API I'm writing locally. That's just causing problems. I don't need to authenticate to my own API in my own home lab. I can trust me, okay? I'm the admin here. We need a shirt that says I'm the admin here.

Yeah, it's true. Yeah, look at me, I'm the admin here. If anything though, my adoption of Tailscale over the last few years has made that stance of mine even stronger because I have no ports like NADA opening my firewall anymore, nothing.

I only have Plex.

Oh yeah, maybe for the-

That's my only port open.

The advertisement of the server for friends.

Right, and we do external watching for my kids mainly. I mean, I do have to watch Silicon Valley in a pinch here and there, okay? So I might be, you know, on the subway or something like that, which we have no subways here. I'm just making a story up. I gotta get my-

This is America, we don't do subways.

You know, we don't have subways around here. No, the ground's too hard here in Austin to do a subby. Like, it's rock under there, okay? Oh yeah, it's true. I mean, I think Elon may be trying to find a way to make a hole, but we'll see.

My LAN is a trusted zone, completely.

And how crazy do you VLAN a lot then? I imagine you probably have an IoT VLAN at least, right?

I do, yeah. So like when some random piece of IoT junk arrives off Amazon, it's not getting on the same VLAN as my server. That's just not happening, but other than that.

It might be shipping software onto it that's got a rootkit or something like that that could be trying to infiltrate, you know, or like port stiff. You just don't know.

You just don't know. It could do that. It could just have an algorithm that crawls open Samba shares and lists the file directories.

Oh gosh, that's fearful.

I don't know. I mean, who knows what these things are doing or capable of? ESP devices over the years have gotten so capable that I watched a video just last night of a guy that built using one of these ESP 32s and a little display, like a need for speed kind of like mini map for his car. And he had like 2.5 million map tiles that he generated for the UK and like had petrol stations on it and restaurants. And it was really cool actually. So they're very capable devices is my point. So you just don't know what they're up to.

Gosh, do we want to go to Bootsy? I mean, do you have anything good to say about Bootsy? What's your thoughts on Bootsy?

I think it's a really interesting development. I don't know a huge amount about Bootsy technically, but in terms of where it fits in the space, I think it's a logical conclusion of where things should go next. You know, the immutable atomic OS.

Are you able to describe Bootsy so that folks can follow? I'll try, but I won't do it.

Yeah, well you go first and I'll look it up.

Okay, geez. What I know about Bootsy is out of Red Hat. It's part of like their world essentially. And they're setting this up. It's essentially a system that can change itself underneath. And it's, what's the right word? That you can't change the system. It's immutable. So what you have is what you have and you can update underneath your data, but that's the part I don't get. And I know that Steam Deck and some other things like these gaming things are doing on top of it is really cool. I just don't know what's cool about it. Aside from being able to build your own Linux distro for yourself, potentially. That's kind of where I've been dabbling at. And when I say dabbling, just a moment. And I got scared and I ran away.

Did you go to Red Hat Summit this year per chance?

I did not, but I did go to Texas Linux Fest.

Oh, it's the first one I haven't been to in maybe seven or eight years.

I know, I think I reached out last year and I was like, hey, I might be there and I didn't end up going. And me and a buddy went and thank you to Carl George from Red Hat for the invite and all the folks who are in that conference. It's an amazing conference. If you're even close to Texas during Texas Linux Fest, you're missing out if you don't go. Seriously, it was a blast. I was at two workshops, both amazing. It just is like one of those really feel good regional Linux conferences. So you get that regional feel where it's sort of small and intimate, but all the right people there to sort of still go deep into this world of Linux.

Did Carl share with you any of his pocket meat?

No, does he have pocket meat?

Carl George, I met Carl for the first time. It must've been 2017 or 18 at Texas. And he was like, hey, Alex, you know the way he does. Hey, Alex, would you like some of my pockets?

He's a very jolly person.

You're a complete stranger. Why are you offering me your meat? But it turns out he lives in San Antonio and there's a meat market near his house. And so he came out for all things open in October to Raleigh and I placed, literally placed an order with him to bring me vacuum packed packages of San Antonio meat market meat to Raleigh so that I could get my fix of Texas barbecue.

Yes, Texas barbecue. Yes, good old Texas barbecue is the best.

If I lived in Austin, I'd be the size of a house. I mean, it would be a problem, I think.

It would be a problem. It's a problem around here, okay? Texas is big and so are the people. People in Texas are big people. But not everybody, but some of them, pretty big.

I still need to make it out of something like Salt Lick or something like that. I mean, I've done the Franklin's, I've done the, what's it called, down by the river, Terry Black's barbecue.

Yeah.

I need to do some of the, you know, some of the others.

If you came here, I would happy take you. And we can take Carl too. He can come too. Carl, you're invited, Carl.

Be an imposition, but you'd find time.

Yeah, yeah, yeah. I would probably, so I've been wanting to take him, we've been meaning to meet Carl and I at Salt Lick, which I live about 10 minutes from the original Salt Lick. So the one that began it all for Salt Lick. And it's like going to somebody's backyard ranch in their house. It's so cool that it's not even like restaurant fields. It's like somebody stood up a little barbecue in their house and it just started to morph over time.

Well, that's probably exactly what it was.

Well, that's exactly what it was. Yeah, it was like, hey, let's make some barbecue. Oh, people come and eat it. Oh, cool, let's keep doing this. And there you go. I like, there's nothing wrong with Salt Lick. They're good. It's not my favorite though, but it's still amazing. So you can't go wrong. I would take you to The Switch, which is near here. I think it's under known and it's just so awesome, man. They've got a really good platter. They've got an amazing chopped brisket, baked potato. I mean, just like a little baked potato with chopped brisket in it. I mean, come on, is your mouth watering?

I'm hungry now. We're not far off lunchtime for me.

The Switch, I would take you there.

Yeah, that's quite, I'm just looking at it on a map. That's quite a ways outside of the city limits too.

It is, but it's really close for me. So I live about 10 minutes west of The Switch.

Nice.

So I live in that area there. Okay, Texas Barbecue, Texas Linux Fest.

Well, we were talking about Boot C.

We were talking about Boot C briefly. Bruteless, well, we don't have to go into Boot C. I'm more, you know what I'm interested in? I'm interested in number. This number. Oh, right. Tell me if this number matters to you. Five, zero, comma, zero, zero, zero. What does that mean to you?

50,000, yes.

50,000, right? Congratulations.

I think I know where you're going. Thank you. Yeah, I think I know where you're going with this, but.

That's a feat. I mean, you've grown the Tailskill channel on YouTube from, I don't know what number, but 50,000 is nothing to shake a stick at,

and that's awesome. It was about 1,600 subscribers when I took over, yeah. It's been a lot of work.

Does that make you happy? Does it make you sad? Do you feel content, discontent? What are your feels right now when I say the number five, zero, comma, zero, zero, zero?

You know, to some degree, the subscriber count itself doesn't matter that much because you can put out a video that gets 2,000 views and you can put out a video that gets 50,000 views or 100,000 or whatever, and there's just no connection between the number of subscribers to the number of views you're gonna get, but it is a nice vanity metric, and it is a nice thing to turn around to management and say, look, we've got big number. It now says 50,000, and it's a nice signal, and I really appreciate everyone that has subscribed to the channel, of course. It wouldn't, I mean, I wouldn't be able to put a roof over my head without people watching, so it's quite humbling in a way, but it also in some ways doesn't matter that much. Like, maybe understand what I'm trying to say about it.

In the grand scheme of things, right?

Yeah. I also kind of just feel like we're just getting started. I've just taken over as the head of developer relations for Tailscale about two weeks ago now, so we're putting together a plan to build out a team to support me on the back end, because up until today, it's been just me, so I've been editing, scripting, filming, like doing just everything.

It's a lot of work, huh? It's too much.

It's too much. Yeah, it's too much. It's too much. It is too much. So we're putting together a plan, and there'll be some job postings goes up in the new year, so keep an eye on the Tailscale Careers site, little pluggy-plug. If you want to work with my team, come do some video editing, come do some DevRel stuff. It'll be a good time.

Yeah. Well, I've been following, and actually when I saw that number, I was like, let me go check it out, and I think we're at like 52,000 now, so I think since that post, a few more thousand have subscribed, and then I saw a button there that said Subscribe, and I'm like, gosh, Adam, you haven't subscribed to this yet, which is so weird, because that's how the algorithm works, which is probably why you say it doesn't matter if you have 50,000 subscribers or not, because I watch your stuff, but I'm not a subscriber. Yeah. And that's a weird thing. It's funny. I kind of like that. I like as a consumer not having to subscribe to everything.

