LINUX Unplugged - 371: Cabin Fever

Episode Date: September 15, 2020

Friends join us to discuss Cabin, a proposal that encourages more Linux apps and fewer distros. Plus, we debate the value that the Ubuntu community brings to Canonical, and share a pick for audiobook fans. Chapters: 0:00 Pre-Show 0:48 Intro 0:54 SPONSOR: A Cloud Guru 2:25 Future of Ubuntu Community 6:51 Ubuntu Community: Popey Responds 9:31 Ubuntu Community: Stuart Langridge Responds 16:26 Ubuntu Community: Mark Shuttleworth Responds 17:30 BTRFS Workflow Developments 19:09 Linux Kernel 5.9 Performance Regression 24:48 SPONSOR: Linode 27:34 Cabin 29:48 Cabin: More Apps, Fewer Distros 33:41 Cabin: Building Small Apps 36:40 Cabin: What is a Cabin App? 44:34 SPONSOR: A Cloud Guru 45:20 Feedback: Fedora 33 Bug-A-Thon 47:53 Goin' Indy Update 49:40 Submit Your Linux Prepper Ideas 50:11 Feedback: Dev IDEs 54:15 Feedback: Nextcloud 58:20 Picks: Cozy 1:00:25 Outro 1:01:38 Post-Show Special Guests: Alan Pope, Drew DeVore, and Stuart Langridge.

Everybody loves an Amiga story, and that's why we are first to bring you the news that the Amiga fast file system is returning to the Linux kernel. Who needs this? I think just about no one. Some SUSE developer. You know it. I mean, I would bet you, search that article. I bet you three bucks that it's a SUSE thing. It just sounds like them.

It just sounds like them. Oh, yep. Okay. A SUSE developer and kernel maintainer, David Sterba. Yeah. Oh, yeah. A Seuss developer. I guess he was just reviewing it and committing. Oh, I can't blame him. Someone else authored the patch.

Yeah, but what are the chances somebody from Seuss is attached somehow? What are the chances? I kid because I love him. Hello, friends, and welcome to 371 of your weekly Linux talk show. My name is Chris. My name is Wes. Hello, Mr. Payne. Did you know that this episode is brought to you by a cloud guru?

Are you looking to get a high-paying career, maybe move into the cloud and make some good change? Well, there's no better place to start than getting a certification. ACG has helped more than 2 million people scale up on the cloud. AWS, Azure, and Google Cloud Platform. Head over to cloudguru.com and get started. Now, we have a great show today. It's a special smoky edition. We're recording it a little early because I'm hitting the road and we're getting slow cooked while we do it. It's a cold smoke. It is a cold smoke. It's a wood smoke that we're surrounded in. So if you hear Wes coughing during the show, it's definitely the wood smoke.

Before we go any further, I've got to say hello to Drew. Hello, sir. Hello. Hi, Drew. Wearing your Sunday pajamas, I see. Wear them on Tuesday, too. You know, it's always nice to have your Sunday pajamas on Tuesday. And, of course, everybody in the mumble rooms in their pajamas.

Time-appropriate greetings, mumble room. Hello. Hello. Hello. Hello. A special hello to Mr. Stuart Language. Thanks for joining us, Stuart. Hi.

Thank you. Hi. Hi. We're going to be talking about Cabin in a little bit. I mean, we got a few things to get to today. We got more than just that, Stuart. Hi. Hi. We're going to be talking about Cabin in a little bit. I mean, we got a few things to get to today. We got more than just that, Stuart. Jeez. But we're going to get to that a little bit.

So do feel free to jump in on any topics as we go. Hello, Popey. It's good to see you too. Hello. Hi. You'll be joining us for the Cabin chat. But you know what, Popey? I also give you very special permission to jump in on any topic.