Well, because I genuinely think that good content, the people, there's a bunch of tropes on YouTube, and the people that rely on the like, comment, subscribe, smash that like button, all those kind of tropes that exist on YouTube, if the content's good enough, you don't need to say any of that stuff. People watch people. They don't watch, they're not watching the Tailscale channel. They're just watching me and what I'm doing. It's a really, like you look at Linus and how he's had to grow his channel sort of beyond the cult of just Linus, and obviously he's way down the road from where I am, but it's really interesting that he'll just make random cameos in videos that are hosted ostensibly by someone else on his team, and then he'll just helicopter in for five minutes and say a couple of pieces to camera and then go away again. And you know that he hasn't done anything like post-production or scripting, or really been involved in the video at all, but they're really struggling to enter that post-PC era, that post-Linus era. Hopefully, Tailscale won't suffer from that problem because I don't really chase the fame aspect of it, or I don't want to be famous or anything like that, but I do want people to be as excited about stuff that I am as I am, and that's really my driving goal is to just, like this week, for example, I set up a self-hosted image instance, you know, the Google Photos replacement.

Yes, tell me about this.

Done by Alex and his team at Futo and Louis Rossman in Austin, no less, so they're on your doorstep, dude. So I set this up and I was like, I can delete Google Photos now. And I finally did it. I've actually finally deleted my quarter of a million pictures in Google Photos. It took me all weekend. You have to go through the web UI and select a maximum of like 10,000 images at a time. It's-

Really, one million for real?

Yeah, it was like quarter of a million pictures. Yeah, it was quite a lot. Wow. But Google make it really difficult to leave. But anyway, I'm off the Google source, finally. It's about time the self-hosted guy actually did that. But I was like, well, I bought my family for Christmas when my daughter was born five years ago, some of those little Google picture frames. And so we just had this shared album where I'd just drop photos in every few months and then it would update their picture frames remotely. And I'm like, oh, oops, I guess I've just broken. So I've just broken that workflow for them.

So I just came in- Just a favorites, basically.

Effectively, yeah. And so I came across this project called Image Frame. And this lets you turn any device, because it's just a web view. So you can run it on Android iOS and Apple TV and Android TV, just a web browser even. And it lets you turn any of those devices into essentially a Google Photos frame. So I can just drop pictures into an image album on my self-hosted server, connect this Android tablet that I'm going to give to my family for Christmas, back over Tailscale to my house. And the same functionality exists, but it's all completely self-hosted and completely- I own the whole shebang. That's what I love about this stuff. Getting people excited about that real world use case. Yeah, Tailscale's cool. Image is cool and all that. People don't wake up in the morning and go, you know what I'm going to do? I'm going to sign up for a new VPN. And it's like the now what problem? How do I actually use this stuff? Make it useful.

That's really what- Right. You know, I think my usage of Tailscale is pretty simple. I think I just connect my services and I don't do much really. I'm not a deep Tailscale user and I kind of feel bad. So I'm actually thinking about having Claude help me go deeper because I'm just like, I don't know how else I can use it personally. Aside from the obvious connected tissue of my home lab, that's really just it. And it's really just mainly, kind of show my colors here a little bit, but I think for me, like I will SSH into Cineplex, which is what I call my Plex machine. And so I just type SSH Cineplex because that's what it's named in Tailscale. It's the name of the machine, right? It's the host name. That's really, sadly, potentially, maybe you'll be sad about this. That's my use of Tailscale. It's just easier host name SSH around my home lab.

It's great. It's fun.

And the occasional external view of something of a service running. I don't have a lot of, I could do better with that. I got other problems that I'm dealing with in my home lab, which is just writing a few more pieces of fun software for me.

You know, home lab.

Well, friends, I'm here with a good friend of mine, again, Kyle Galbraith, co-founder and CEO of Depo.dev. Kyle, we are in an era of disruption, right? I would also describe it as rethinking what we thought was true. And I guess that's kind of the definition of disruption, but from your perspective, how are teams, reliability teams, CICD, pipeline teams, how are they all rethinking things and where does Depo fit into that?

In the conversations that I have with customers, a lot of DevOps teams, platform teams, site reliability teams, they're really looking at this new era of software engineering

that we're all living in. And they're starting to question, like, the bottleneck is no longer the act of writing code. The bottleneck is shifting. The most time consuming part is integrating the code. It's everything that comes after. It's the build, it's the pull request review, it's the deployment, it's the getting it into production.

Once it's in productions, it's scaling up support teams to support it, it's adding documentation, all of these downstream problems. And so through the lens of Depo, what we're really starting to think about is there's a very realistic possibility that

within the next two to three years, maybe even sooner, that we're going to enter a world where an engineering team of three people could theoretically have the velocity of an engineering team of 300 people. And what's the consequences of that? What's the consequences of the code velocity spiking up to that level with such a small team? There's no way three engineers are going to be able to code review all of the code that's being created if there's three engineers and 297 agents also creating features and fixing bugs. So that's just like from a pull request perspective. But then you think about it through a build lens too of if your builds take 20 minutes with three humans, and now you're going to have three humans and 297 agents also running, well, like you definitely don't want your builds taking 20 minutes because now like the entire pinch point is the build pipeline.

And so we're starting to think a lot about how do we eliminate the bottlenecks that come downstream and what can we do with Depo that streamlines that?

So obviously friends, we are in an era of disruption. Things are changing, you know it, I know it. That's how it is. And the thing with production and what Kyle's talking about here is how in the world do you get your builds to be faster? How do you get them to be more reliable, faster, more observability around those deployments? You need it, it's required and Depo is there to help you. So a good first step is to go to depo.dev, get faster, try the trial, it's too easy. Again, depo.dev is where to go. It all begins at depo.dev.

So how'd you do your backups, Adam?

Gosh, I don't have backups, Alex. I got raid, okay, I'm just kidding. I really don't have offsite backups, okay? I'm not practicing three, two, one. I'm in a situation, it could go down right in this moment. It could catch on fire over there. And I will have to say, you know, get the thing, get the extinguisher and put it out and my stuff is gone. So I don't have backups.

Oh, well, there goes one of my pitches for you would be an offsite backup.

Well, that could be a pitch though, I need an offsite backup. So what would you do? How would you do it?

Well, I have a friend in Canada, so I'll tell you how I do it. And so I have a ZFS array in the basement, which just got, I think, five 20 terabyte hard drives in it. And then I use, it's a program called Z REPL, Z-R-E-P-L for Z replicate, I guess. And it essentially mirrors all of the ZFS blocks to Canada, which is where my friend lives. And he's got a similar sized server. He actually runs an MSP. So he just carved out a few use of rack space for me to use there. And so I have, he's a very nice friend. I'm sure he would offer you some rack space if you paid him a few Canadian rubles every month. And I just use the REPL over Tailscale because it's an SSH based protocol for the replication.

Could you do the same thing with Hetzner by any chance? Like, I don't know enough about Hetzner. I know you could do a lot of cool things with production machines. I know you can do some pretty decent size builds because you can choose your CPU and stuff like that. But do they have like that level of like ZFS, ZFS to me? Cause we're different.

My brain is so confused about that letter because it's in my name, Alex KTZ, Alex KTZ. People always give me a hard time about that one. And I'm like, ah.

I'm not giving you that time. I'm just being sure that I say ZFS, that's it.

I just pick whatever.

Could you do that with Hetzner though? Could you do that with like, I don't have the friend. I wouldn't mind the friend, but like for those who don't have the friend, could you, how could you do the same thing?

There's a couple of ways. I mean, Hetz, the trouble with any cloud provider is that once you get over a few hundred gigs of storage, the cost becomes prohibitive pretty fast. So there's a service called, and I'm not affiliated with them in any way, but there's a service called zfs.rent. Go take a look. What you do is you send them a hard, you send them a hard drive and they put it into a slot in one of their servers, I think in California, but don't quote me on that. And you pay them 10 bucks a month per slot that you occupy. So if you want 20 terabytes of offsite backup, you send them a 20 terabyte hard drive. It can be direct shipped from Amazon or Best Buy or wherever. They'll rack the drive for you. And then they'll give you an IPv4 address and a VM. And then you use that as your zfs replication target offsite for 10 bucks a month, job done.

That is good. That's a good find. I should get them on the show. This is a really, like, I wish I thought about this. zfs.rent is the coolest domain ever. And then, you know, any notable person with RAID in their house has gotta be running zfs. And if it's, I mean, if it's a butterfess, sure. Okay. Btrfs.

There just aren't many alternatives. I mean, I know butter gets a bit of a hard time, perhaps unfairly, but zfs at this point, really, for me, there's nothing that comes close in terms of the reliability, the development trajectory that it has with the IX system support and Clara systems like with Alan Jude in Canada as well. TrueNAS, of course, has been industry standard for sort of normal people for a long time. I say, you know, non-enterprise deployments of like iSCSI storage and anything you want to do for a small, medium business, whatever. Like, TrueNAS is the answer. I just can't think of anything better. Like, bcashfs is kind of promising, but then there was a whole drama recently with Kent Overstreet and Linus Torvalds in the Linux kernel mailing lists and development and trying to sneak some weirdness there, so.

I'm unfamiliar.

Oh, go look it up. It's a whole tranche of stuff, drama.

Yeah. Oh, gosh. Just what we need is more drama.

Well, bcash got accepted into the Linux kernel, which you would think would be an amazing thing, right? We've now got another file system native in the Linux kernel, because this is one of the downsides of zfs, is the licensing drama. You familiar with all that? Yeah.

Loosely. And I know about the origination in Sun and the accidental open-sourcing. I've talked to Matt Ahrens before. We have a good friend at Oxide, Brian Cantrill. So I kind of have some of the backstory of the Sundays and I guess not Sundays, like the day of the week, but S-U-N space days, capital S-U-N. So I got a little bit of history. Some microsystems. Yeah, some microsystems. And it's got an interesting history. And I think even the accidental open-sourcing of it was really wild. And then now you have open zfs. I don't know. And Alan Jew came on the podcast and talked about, and really got me wanting to become a BSD nerd. And I tried for a little bit and I just did enough to be familiar, but just was like, okay, Linux is really for me. I just don't need what BSD has. And it felt like an uphill battle for me personally, but I get it. I get it. But the licensing thing is really where it's at.

Mm-hmm.

The issue at least.

The trouble with BSD is I learned Linux first. So all my muscle memory is Linux.

Yeah, same. That was me.

And just stuff like the way that grep works or sed works, or just basic command line utilities like that. It's just different enough in BSD land that I'm like, I'll just go back to what I know.

Yeah.

There's a lot to like about the BSD universe. There really is. But it's just never stuck for me.

Yeah. So offsite backups, zfs.ran. Unaffiliated, I don't have any experience with it. So if you check it out, yo, your mileage may vary. Report back, of course.

The other option you've got, it's just a Raspberry Pi with a USB hard drive and shove it under the stairs at your mom's house or something, you know?

So how does that work then? I mean, ZFS is so RAM intensive, but I guess for a backup, you don't need a lot of RAM, right? Precisely. Is that the idea? It's just a backup so the arc can be small.

You are not going to be limited by the performance of the Raspberry Pi. You're going to be limited by the performance, presumably, of your internet pipe.

Right.

That's gonna be the bottleneck. And even if you do end up being limited by some local thing on the Raspberry Pi, it's a backup. Does it matter? It's gonna run overnight, every night, for the next five, 10, 20 years, whatever. So if it takes an hour or it takes six hours, who cares?

Right? I have to do this. Like, you're, okay, this is my, you've given me a Christmas present. Thank you, Alex.

This is my new- This is what motivates me, dude. It's just like solving real problems for people. So I'm glad to hear that.

I like that. Okay. So image is what you stood up. You're off of Google Images, if I understand correctly. There's no favorites list. Like, I try to presume in Google anymore, you've gotten a whole new hosted system that is image all the way, and your family is still happy with their Google Frames?

The family, one of the family texted me this morning and said, my photo frames stopped working. And I'm like, oops. I might have pulled the trigger a bit too soon on the deletion, but I just wanted to stop paying Google eight bucks a month for the two terabytes, you know? So.

Yeah, right.

How many drives do you buy? Well, for eight bucks a month, amortized over five years, I can buy probably two 20 terabyte drives.

Yeah, it's not the end of the world. I mean, there is some value there that you don't have to manage it, but now you get your friend. You get your friend in Canada?

Well, I'm also not training Gemini's future models with my images and the images of my daughter. And there's a huge digital sovereignty aspect to my motivations for getting off the big tech train when it comes to really deeply personal data like photos and documents. And I mean, I'm sure many of us use Gmail even, and I've been a Gmail user since I guess, 2004, five era, like when it was cool and a beta product. And I look back at the original emails and I'm like, oh my God, every email I've had in the last 22 years, they know what it is. It's a lot of data. It's a big data set, so.

You get paid to run a home app though. Would you say that it's fair to say you get paid to run a home lab for the most part or play with a home lab? When you get paid to make videos.

The phrase, if you do what you love, you never work another day in your life is very true for me. I'm extremely privileged and lucky to get to do what I do. Like Thanksgiving weekend was just a couple of weeks ago as we record. And I decided that I was gonna learn Kubernetes over the weekend.

Oh, how'd that go?

And I spent the whole weekend glued to my laptop and I had a great time. And I've now got a three node Kubernetes cluster in a closet and it works great. And then I was able to turn that into a piece of content for work. So, for me, I'm just very privileged. Like this is what I'd be doing anyway. So the fact that I get paid to do it is like just a huge bonus for me.

Yeah. Can we maybe touch quickly on, I guess as long as you'd like to actually, I said quickly, but however, these are the fields that I think get exposed through vibe based podcasting, let's just say, right? I got this vibe that going to Tailscale from being a podcaster, I don't know what else you did then. So I'm not gonna diminish what you were doing. I don't know. Like I know you were running the podcast then you got the job at Tailscale. What were you doing before? Were you consulting or like, how was that? What I'm trying to get to is like, how was it to go from like podcaster that I knew you have at least, and then now you work for Tailscale to produce content there?

I was a Red Hat consultant in Europe for a year or two. And then I moved to the States and I got a job as a TAM, sort of managing OpenShift accounts and things. It was fine. It was pretty boring. As long as the customers were happy.

What's a TAM?

A technical account manager, I'm sorry. So yeah, I basically have, I don't know, half a dozen accounts assigned to me and I would have weekly calls with those customers, big spend customers, and make sure that they were getting value for money and that their support tickets were being slipped through the queue correctly. And if there was anything new coming with it, like I was there during the OpenShift 3 to OpenShift 4 transition, which was a big shift from a RHEL base to an atomic immutable base OS image and re-architecture of like around Core OS and Project Atomic and all that stuff for Red Hat. Just helping customers navigate the complexities of all that stuff and just telling them what they need to know ahead of time. So there's a lot of skills in that that actually translate pretty well to video. There's also a lot of skills that translate from being on the Genius Bar back in the day to being personable and being on video and figuring out what someone's problem is and then presenting it to them in a way that they can understand. So it's just where I am now is like the summation of an entire 20-year journey, really, just failing upwards.

Failing upwards. Talk to me about the personal feels, though, of getting to go from, I guess, seems like interesting work, but maybe not so fulfilling based on what you're saying. You were running self-hosted at the time, I believe, right? Yeah, yeah. The podcast. And that podcast has since ended. And then you go to TSL scale, like being able to go and create content. Like, hey, I used to do this stuff for free, you know? Marginally for free. Sure, you probably had sponsors to maybe cover some rent or some things like that and some pieces, but not the full paycheck, necessarily, to get to do it as a job.

Well, you've got two routes to go as a content creator, don't you? You've got one, which is to do it the really hard way and do it yourself and bootstrap your own small business and interface with all these different companies and manage sponsorships and negotiate every cent that they give you. And then you've also got to be a creator and an editor. And you've got to be five to 10 different roles in one. I chose to go an equally difficult but totally different path of trying to do that, but within a corporate organization, a startup organization. I've been really lucky in so much as that the Tailscale channel doesn't have to pay for itself in the traditional way that, say, any other content creator has to pay for itself, because it's a brand awareness play, ostensibly. That's the purpose of the YouTube channel is just to say, hey, look at Tailscale. Look at the cool stuff you can do with Tailscale. And that's it. We're not reviewing products. We're not taking paid placement sponsorships. We're not doing any of that other stuff, which I kind of find a little bit dirty, to be honest. I don't ever really enjoy doing that stuff. It's a necessary evil, I understand. Everybody's got to eat. But the attention that you're selling as a content creator, you're monetizing that attention for, well, money, I suppose. So yeah, I mean, it's nice to have insurance. It's nice to have a paycheck and all that kind of stuff, doing it the Tailscale way. I don't know where it will end up, though. I mean, I really love what I do. I mean, the fact I get a paycheck reliably every month is fantastic. But there's that phrase in, of course, you've got Mike Judge playing behind you on Silicon Valley. You remember Office Space, right, where he has the interview with the Bobs?

Yes.

If Initec ships a few extra units, I don't see an extra dime. Where's the motivation?

Right.

And of course, my motivation is I get a paycheck every month. But if the channel goes from 1,600 subscribers to 53,000 subscribers, I still get paid the same. I suppose the validation is that I can.