Just any topic, Popey. Oh, you're too kind. All right. Well, let's start in the news. Not the Amiga file system update, but the future of the Ubuntu community. Now, I'm going to warn you, I got up real early today, so I might have a slightly spicier than usual take on this one. And any guidance from our friends in the Mumba Room would be appreciated. But I came across this thread on the discourse.ubuntu.com site. You may have heard of it. It's a new thing. Brand new. You should go check it out. No, not really. And I'm not going to name

names at first here, but eventually this conversation led to Mark Solworth coming out of the woodwork and responding to this. And it starts kind of directly at him. He says, this is the original poster. I've seen discussions about the abrupt loss of leadership in the community. And this is a sad event. I'm not entirely, however, surprised. If you've been around the project for some years, you'll notice that there's been a progression towards de-investment in the community for many, many years. But the thing that's most saddening, though, is that Mark, who has benefited from countless thousands of volunteer hours, which are certainly worth millions of dollars, didn't have the respect for the community to even articulate why he abandoned the community and has been silent on the collapse of governance,

which he played a part in since he, after all, is the project leader. Now, it's hard to read those two first paragraphs and not feel like there's a tone in there. And the tone that I impart on that is one of all dad does is work all day and night, two different jobs to put food on the table and a roof over our house. Daddy's too busy keeping the business running to pay attention to us. And we're sick of it. That's what I hear when I read that. And I appreciate that this is a complex issue, but the structure of this is, it feels so privileged. It reads to me as so extremely privileged. I grant that there is some legitimacy to the complaint. There is value derived from the community, and there is a certain level of investment that canonical should be making

in the community to perpetuate that i i grant that argument but i feel like this tact is it's almost designed it seems like to elicit a response from shuttleworth it's quite provocative and yeah seems to warrant a response so mark chimes in eventually pretty far down in the thread because you know something like this, people can, they all got to jump in with their two cents. Mark responds in part, I'm not absent. In fact, for the past few years, I've set aside all other interests and concerns to help Ubuntu get into a position of long-term sustainability.

This has been an amazingly difficult job. But I set my mind to it precisely because I care that the Ubuntu community has a backbone which is durable. I'm rather frustrated at my own team because I have long allocated a headcount for community lead at Canonical, a post which has not been filled. It's necessary to have a dedicated lead for this. Not so much because the community needs leadership, but because its self-motivated leaders need support. Think of the role more as community secretary than community advocate, helping to get complicated pieces lined up to empower others to be great. The project has continued to grow in complexity and capability. There are more people than ever working on it, more people than ever

making demands on it. So getting things done requires patience and coordination. Helping motivated community leaders to be effective in driving their work forward is important to me. You could really read a lot into this, and I'm curious to get everyone's take on this. One of the last paragraphs he writes, Mark Shuttleworth writes, I watched how CC members stopped coming coming to meetings stopped organizing their meetings stopped driving activity this is obviously not a universal picture there have been harder working and less hard-working ccc bodies and there have been more effective and less effective cc members i understand it's hard to put a lot of effort into something that doesn't seem to correspond directly to a specific project or outcome and he kind of talks about how he had

other issues with decision making. I'm just going to take a stab at the dark. And I'm going to say, you know, Mark's got a limited amount of time. And he looked at the effectiveness he was deriving from this community council and decided his time was better spent investing in making canonical sustainable and durable. Yeah, it's also, it's, yeah, just hard, right? The limited amount of time, a little amount of oversight that one can have over all these things.

And if no one else is doing those, if it's not actually being delegated or being handled, yeah, at some point stuff slips through the cracks. Pobi, from your vantage point, do you feel like the role that community plays in the creation of Ubuntu has shifted over time? Oh, definitely. Things have changed over time. Every open source project does. I think I recognize a number of things in this

conversation one thing i do have to point out is i'm glad that ubuntu has the kind of community that will pick up on these things and will tell us hey you're doing something wrong and we'll come to a place they know we're gonna see like the ubuntu discourse and we'll lay out their stall for how they think things should be improved. And whether you think this was done to trigger Mark or whether you think Ben, who wrote the first post, had good thoughts in mind when he wrote it or not, the outcome is good because it did ping Mark, obviously, and it made him think about this and articulate his thoughts. And I've been on the CC about 10 years ago before I started working for Canonical. I was on the community council,

and I certainly recognize some of his comments about the CC goes between periods of not having anything to do because they're quite a reactive group and having really difficult problems to deal with. And that's very hard going from nought to a hundred, like very quickly and having to deal with problems. I think one of the big problems with some of the community councils, what I mean by some of is, you know, people's term runs out and they may get voted in again or they may not. And so the staff that people change

over time i think one of the mistakes is people feel that they have to wait for mark to say yes to something and that's not the case when you're voted into the cc it's recognition that you already have some level of recognition in the community you already have autonomy you already have some level of recognition in the community. You already have autonomy. You already have a position of leadership. And so you should feel empowered to go and fix things or go and have conversations with people. And I think people in previous CCs have felt like they had to run everything past Mark. And that's just not the case.