You could negotiate, though, right? Well, I suppose. You could hinge some hallmarks and some milestones. But you get to hire, too, so you get to make some friends.

Yeah, yeah. I suppose the validation is the promotion, is the fact that I'm still employed, is lots of different positive signals. But it's just these are the sorts of things that you think about in the shower. And just at like 2 AM, your mind starts going. You're like, well, how can I maximize just everything? How can I maximize it for Tailscale? How can I maximize it for Alex? How can I make sure that the audience is getting the most out of it, too? Because ultimately, if the audience is happy, that will drive everything else down the road. So really, my focus isn't on me. It's not on Tailscale. It's on what people want to see and need to hear and want to hear. Some of my best performing content has been hour-long deep dive tutorials in how to set up a self-hosted server from scratch. It's got nothing to do with Tailscale, does it? But it's on the Tailscale channel. And so it's a really weird, unique thing. And I've started to see other competitors of ours copying that strategy, which is kind of interesting. But yeah, it's a very unique way of doing DevRel. Because if you think about what developer relations typically do, they'll be hacking away on SDKs and API code on the back end. And people don't really see it unless you're a developer and you want to integrate with, back in the day, Twitter or something. And then they might go to a conference and talk to a room full of a couple of hundred people. And that talk is never seen again. Whereas what we're doing is, effectively, that same cycle, that same loop, but we're not focused on the developer aspect as much. We're focused on the end user, the top-of-funnel user, mythical person. And getting them excited about using Tailscale for free, and hopefully they have such a good time that they then go, why aren't we doing this at work? Why are we doing OpenVPN? Why are we? And hopefully then that motion turns into a sale two, three, four years down the road, whatever it turns out to be. It's a fascinating space. And I don't know, I kind of feel a bit weird talking about it publicly, but there you go.

I'm enjoying it. I'm sorry I feel weird about it. Do you feel like?

It's just a bit navel-gazey, isn't it? But anyway.

Well, I suppose you're the guest here, though, right? You're the friend here. So you have to navel your gaze a little tiny bit just to be, to fulfill your role. Come on, Alex. You can talk about me if you want to, but that's not going to be fun. The question I have for you, I suppose, is you work at Tailscale. This is going to be on YouTube, and this may be seen by other people. How honest can you be about your curiosity? Do you have an itch? And if you have an itch, do you feel like you have the freedom to speak publicly about the curiosities as a developer, as an engineer, as a Linux nerd, as a home lab, or as a self-host, or whatever?

I mentioned a few minutes ago that the channel to date has basically been just me. That comes with a huge bonus of it's just me that decides what gets put out. So to date, the last two and a half years that I've been at Tailscale, yeah, I've had basically complete creative freedom, which is both a huge blessing and a massive responsibility at the same time, because you're representing the brand. And for example, Kevin Purdie, who is ex-Arse Technica, who we hired earlier this year to write for our blog, he's an amazing writer, by the way. You should totally go and look at the Tailscale blog. He wrote a piece about how to jailbreak your Kindle and put Tailscale on it. And it was on Hacker News yesterday, and we saw a huge spike on the back end. But putting something about jailbreaking a device and breaking a EULA on a corporate blog, there are just certain topics you can't really touch. And that one, it falls under sort of fair use, and it's fine, we think, legally speaking. But there are huge trunches of self-hosting that I just cannot talk about. Wink, wink, nudge, nudge, the stuff behind you is what I'm talking about.

For sure.

And we can't talk about circumventing geopolitical, geoblock, content-right restrictions, and circumventing copyright, obviously, and everything that goes with it. And for me, the reality was some of that stuff was the reason I even got into self-hosting in the first place. So I have to find a whole other seam of stuff to talk about, which can be challenging, admittedly, sometimes. But then we have product release weeks, and you sort of think to yourself, right, well, now I'm going to actually put together an advert, effectively, for the products. But how am I going to do this in a way that doesn't feel like an advert? And so it can be tricky, but also very creatively challenging and rewarding at the same time.

What are you doing in your home lab currently that isn't so much telescope, but more just like, I know you just mentioned image, but could you describe the level of home lab do you have? Do you have a massive unified system? Are you rocking something else? Take me through from gateway to client.

Yeah, unified fanboy over here, self-admitted. You are?

OK.

Yeah, yeah.

It's hard not to like them, right?

It really is, apart from the price. There's definitely a unified tax.

Yeah, the price definitely does get you. Yeah.

I've got the UCG fiber, their new fiber gateway that's really nice. And I replaced a, I think I had a dream machine pro or something. When we were getting ready to sell the house, I just downsized everything from sort of rack gear as much as I could to the sort of small utility switches. Yeah.

Nice.

So I've got like a ton of those Lenovo mini PCs, the Kubernetes cluster I mentioned a few minutes ago. I've got a couple of MS01 minis forum devices doing like all the heavy lifting. I've got 100 terabytes of spinning rust in the basement, give or take, running ZFS that I talked about. That's in a 19-inch Sliger case. But yes, it's pretty minimal, dare I say, for what I do. Like I need a set of infrastructure to run tail scale stuff on for the videos. And then I need a separate set of stuff to run my home sort of pseudo prod. And yeah, I try and keep it as simple as I can, like 10 gig for all the trunks, the unified aggregation switches for like $260 are great.

What about any applications? You mentioned Kubernetes, and that's kind of new. It's like three weeks old for you, I would say, right? That's since Thanksgiving, three weeks?

Yeah, roughly.

So what are you running on there? You wanted to learn it, but what are you running, what kind of applications are on there?

Nothing yet. I don't trust myself with it. Nothing, that's cool. My media server primarily is my main application server. So that has Plex on it, that has Jellyfin, that has all the other stuff, Paperless, Nextcloud, Smokeping.

Interesting. Plex and Jellyfin?

Yeah, because Plex keep turning the screw, don't they, about trying to charge you money to stream your own files, that kind of stuff.

Are they? I haven't experienced that personally, but what's the latest?

I'm ready for their rug pull. It was about two weeks ago, there was another nail in the coffin there. I mean, look, I understand, like Plex need to get money from their audience to support development. The trouble with it is that they gave it all away for free at the beginning, or they gave lifetime passes away for too cheap. I bought Plex lifetime for 75 bucks 12 years ago, and they haven't seen another dime from me since.

Except for a bunch of marketing and new users, potentially.

Well, okay, in my specific case, maybe, but that's not the average.

I guess me too, a little bit, but yeah.

We're not the average user, are we?

No, I would say probably not.

That's not a good business model, like recurring revenue. That's what you need to build a business around. And it's interesting, if Plex launched today with its client app suite, with its fit and finish and polish and codec support and everything else, I'd probably pay five, 10 bucks a month for it and not think twice. But because it has been free for so long, or in my mind, at least I paid that lifetime fee and it is free forever after that. It's tough, you can't ever really charge for something that used to be free or used to be on a lifetime tier. So they kind of ruined their business model 15 years ago and they're kind of paying for it now. So it's tough for them.

That is true, that is true. I like Plex amp. So outside the typical Plex ecosystem, I listen to a lot of my music externally, hence the port. And Plex amp is a dedicated, you got it on an iPhone, I believe, I think you have it as a Mac app via the iPad app, I guess. I don't know the other platforms, I guess maybe on Linux potentially, Android, I'm sure. And Plex has done a great job of having like client apps kind of everywhere. And so you're right. I guess if they had an open core, right? An open core that if I could self host, which I guess we already do self host, then maybe not, I guess you're right. Like how do you charge for that? I suppose maybe particular features, like you said, codex or like I do. So an example of maybe how they can make money, hey, don't do this by the way, is I watch a lot of 4K content in my home lab, in my home. And so maybe I'm a unique, kind of maybe a super user, so to speak. I'm the kind of person you may, I'm an enterprise customer, so to speak to Plex, maybe to get HDR from my 4K and have that codec, which requires a certain amount of developer tooling and engineering to make that work. Well, maybe you pay for things like that, like certain or lossless, I guess that would be kind of sad to force you to compress your own music. But if you had lossless on your WAV files, and that wasn't gated, but maybe you gated to get money, but you're right, it is kind of a road pool to the lifetime customers. Maybe there's a lot more customers they can get that aren't lifetime and they can just kind of give us the past because we've been invested for so long.

I don't know. And do you think piracy is gonna have a moment as inflation continues to bite and people just can't afford, as the media space continues to homogenize, you saw the Warner Brothers acquisition by Netflix recently. Those massive corporate mergers are never good for consumers. They're just, it's always gonna result in long-term, higher prices, less choice, less content.

More gates.

Yeah.

More tall boots.

I do think piracy will have its moment over the next few years as a result of all that homogenization. We kind of had a nice honeymoon period for the last decade led by Spotify, I think, where that was all you can eat, but you're starting to see famous artists now pulling out of Spotify because they don't pay their artists enough. And people like Tidal are coming along and Cobuz and you end up with this fragmented ecosystem. And the same thing's happening with video. Arguably, it's already happened with video because of just how lucrative a streaming service can be. I remember the original promise of Netflix was that it would just be one subscription and you'd get all the movies and all the TV that you can eat. And then that kind of fell by the wayside as the studios realized how much they were-

They're making all of our money.

Yeah, exactly. So video, as you well know, is expensive to produce, let alone blockbuster scale graphics effects and entire industry's worth of crews and sound people and costume and set design and everything that goes into making a good movie. You've got to recoup somehow.

Yeah. Do you have Targets near you? Like Target stores in Raleigh?

Oh yeah, Le Tagee, yeah.

Tagee, yeah, that's true, Tagee. Do you, is that one of your places? Does your wife love Tagee?

Well, my five-year-old does because most times she leaves with some new thing.

My five-year-old loves, well, he's going on six now, but he loves. Let's go to Target, mom, because Target means maybe something for him. The reason why I asked that question was because I don't really care to go to Target, necessarily. I mean, let's be honest, I don't think I like to go to any store unnecessarily. Like if I don't have a mission, what's the reason to be here, to browse? What's the, come on, what's the point here? But apparently it's to spend time with your family and so I'm cool with that, I want to do that, okay? I am okay with that.

I can do it at home without spending money.

We don't have to go to Target and buy the snacks we don't need and the things that we, to be attracted by the end caps or whatever. I'm just gonna not play that game. But I do play that game. And I used to happily play that game because I could be like, hey babe, I'm gonna go over to the Blu-ray section for a little bit.

Oh yeah.

It's not there anymore.

No.

It's not there anymore, Alex. Like, what are we doing? What's happening? The only place you can buy my hard disk content is from the Ultimate Bookstore, Amazon, you know?

You will own nothing and you will like it. Isn't that Cory Doctorow's identification culmination?

Well, I think that was actually the New World Order or maybe the New World Forum or something like that. They did that and then obviously it's part of the identification that he's, you know, extrapolated from that. If Cory was here, he'd probably word salad me for 30 minutes on how I'm wrong. But I love Cory. He's the best. I really love, I love, absolutely love Cory. It's been too long, actually. We should get Cory back on the pod soon. But yeah, you can't really go to the physical store. And I mean, you can't buy CDs anymore either. So whatever. But like, do not take my Blu-rays away, okay? Like, I want to own the film and I want to have it in perpetuity. Like, I don't want to buy the white album on every format. Just let me buy the one and I'm done, right? Let me have the lossless version of it, the blessed version of it.

You mean you don't want to buy the remastered remix deluxe super edition box set on 33 RPM vinyl?

You know, I'm tempted. Reissue. I'm tempted. I'm tempted. But then I would just be buying, I would be like filling somebody's pocket who bought the license to that, those masters, you know? Anyways, whatever. What are your thoughts on the lack of media or the fact that it's only buyable on Amazon?

I think piracy will have its moment.

How will it have its moment though if there's no media to pirate? How will you pirate?

Oh, they're sneaky. They'll find ways to rip it in other ways, I think.

From the stream? Like encryption, like de-encrypt it?

Yeah, I mean, well, ultimately it's displaying on a screen somewhere and there are ways to, I don't know how much we should talk about this. There are ways to circumvent pretty much any DRM if you try hard enough.

Okay. I don't try very hard personally, but I- No, nor do I. I feel the message is something I do who does, yeah.

That's for other people to do, right? I just, I might benefit off their fruit on occasion. Forbidden fruit.

I try to buy all my media, honestly. I mean, I really don't, I'm not calling you out. And I don't even know what, I don't know what would make me want to be, I guess, a pirate or a consumer of pirated material. In this stage of my life. When I was younger, yes, cause like I didn't know any better and I also was poor. And I'm still like, just like marginally less, a little less poor.

Well, we talked about this a little earlier on, didn't we? We talked about the experience, the overall experience of Docker versus Podman.

Right, okay.

And the overall user experience of having five or more subscriptions and having to go to a store and buy a physical disk, ultimately is worse than what you can get if you put your eye patch on.

I see.

The overall user experience is just better. And I think there's a, I don't even know if the movie studios can coalesce around any single service. I don't think it's possible, but until they do, like the quality is better, the bit rate, you know, they're not gonna do rug pulls and remove content. You know, like friends famously was on Netflix and now isn't, it's somewhere else. So like that stuff happens all the time. One of my favorite cartoons, Final Space, just got completely deleted from the legal internet. You could, you know. Really? Yeah, you can't, someone's life's work, the creator of that piece of work is just gone.

And what is it called?

Final Space.

What is this Final Space? I don't know about this.

It's a stupid cartoon, but it's quite entertaining. It's got this cute little green thing in it called Mooncake.

Okay.

Anyway, it's a fun little thing to put on late at night when you just wanna vegetate.

Is this not on Netflix then?

I don't know. It used to be on one of them. I forget where.

Let's see here. I landed on a Google result. Oh, well, at least it's not available on ad supported plan because I am an ad supported user of Netflix. I guess that's more rug pulls. Like let me just increase your price or give you, actually it's not as supported. Actually it is, but it's cause I get it from T-Mobile. So I have a phone and they give me Netflix and I'm like, okay, I'll get it for free. Why would I pay a little bit more to get the ads removed? I don't mind them. And honestly, Netflix does a pretty good job with the ads. I mean, they're not that invasive. I hate the pause screen. I wanna pause to see the thing, not the ad.

I'll do everything I can to avoid adverts. I'll pay for YouTube premium. I will run smart tube to watch YouTube. I will pay for a Netflix ad free tier. The reality is that I just don't wanna support the behavior of companies that will do these kinds, like Final Space is a great example. So not only was it canceled before the show's narrative finished, which happens all the time on streaming services, but they also just deleted the entire series from Netflix so that they could use it as a tax write-off. I just don't wanna support that behavior. It's the same thing with Google Photos. It's the same thing with media. I think without sounding too much like a mouthpiece for like Corey or something, like genuinely it's egregious what a lot of the industry, the big tech industry is getting away with. Using our data to train their models to then sell us back even more. Like the AI bubble, whatever you wanna call it, it's an industrial bubble. So there will be some useful stuff comes out the other side of it. But when it goes pop, I think it's gonna be very painful for people.

Interesting. Well friends, I'm here with my good friend, Chris Kelly over at Augment Code. Chris, I'm a fan. I use Augie on the daily. It's one of my daily drivers. I use Clod Code, I use Augment, Augie. And I also use Amp Code and others. But Augie, I keep going back to it. And here's where I'm at. I feel like not enough of our audience knows about Augment Code, not enough about Augie, the CLI. It's amazing. I love it. What can you share?

Yeah, we often say Augment is the best coding assistant you've never heard of. And that's both frustrating as someone that works there and it's like very proud of the work we've done, but also like inspiring. Like we wanna go and sort of punch above our weight. We just like, we aren't anthropic and we aren't open AI. And so the quality of the product itself, with our context engine, once you do touch it, people are like just blown away by that. And so like that keeps me going every day.

So not to bear the lead here, but this is a paid spot. You are sponsoring this show to get this awareness. Now at the same time, we're selective and I love to use your tool, but there is in the world. So a lot of developers look at the space and they say, okay, well, how long can this work? How long is this sustainable? In the case of cursor or a windsurf, or you pick the name and you think discounted tokens, help me shape a lens for audience.

I think it's a lot of awareness, right? Like cursor got a lot of publicity early on for like fast revenue growth, which well-deserved.

I think, frankly, some of the media gets the story wrong and that like, if I gave you a dollar 50 for every dollar you sent me, I'd be the fastest growing startup in the valley. And so when you're selling discounted tokens, yes, of course you're gonna grow very fast, but all of that money plus more goes to the model providers.

So I think the real story is the story of Anthropic and being an API provider, I think the market has just moved so fast and there's so many pieces of competition out there that it's just hard to get noticed.