And, you know, sometimes he didn't turn up to meetings because he was in a business call or on a plane or, you know, there's reasons why everyone can't be at every single meeting. But it did tail off about a year ago or so when people kind of gave up. And I'm glad that this thread has been triggered because it's reinvigorated Mark and reinvigorated the people around Mark who hopefully will solve this problem and reboot the CC. Fair enough. In fact, there has been a development before we get there, but I wanted to ask Stuart, and I'm trying to think of a way, I was trying to think of a way to phrase this that doesn't say, Stuart, you're old, you've observed a lot, but I mean, Stuart, you have been around for a while in the community and you must have observed there's a different

value, it it seems like companies many companies but canonical in particular are driving from the community today than they did say 10 years ago i would agree i mean i don't really have a position on this um i've been uh i have watched their conversation with interest obviously obviously. But I think one of the things that Canonical has done, as all long-term open source projects do, is think more about sustainability. This is why, for example, things like the Unity desktop went away, even though it was marvelous,

because Canonical need to think about how can we ensure not only that Ubuntu exists today, but ubuntu continues to exist and carries on doing the best it can for the most people and so i think there was something of a sort of a community free for all in the early days and now as it's got a bit more professional and some people have kind of gone well, it doesn't seem as much fun anymore, which is undeniably possibly the case. But on the other hand, it means that I still have a desktop

that I use every day and have done for, when did it first come out? 2004, so 16 years. And I'd rather that was the case than the the whole thing fizzled like so many other projects have done when someone starts saying but i've got bored with working on this and i'm not being paid for it yeah i could definitely see a similarity when we started taking ubuntu more seriously as a product and started using it in production more this attitude shifted a bit and it became less of this fun thing that we're playing around with as a community and more of a serious thing that we depend on. That's the thing. I mean, as you

mentioned, I am older than some of the people in the community and have said, I'm not that old. But I think the other thing is that fiddling about with your desktop is to some extent, at least a, a young person's thing to do. In my experience, it certainly was when I was younger, I used to enjoy things like changing between different distributions and so on. But now my views have radically changed on that to the point where my desktop is the thing which launches applications and applications are what I care about.

So I can get things done, whether they're fun things or work things or whatever. And so the importance of talking about the desktop itself has, in my experience and in my life, fallen by the wayside a bit right i don't know whether this is because i've got old or i've got professional and boring or the linux community has got professional and boring or all the above or possibly yeah a little from column a little from column b little from column c yeah right i mean i think too is you know once he's taken over the cloud right what you're using is more of the package infrastructure, things like that. And the desktop isn't really relevant. Right. Yeah. I mean, it's absolutely something that matters to us. But to most of the market, it's not necessary for what you're deploying on the server. caring about the desktop and people wanting to get involved in maintaining it because it was fun as Stuart said back at the start it was fun like we're breaking new ground and it's

innovative and we're doing new things and we're making the desktop usable by normal human beings as the old strap line used to go but it turns out those things have to be maintained and if you commit to a five-year support term for an lts or a 10-year support term for an lts someone's got to keep cranking out those packages and that gets really boring really fast and so yeah people move on to the next big shiny and canonical have still continued to finance people to support all those LTS releases for all those years. And some of the community people have just moved on because it's just not fun and interesting anymore. I certainly would agree. When you say people move on to the next big shiny, I think people who are involved in maintaining the desktop for fun in their spare time are entitled to go to where the fun is and the fun

honestly is not as you say in maintaining a thing that already exists making sure it stays stable and so on so i think some of the motivated community developers it's not as much fun to work on ubuntu the desktop as it is to work on some other desktop who are doing radical changes of things because they're not supporting millions upon millions of users who don't want their desktop to change every 30 seconds. But if you do want to hack on this stuff, then yeah, go somewhere where they're doing something radical and innovative. That's a really good idea. Right.