So friends, I love augment code and I love using Augie and I highly recommend you use it. I love using Augie. I can hand Augie a well-defined specification, a well-defined PEP, as I call them in my world, an agent flow and it executes flawlessly. So the cool thing about Augie that I love most really is that context engine and I can hand it a task and it can just churn away on my well-defined plan and just never bother me and accomplish the mission. It is so cool leveraging the latest models, the context engine and all the fun things behind the scenes in that awesome CLI. So yes, go try it out, augmentcode.com. Right in the top there is a CLI icon, a terminal icon. Click that, install it and change your world. It's gonna be awesome, augmentcode.com. And also by our friends at Framer. You know, most design tools, they lock you behind a paywall. Well, Framer flips that script. It is a free full feature design tool that does something most site builders cannot. It's actually designed for designers. Framer has already built the fastest way to publish beautiful production-ready websites but with design pages, they've defined and redefined what it means to design for the web. This isn't a Webflow clone or a WordPress competitor. It is a true design platform, vectors, 3D transforms, gradients, wireframes, all the tools you actually use and they're all free. Unlimited projects, unlimited pages, unlimited collaborators and here's the kicker. You design, you iterate and publish all in one place. There's no Figma handoff. There's no messy HTML imports. There's no tool switching. So if you're ready to design and publish in one tool, start creating for free at framer.com slash design and use the promo code change log for a free month of Framer Pro. Once again, framer.com slash design. You said AI, we have to talk about AI.

We haven't talked about AI yet.

You said that.

You look at the utility that people get from these things. Certainly like when I'm writing a video script, I'll run a bit that I'm stuck on through it and say, is this, what do you get? What do you understand as to what I'm saying here? Does this make sense? Like I'll do it into a voice memo and dictate it and then give it to Claude and say, make, to polish this explanation about something or tweak this bit of code and make this do this instead of that or whatever. Historically, if you look at industrial revolution, like we're going back a couple of hundred years here, there was a huge amount of speculation in the industrial revolution era. A lot of people made a lot of money and that's happening now in the AI space and not every idea that should have been funded got funded. If I mean the other way around, there were a bunch of ideas that shouldn't have been funded that did get funded simply because people were hoping, speculating, right? We've seen this many times over the years. But these industrial revolutions and the AI revolution is one of those because it's a tooling thing as opposed to say 2008, which was a made up financial banking bubble, whatever. There was nothing material, moving humanity forward there. So to say what you like about AI, but it's here to stay, I think. And the big players like OpenAI, Anthropic, Google, of course, they'll be here to stay because every other startup is building their tooling on those big guys' models. They're not doing the R&D spend to train their own models. They're just using Anthropic models for an API key behind the scenes. So I think the big guys will be okay, but it's the circle of money that has me concerned, like OpenAI giving Nvidia money, Nvidia investing back in OpenAI and so on. That's worrying.

Yeah. What is your relationship with AI? Are you a daily active user? Are you, what's your stance so far?

Yeah, I'll reach for it quite often now, yeah.

Okay. Do you have the itch right now on the pod to be like, can I just clod something real quick, just cause you have to?

No. No. I mean, if I'm stuck trying to explain something, I might, instead of where I'd have reached for a Google search a year ago, I'd probably reach for a chat now instead with a clod or a chat GPT, whatever, but it's just replaced Google entirely for me. Yeah. Whether that remains once the adverts start getting dropped, we'll see.

But yeah, I've seen some things shift in chat GPT that I talked about on a podcast recently. We're talking about trying to determine which podcasts were the better AI focused podcasts to listen to. And we're hoping to find practical AI in that list, which is a show we co-produced for a while with Chris Benson and Daniel Whitenack. And they've since taken the show on their own. We still are working with them behind the scenes, but we have some skin in the game, let's just say. And so we're just curious, how did it rank whenever you search for it in chat? And when I did that in chat GPT in particular, at the bottom area where we would put a new prompt in after you put the initial one, you've read the whole thing, well down at the bottom it said, create a playlist on Spotify. And I was just like, may I get the connection there? But I don't, because I'm not trying to make a playlist. I didn't prompt you about a playlist. I guess I kind of did by asking you about a list of podcasts. So maybe it's to some degree relevant, but like they're probably getting money in some way, shape or form or a kickback. I'm assuming.

But if they aren't now, that is certainly a business model they will deploy in the future. Yeah. When the speculative bubble does whatever it does, they haven't turned on ads yet. They haven't turned on X-rated stuff yet. There's a whole bunch of levers that Sam Altman and people at his level can pull to monetize that they aren't doing right now. Just the whole space doesn't, long-term at the moment doesn't add up because the energy requirements to power it are just so vast. I mean, they're spinning up nuclear power stations to power this stuff. We can't manufacture enough NAND chips. We're seeing that it's just bonkers. It doesn't add up long-term, but.

And when you mean NAND chips, you mean, are you talking about like RAM type things?

Yeah, yeah. So the little black chips on your RAM sticks, those are NAND flash chips. They go on M.2 SSDs. They go on RAM sticks. They go everywhere, in everything. Literally everything that has electronics in it has some kind of NAND flash storage somewhere on it. And so it's gonna affect regular people really soon. Not, I mean, right now as a PC,

they're on- RAM plus you have to go through the roof right now. I mean, like, have you tried to buy RAM right now?

No. Yes, just building the Kubernetes cluster a couple of weeks ago.

Yeah, I mean, thankfully I bought some RAM recently for a build of mine just before that spike. And I was like, I was kind of sad because I needed, I actually meant to buy a set of four DIMMs as one kit, but like a full, I bought two disparate kits of two thinking I could just put them together. And anyways, I will spare the details, but it's just like with this particular processor, it's an AMD processor. And it's got challenges with like a non-kitted or what do they call those? You probably know the term, when it bend. They're not bend to each other. And I didn't, I'm learning, right? Like any home lab, I've never had to buy four sticks of RAM and concern myself if they were from the same bend or bend together.

It's just tricky. RAM on the new chips, it's very sensitive to the timing.

They are, and especially the faster speeds. I mean, I think when you're on the edge like that, you're going to the faster speeds, then you should expect some dragons to some degree. But the dragon I didn't expect was the cost dragon. Cause I was like, I got to swap these out. I was like, Oh my gosh, what happened to the prices? Like I didn't pay that two weeks ago. And then now it's like astronomically priced. I would even like, I'm like, okay, fine, I'm cool with it. I'll just only use two for now. I'll make do.

Yeah, it's tough. Like, I mean, you should know everything. Like even daft stuff that you don't even think needs RAM or needs a NAND chip in it, like in your car, new cars, they're probably expensive enough to absorb a few hundred dollars here, a few hundred dollars there, but stuff like, I don't know, a coffee pot, the smart coffee pot will have a NAND chip in it. And that's not quite enough margin to absorb a three, four X price increase without that price going up. And yeah, it's just, it's going to be tough.

Yeah, that's interesting. So that's the negative side, I would say of AI. What do you have a positive side to AI? How has it impacted your home lab?

Well, take a look at this website, quicksync.ktz.me. 100% vibe code slop.

Oh gosh, what is this?

This is a benchmarking website that I put together over the weekend that allows people to benchmark the transcoding performance of their Intel CPUs. So trying to answer the question of which CPU do I buy for my media server?

Oh my gosh, this is spectacular.

And this was me and Claude for about two or three evenings this weekend. So my positives around AI are that I know just enough to be dangerous as someone that's been adjacent to developer technologies for the last 15 years. I'm not very good at actually writing the code. It takes me way too long and I get bored before I get finished. This allowed me to get something finished and out the door way faster than I ever could have done by myself. OK, I didn't learn the intricacies of how the API calls work and what have you under the hood. But honestly, I don't need to. It's just a flat JSON file that gets stored in Cloudflare R2. This page loads that JSON file client side and then renders the web page locally. And that's all I need to know. As kind of like a normal developer adjacent person, I don't need to know every intricacy. So for me, it's like this has allowed something to exist in the world that I would have really liked to have seen 10 years ago when building my first server. And it's taken the information from being a flat text file to being something interactive that people can kind of enjoy, hopefully. So you know.

Help me understand what I'm actually seeing here. So how are you getting all these benchmarks?

It's crowdsourced. So you see up in the top right hand corner, you can clone the Git repo to your server. And then you could just run the benchmark script. And then it will pull down a Jellyfin container because it just uses FFmpeg. It pulls down some drone footage that I edited together a couple of years ago into a test file. And then it would just encode those files in different codecs. So it would do a CPU software encode to give you an idea of the raw horsepower of your CPU. It will then jump into the Quick Sync engine, which is the hardware transcoding thing built into your CPUs. So it does 1080p. It does 4K 8-bit HEVC, and then 4K 10-bit HEVC hardware encodes. And so you can see on the graph, the performance of the newer chips is a lot better, as you would expect. But the energy usage also goes up. So it's like trying to find that sweet spot between energy usage, performance, price, et cetera. Crowdsourced information, this is not a scientific experiment of super high accuracy, but it's the best we've got in the real world.

Is there a plan for an API? To do what? I guess pull it, look at it. I guess you do have to have the website. I guess an API just lets you have access to the data in a different way that isn't a flat file. It's JSON, a lot easier to sort of parse.