Which is, and Ubuntu is not, I mean, when it came out, it was this incredibly big change. The idea of having a sense of design and being design-led in the Linux world was itself a radical reinterpretation of the text. And it brought calm to chaos in a way too, because there was app selection that took place instead of having three mail apps. Yeah, exactly. So having opinions, being prepared to state them and back them up and back them up with money made a really big difference. But now Ubuntu is not the new play thing. It's the old statesman of the Linux world. It's what 95% of people actually out

there using Linux are using. Those people are not interested in hacking on the desktop. It's just the thing that launches their applications. Yeah. Or maybe even more starkly put, the thing that launches their web browser. Minimek, you think maybe the role falls upon us to reach out to a newer generation? Yeah, we shouldn't forget the talking. I see all the arguments, and I'm also one of the older generation, so I normally stay with stock installation. Maybe I change my desktop manager. But we need the younger generation, and these are the hackers.

Like Wimpy mentioned two weeks ago, we need to find new channels and new ways to communicate with them, give them access to the hacking. And yeah, so they have really fun to do the thing and then discover Linux like that. I could see just trying to be there to support their journey and learning and being open to whatever that path might take them. Cause it's probably going to be a lot different than our path was when we were younger because the industry has changed significantly.

As we're recording this, though, the story has updated a bit. Mark Shutterworth announced that now in cooperation with an Ubuntu member and former community council member, I'm going to say Walter Lapchinsky. What do you think? I think you did pretty good. Really? We call him Wexel because it's easy to just call him Wexel. Well, Wexel will be working to restore the community council. And Mark, having considered over the weekend, I think I'll take Wexel's offer to help run the process. Let's

go ahead and call for nominations to the community council. Well, how about that? He also thanks Wexel for the conversation and getting it going. And then he ends with an apology for having dropped the ball. It seems like he really did reflect on this over the weekend. And that seems like a positive development, too. Right. I mean, I can see Big Default, too, if there's been, you know, mixed levels of availability and seemingly excitement around this process. You know, he does note that it's nice. Maybe it is a good idea to have this in place, even if it isn't particularly active. nice, maybe it is a good idea to have this in place, even if it isn't particularly active. So maybe there just needs to be some discussion and clarification here around what the roles should be and, you know, what it really means to be a part of it.

Well, speaking of other positive developments, there's been a lot of positive developments for the workflow development of ButterFS. Now, Wes, I'm going to go and say Yosef Bakik. Well, this is one of the Butterfest developers who I follow. And, you know, we'll have to check after the show because his talk over the summer we'd mentioned before about Butterfest development and use in Facebook, that's out now, which we'll have added to the show notes. Right. Thank you. Yes. Which I did catch a little bit of. It's, you know, it's a file

system talk. So prepare yourself for a lot of excitement. But some fascinating insights into how Facebook uses it. And it looks like after some discussions over the weekend and working out how they want their future development workflow to work, they're on track to improve things quite a bit. Yeah. How about this? All of our patch submissions will be tracked as individual GitHub issues. So we know what needs to be reviewed and what is pending. hub issues. So we know what needs to be reviewed and what is pending. We're also tracking XFS test failures in our FS test tree, and we'll start staging fixes there so we don't step on each

other's toes when adding new tests. I like this line here. Getting a group of developers who have different companies and different bosses on the same page is tricky, but we think this is a step in the right direction. And right, I mean, yeah, there's a lot of interest in different folks helping maintain ButterFS. It's getting serious. It can't be easy. Yeah, and it's good that they recognize as a team they got some workflow stuff to work out. I mean, it's not like everything's solved. Yeah, sometimes you gotta, you know, do a little work on the ways that you work.

That way you can work. That is so wise, Wes. Somebody should put that on a t-shirt, you know? That could be what drives industry for years to come. Oh, I think it will. And that and the next Linux kernel, which Mr. Larble over at Pharonix seems to have been hot on the trot

on a performance regression and even gotten a bit of a conversation back and forth with Mr. Linus on what the heck's going on. I mean, you know he loves benchmarks. And you know, we love a good kernel story. So anytime we have an excuse to cram one in, we do. And like, you know, you got to talk about the kernel.

Of course. You know, it deserves its respect. And last week, Foronix reported that there was a performance regression in Linux 5.9. When you go from like, say, 5.0 and you compare it to 5.9, pretty noticeable performance drop in things like apache and nginx workloads so it would directly impact quite a bit of cloud servers if deployed

not necessarily great and so it seems that he he sent a note to linus and linus took a look at it and kind of gives an explanation on what's going on and i think that's fascinating if you're interested in what the actual real world performance results were, they actually are pretty noticeable. Yeah. Not good. You know, I thought it was interesting that there had been some other sort of tests on, like, a hack bench, other systems that the kernel uses. But it was actually, you know, Michael's benchmarks, because they had tests that Linus at least felt were a little more real-world, that stood out and, you know, gave some more attention to this issue. That's kind of a nice feather in his cap.

Yeah. So here's how Torvalds originally summed up the issue with all of the interesting technical details in an email to Michael following the bisect report. Mr. Payne, take it away. Let's say that a page is locked by some user. Doesn't matter why. It might be IO. It might just be for any number of other reasons that, you know,

I need to get and make sure that this page stays consistent, I need to lock. And while that's happening, a number of other processes come in and want to lock it. What we do, and what we did before that commit to, is to queue them up in the hashed page wait queue. And a flag is set on the page structure to say, this page has waiters. We queue things up in order so that the oldest waiter is first. That is meant to be about fairness, but you'll see later why that didn't actually used to matter. Look at Linus the storyteller here. I love it. Anyway, this thing didn't change fundamentally by that commit that you bisected the performance regression to. Yes, the commit changes that queuing, but not in any fundamental significant

way. The change is incidental to the big change, which is the what happens at wake up. Right. So that's what the big difference here is. It's that wake up behavior. Both before and after, the basic trigger is the same. The process that holds the page lock does an unlock page call. And as it unlocks, it also checks that do we have waiters bit on the page structure and goes off to process that waitlist. And here's the big difference. We used to just wake things up on that waitlist. And that was it. We had various anti herd behavior. So we'd only wake up one exclusive waiter process, i.e. somebody who was doing a lock page. So we wouldn't have all these waiters suddenly wake up.

But basically we just did, oh, we've now released the page, so wake up the waiters. That's the simple end of it. That sounds obvious, but actually it has a rather non-obvious effect. In particular, the page is now unlocked and the waiters have been woken. But they aren't necessarily running. They have become runnable, but particularly under load, it's probably sometime before those waiting processes actually get any CPU time. In the meantime, all the processes that weren't on the wait queue and are runnable are free to come in and take that page lock again before the person that's supposed to take it has actually gotten CPU time.

I see. So I barely follow, but it seems like you have this situation where it's waiting to wake up, it's waiting to wake up, and then there's this wait queue. I don't really follow the technical details of this, but it seems like they're on the case. Well, so it's kind of like this. You're waiting for something. Maybe it's your turn

to head up to the buffet line, if you remember those. Alright, now I'm following. Yeah, right? So you're in your queue, you're ready to go to the buffet line, but it takes some time to actually get from where you're waiting all the way over to where the buffet is. And even though you've now got the magic spoon that you're going to use to ladle all your food from the buffet,

someone else can just come sneak in, and since no one's currently using the buffet in between, well, they'll just take your place. And then there's some like more nuances here where then actually the person would notice that the page had been locked before they got there, you know, the one who was supposed to get it, but then the way the kernel

scheduled is they'd end up back at the end of the wait queue again. So they were at the front, didn't get the resource, had to go all the way to the back. And because of this, that you'd notice these big regressions where suddenly you have these horrible latency spikes. So to round out this analogy, is it basically brought not to the front of the line?

Exactly. That's how you lose out on the pudding, Wes, is you get sent to the back of the line. That's super unfair. So Linus' change was aimed at trying to make this more fair. But unfortunately, that can have bad places for throughput because while you might have a less maximum latency, probably your minimum latency is going to be worse

because it's more fair. And at the same time, this has been unfair for a long time. And so that just means sort of accidentally a lot of the other parts of the kernel have been built around that. And so changing it touches a lot of other subsystems. It's just peculiar why you see problems with Apache, but you also see a bunch of

issues possibly in file system benchmarks. Linus is trying some stuff. He's kind of ideating, playing around with some solutions, looking for feedback. So probably this will be something we see fixed sooner rather than later. Okay. Well, I'm glad they're on it. That's my summary of it. Go get them. So we don't have to.