It pulls it down as a JSON file right now. So I guess I could just pull a link to the raw JSON file. Sure. There's nothing proprietary or sensitive in there. The only user identifying information for GDPR reasons that I wanted to keep was an optional submitter ID. I don't record IP addresses or anything like that. So when you submit your results, you've got the option to give me your username if you want to. You don't have to. But everything else is just generic CPU score this. So yeah, I mean, there's nothing sensitive in there.

Wow, I mean, I like this kind of stuff. This is those kind of things like you had said. This is where I'm really, this is where I'm truly enamored is that, like you said, you know just enough to be dangerous, not enough to actually code it yourself. And if you did, even if you did, you'd probably get bored because it would take you weeks or months or just too much of your life taken away. But you have the knowledge to connect different dots. That someone, only someone like you, or maybe a small few like you, would have those dots to connect. And then now you have this superpower, essentially, to take that into an idea, architect it into some version of a platform.

And then it makes me more productive. It makes me more productive.

And this will pay dividends in the future, though. I mean, I don't know if this has a sad ending or happy ending because at some point, maybe Intel CPUs go away. It's a whole different subject.

I'd like to expand this to support things like Nvidia GPUs because obviously they have the NVENC encoders in them. The newer AMD CPUs have hardware transcoding as well. So expanding it to support different vendors, Intel Arc GPUs, of course, are important to test as well. I just don't have any of those on hand. So I do want to expand this as we go. But for right now, at least, it's just Quicksync. But over time, yeah.

So I'm looking at Morgan Peterman's, I love that last name, Peterman, his profile because he is the dependency that you came from. You forked his Quicksync calc, I assume, to build on top of.

He and I worked together at Red Hat. And we were bored one afternoon and put together this initial version of the script. So I just forked it and built on top of it and it's where we are. It's open source, baby. It's in action.

So his commit was like last year, so 22 commits there. You're now at 127. So 100 or so commits on top of the original Quicksync calc. I imagine you'll probably contribute back to upstream potentially, just leave it as its own thing, maybe expand it.

I think it was always in my attention that my repo would be the primary, I think, in this case. Because I wanted to give Morgan a lot of credit because he did a lot of the work on the original script. But in terms of long-term ownership, that was what I was more interested in. So we could contribute back to upstream.

I'm not trying to call out what you, I'm just thinking of how do you architect it. Because it's not called Quicksync benchmarks. But if you're talking about NVIDIA GPU-type things, then it goes from Quicksync benchmarks to something different, a different life to it. It morphs at that point, which is really interesting about this curiosity because you scratch one itch. And then next thing you know, you're like, well, now I can do that. And I've got this idea that connects there. And that idea influences this and enables that.

In a way that I couldn't do just as a solo dude before AI came along. Because you're at the velocity where the ideas can flow. They still flow faster than what the AI can generate, but not that much faster.

Yeah. That's the phenomenon that I think a lot of folks are experiencing. And I'm curious, what were some of the things that you did to do a little less vibing and a little more coding? I know there's this pejorative nature to the word vibe code. Or if you have vibe coded something, it's like, oh, come on now. You don't know Rust. You didn't write that yourself? Seriously?

Yeah. I spent a long time in the planning phase with it. So before it wrote a single line of code, I probably spent two, three, four hours going through what the various pitfalls of different decisions might be. And then once we talked through all of that, the actual coding part only took about 20 minutes. It was kind of stupid how fast it was in the end.

It is. I'm with you on that. Plan for hours, code for five minutes.

But only if you've taken the time. Otherwise, it's just guessing.

Yeah, exactly. If you didn't plan all that, and you didn't think through, and you didn't go back to the specs or the RFC, or just something that was the crucial piece that was the one thing that would have stuck AI because it would make all these assumptions in a plan that wasn't well-formed, well, now you've handed it a truly well-formed potentially even with an implementation mandate sitting next to it. That's kind of like not the implementation, but the various things of proper code or different ways you want to do things so that when you get to the implementation stage, it's not guessing. Because you've already thought through a lot of the, like I said, pitfalls. That's interesting.

Like I said, I know just enough to be dangerous now.

Well, I think that's what we need is the thinkers in this case. And that's really where I'm camping out at. I see this new power available to people who can wield it because they have some experience in history, some wisdom, so to speak.

But the flip side of that, though, is I didn't have to collaborate with anybody on this. Whereas in the original case, I spent a lot of time collaborating with Captain Morgan, right? These days, I can just sit in my room and do it by myself. And so the collaborative spirit and nature of open source kind of diminishes a little bit because I'm so empowered by it. And that's a risk. I don't know how big of a risk, but it's a risk in all of this.

That doesn't bother me, really. I mean, not necessarily because even my own personal usage of open source hasn't been in the collaborative. It's been in the consumption and the availability. And I would say maybe the hyping and the popularizing and the showcasing. I haven't really had a lot of collaboration personally in open source. So that doesn't really personally bother me. I'm also kind of a fan of open source, not open contribution. I don't think you have to. They're not cut from the same cloth necessarily. I don't think you have to be open source and open to contribution. There is a way you can vision an idea or have a vision for it. And sure, you may put it out there as open source because you want to share. But it doesn't mean you really want the feedback loop because you've got a plan.

The other downside, of course, is that how did these models get smart in the first place? They got smart off the back of all of our open source sharing. And now we're just feeding back themselves into themselves. There's got to be a reckoning.

I mean, what a unique thing to think about. So honestly, this is really an interesting thought experiment I try to push myself to, which is kind of a mind bender. I try to think hundreds of years or at least a certain amount of years in advance from today that future atoms or future Alexes look back on these moments. And they're like, wow, thank you so much for putting all this blood, sweat, tears, effort into the open so that we can have this thing that now helps us make I don't know what. I have no idea what will get made. But even in this moment here, you're expressing this. This is a monumental feat, like producing this website and this level of care in a few days and 100 commits. You know what I mean? Solo, wow, where will we be 10 years, 20 years from now because of that truth, because of that openness. I sure hope it doesn't backfire on us. I'm not wise enough to see that future or even speculate. But in the moment, I feel like that is a net positive on the future because it's already net positive to me in this moment.

Yeah, I think there's a bunch of people, a huge tranche of people actually, that kind of miss the forest for the trees in this regard. My take is typically a pragmatic one of, well, if I don't upskill and use these tools and I don't start solving these problems for myself using these tools, someone else will. And then I'll become less relevant as a human. So I think the superpower of these things in the long term is the empowerment they give the individual to solve their own problems and scratch their own itches, which then feed into your collective experience as a human to solve other problems down the road. And you both make each other better in the long run. And it's going to be really interesting how the business models of these AI companies play out over the long term. But I'm optimistic that they won't all be complete.

Optimistic? You don't seem optimistic. You seem hesitant.

I am. I am hesitant. You look at how the last 25 years have gone with Apple and Google and Amazon and Microsoft and the state of the current internet and misinformation and everything else. Let's not repeat those mistakes, I guess what I'm trying to say.

Yeah. Yeah, for sure. Well, for the moment, we can enjoy our feats of brand new quick sync, deep dive benchmarks. I love this, man. quicksync.ktz.me. Link it up, of course. It is on GitHub. Are you open to contributions?

Yeah, yes, please. Yeah? Yeah.

Is there any other way? Did you license this? Was there an original license that you inherited from the?

I think it's MIT.

Upstream? Is it MIT?

Yeah, I don't know. One of them.

I don't see a license file at all, honestly.

Oh.

Yeah, I think it's not licensed necessarily in the moment.

Oh, well, it will be by the time this episode airs. It will be.

Well, MIT FTL, or FTW, sorry, not FTL. I was thinking about something else. Well, Alex, I don't know what else to talk about, man. I know I love talking to you, man. I really, I was thinking, you know what? Who's a good person to chat with at the end of the year? And somebody I haven't talked to in a while. And I knew.

I've run out of all the other options. Let's talk to Alex.

No. No, no, no, no. Not at all. We actually did have a plan for somebody else. And they did fall through. And you were a fill-in. But it wasn't because of that. It was, you know, I like talking to you. I think you got some cool ideas. I mean, this benchmarking is the best, man. And I knew you were going to move to London. And I thought you were an SF for some reason in our back chats and was reminded, yes, Raleigh.

North Carolina. It's a snow day in North Carolina today, actually. We got about an eighth of an inch of snow, and they shut all the schools.

Yeah, right? Like, scary. That's like nothing. I guess before we leave here, do you have a favorite Linux distro at the moment?

Talos Linux. Justin Garrison would be very happy with that.

Yeah, he would be very happy with that. Is it because of your recent KDS explorations?

Yeah, we sort of touched on the boot C kind of immutable nature and disposable declarative nature of operating systems. I'll tell you. If Talos existed for just, like, non-Kubernetes regular Linux, that would be a chef's kiss.

Help me understand. What are we missing?

Ooh, I don't know. Have we got time for this? I just like the fact that the nodes are completely disposable. You know, the storage, the nodes are, like, for years, you've written Ansible playbooks to configure nodes and install packages and do this and do that. And now I have a declarative file that says, system, do this. And when it bootstraps itself, it doesn't know what it is until you give it that instruction. And then it just figures it out. There's no config drift. It's impossible for there to be config drift. It just is what I tell it to be exactly. And the only way to make a change is to update the config so it gets versioned. It gets all the niceties of that.

Is that kind of like atomic, kind of like boot C then? That's kind of what boot C is, right?

Yep, yep, yep. It's the same idea. Different implementation, but same idea.

Yeah, interesting. And so just hypothesize for a moment, in the non-Kubernetes world, how would you see a Talos that isn't that focused on a different alternative? What would that be to you then? Where would you use it?

For a home server. I don't want to have to do an apt-get to upgrade every six weeks and then have something break, and suddenly my server doesn't boot anymore. It would be nice if I could just update a config file, push that to Git, and then it just figures it out and reconciles those changes. A bit like Argo CD and Flux do in the Kubernetes world, but at the OS level. That's what I want. That's the world I want to get to. Push a configuration change to Git, and then my server just figures it out.

How different, and maybe this shows my cards here a little bit, but how much different is that from a script that you wrote, like a literal bash script that just runs your system, runs all the apt-gets and stuff like that, like configures your system on boot. How much different is that?

Well, you've got to put a ton of logic into the bash script to do idempotency. You've got to do a ton of logic. Nick's kind of solved some of these problems a little bit, but it's a very imperfect solution. Let's say you have a list of packages you want to install, and you take one of those packages away. There's now got to be some logic to clean that up, and then there's going to be some residual files, maybe a config file left here, something left over there. You just don't know. And so the nature of declarative stuff is that you can test a lot more easily like a specific implementation of an NVIDIA driver, maybe as a kernel DKMS build failure and all that stuff in the old world, but in the new world, that literally is impossible because it built, it's an immutable thing, and so we know it's going to work.

I think that's what boots used to solve this, which is why I was kind of interested in it, but I just couldn't figure out how to use it. I just don't know yet. I haven't explored it enough to have a good-

I never quite know how to say this guy's name. Is it Jorg Castro? I never quite know how to say his name, but he's running the Fedora Silverblue. There's some cool stuff happening in this space very soon, I think. And yeah, it's going to be really fun when it lands.

We put out a podcast recently called the 4DIMM Problem, which I alluded to earlier. So I have a AMD Ryzen 9 CPU, and I have 4DIMM's RAM. I should be able to do that, right? Well, let's just say there's a 4DIMM problem. And Nabil and Andrew and AJ and Ron are in our Zulip talking about some different stuff, and Andrew O'Brien actually said, I wonder if Adam, this is me, wonder if Adam's tried Bluefin now that he's settled into Fedora. He's mostly a Debian person, as he says here, but he said, I've tried Bluefin based on Fedora Core after a ShipIt episode about it, which should have been a while ago, because ShipIt has since ceased to be an active podcast. Amazing show, obviously. Archives are still there, check it out. No, I have not tried Bluefin, but I am interested in bootsy in this world that is creating this atomic declarative. And now that I hear you talk about Talos in that way, which I have not played with Talos. I have not personally played with Kubernetes really at all. And I probably should spend a Thanksgiving weekend, like you did, just knee deep in it. Like I should have so many things, but I haven't yet. I am interested in it, I just don't, I don't know where to apply it yet. And that's what I would like a guide on. What could I do with it? Because that always gives me a reason to play with it. That's how I get curious, is I have a reason to do it. I mean, the reason why I have different VMs and different things happening is because I learned how to use Proxmox and a bunch of cool stuff around Proxmox and different flavors of Linux from Fedora as a server or Ubuntu as a server or pick your battle, whatever you want to do. So I haven't done that yet though, but I'd like to, I just don't know where to begin.

Linux is an endless rabbit hole. That's the best thing about it.

It sure is, it sure is. So are you going to be at Texas Linux Fest next year?

If I'm in the States, sure, why not?

Okay.

It might be a bit of a long way to come from London, but I don't know, it is a good time.

So hopefully- I think it should be October-ish, is that right? I don't know, I think it was in October this year.

Yeah, that was the thing, because it's normally in March or April time.

Oh, is it, okay.

And then they moved it.

They moved it. I like the new time, it's good.

The issue is for me, it runs up against all things open and we had a company offsite this year. It was just a lot of stuff. So yeah, it's a tough one, but.

And all things open is a staple for you. You don't miss that.

But it's five minutes down the road, so yeah.

You can't miss it, right? You have to go because it's like, well-

It's a great conference, yeah.

It is a great conference. It is an absolutely great, you know, this year they moved it on me too. It's normally at the end of October and this year was at the beginning and I planned some travel earlier in the year. And when I realized that the times collided, I was like, oh man, I can't go this year. And so we didn't have a presence there this year. We're normally, we're kind of like a staple in the hallway track doing our booth thing and podcasting.

I did wonder where you were actually.

Well, that's why. That's why, it's not cause Todd was mean. Todd's amazing. The whole team that runs the conference is amazing. And I love it. And I was really sad to miss it this year, but what I was trying to say though, is that if you're going to be at Texas Linux Fest, we can plan on the Switch and maybe we could do a meetup. Like, hey, we'll call in advance. If we can do some sort of like precursor to it, let's have a little fun. I know that it's usually a Friday, Saturday thing. So we can plan it for Friday. It's usually the day I'm most available. So I'm a family guy and Saturdays are reserved for family. But we could certainly do the Fest all day and maybe a meetup that evening at the Switch for dinner or something like that and make it a fun thing. What do you think?

Let's make it happen.

Let's make it happen. Well, Alex, it was good to talk to you. Merry Christmas to you. I know it's just a few days away or a few weeks away at this moment. Two weeks I believe my kids are counting down. They're excited about it. I'm excited about it as well. And always good talking to you, man. Thank you so much for sharing time with me and just being a little nerd, man, like I am. I love it. Thank you so much, man.

Thank you. I'm going to go and watch Silicon Valley now.

Well, you've been watching me back there. What now?

Now I want to watch it for real.

There's Big Head. Where's Big Head? What's he doing?

What are you doing, Big Head? It's so cringe.

Oh, this is where they're talking and he's like being the bad friend there. He's like, oh man, I think I'm gonna go back home right now. No, you shouldn't go back home right now. But you can't be part of Pied Piper anymore. Okay, yeah, that's what's going on right there. That's episode two of season one, by the way. Alex, you're the best.

Thank you so much. Awesome. Thanks, dude.

Have a good one. So I've made a list. My Christmas to-do list, I suppose. Here in the home lab, image, image frame, that's definitely on my list. It involves zfs.rent. So embarrassing to admit on a podcast that I have no backup plan. Raid, y'all, is not a backup plan. You have to have something offsite. Alex is totally right. I am vulnerable at this moment. And I never imagined that someone would create zfs.rent. I'm gonna see if I can get him on the show. We'll see. I also want to play with Talos. I have not messed with Kubernetes personally yet, much like Alex, a little intimidated by it, but it involves hardware, so that's always fun. So I might build out a three node cluster, or I might just do it via Proxmox with VMs. That can be cool. Just to learn. I do have more plans for applications in my home lab. I've got a very cool read later application I'm working on. You know, it's one part archive. It's one part read later. The project is sort of morphing in my mind on how I can use it, but it's pretty cool. And I'm gonna expose it to myself on the network as an application. And I'm thinking that might be fun to do with Kubernetes. Who knows? Maybe not. I am enjoying Proxmox, so of course that's on my list. And it should be on your list too. But if you haven't yet, head into Zulip. Go to changelog.com slash community. It is free to join. You are welcome here. Hang out with us in Zulip. It's like Slack, but better. And I want to hear what you're doing in your home lab this holiday season. Comment on this episode. Easy to find inside Zulip. We'll see you there. Of course, big thank you to our friends and our sponsors. Tiger Data, love Tiger Data. Agentic Postgres. Our friends over at Depot. Build those builds faster. It makes no sense to wait. Our friends at Augment Code. Augie is one of my daily drivers. I love Augie. You gotta check it out, AugmentCode.com. And of course, a massive thank you to Fly. That's our home, and it should be your home too. Learn more at Fly.io. Of course, the big freak in residence, Brake Master Cylinder. New beats, love those beats. Brake Master, you are awesome. But friends, that's it this week. The show is done. We'll see you on Monday.