Linode.com slash unplugged. This is an exclusive offer for Linux Unplugged listeners. Sign up, go into Linode.com slash unplugged and get a $100 60-day credit towards new accounts. Linode is simple cloud infrastructure that you can deploy in seconds. We shared hosting starting around $5 a month, but then they have a bunch of different options going from there with dedicated CPU plans, GPU compute plans. You can use orchestration management if that's your preferred way to deploy systems. And of course, they have a one-click app market. I've been a Linode customer for a couple of years.

And over that couple of years, I've witnessed this company just grow and evolve their product offering in a really clever way. Linode.com slash unplugged. Of course, they have native SSD storage, 40 gigabit connections coming into the machines, industry-leading fast

processors, and a new revamped cloud manager that's built on open source how about that west and of course you get root access to your box but the thing that i really like and i've never really taken advantage of this before is their command line application i just started using it about i don't know it's been i don't know it's less than a year and i find this to be a really easy way to send up files to their object storage and then link them publicly to shut down machines. I get alerting and reporting on what's happening with my system.

So like I saw over the weekend, Alex was working on a self-hosted development machine that we're setting up. We use Linode to host our matrix server right now, and it just works fantastic. Sure does. We also set up multiple accounts so the guys can log in and work on stuff as they need. Wes Payne can go in there and set up a machine if he needs to, just like Alex can. Super simple to set all of that up. And with $100 credit, just go over there, set up a new account, you get $100 for 60 days just by going to lino.com slash unplugged. There's a lot you can do.

I like that, yeah. A lot you can learn, a lot you can mess around with. I know that's what a lot of our listeners are doing with it right now. Learn a lot you can mess around with. I know that's what a lot of our listeners are doing with it right now. It's a great opportunity to support the show and get great cloud hosting for yourself from someone who's been around for a long time. That's the other thing that's really great,

is they're like a deep community member. Just about every event I've ever gone to, they've been a part of. They've had a booth there. They've been giving away swag. Even our humble LinuxFest Northwest. So it's really awesome now that they're a sponsor. Linode.com slash unplugged.

And a big thank you to Linode for sponsoring the Unplugged program. Linode.com slash unplugged. Chris, I signed up for Linode just last week. Oh, yeah? Yeah. And what are you using it for? I am upskilling, getting up to speed on some technologies in the sysadmin space that I haven't really played with since I've been out of that world. That's such a great use. Right? Oh, it's awesome. I hear that from a lot of our listeners that out of that world. That's such a great use.

Right? Oh, it's awesome. I hear that from a lot of our listeners, that they use it for training. It's a great opportunity. Linode.com slash unplugged. Now, Stuart and Popey are specifically joining us today because they had a proposal. Maybe I'll let Popey tell the story uh didn't quite get the traction they were hoping for originally but it's now kind of been opened up to the community to potentially run with

last year the gnome project announced that they were in collaboration with endless they were doing a community engagement challenge and they pre-announced it last year to launch this year. And the goal of it was really to come up with ways that could encourage more people to get involved in open source software. cash as prizes for people who were successful in getting their projects through this community engagement challenge process and ever interested in the idea of making money I looked at this and thought that might be interesting I'm sure me and some pals can come up with some ideas that could get through this and maybe we'll make a little bit of beer money on the side so i mentioned it to my good friend stewart and we brainstormed for a bit about things that might be interesting and we came up

with this idea for something called cabin which i'll let stewart describe but we we went through a process of creating a proposal and writing plan documents and submitting them almost at the last minute when the deadline for the first phase of this project came about but as you said yes we weren't successful in being selected for the first phase. And those projects go through a filtering process for the further phases, which will follow through the rest of this year and into next year. But we were keen for this to not be lost because we put some effort into creating this proposal. And so we thought we'd publish it. And so Stuart published it because he did the lion's share of the work on Cabin and

came up with the original idea. So he published it on his blog. I wonder if, is it correct in my assumption that this feels like it's informed by maybe a longer held philosophy that you've been developing for a while that we need more apps, less distros? I know I've heard both you and Popey talk about that. Is it, there needs to be maybe a better way to get apps onto Linux? Is that what informed this or what was the nature? As you say, I've been developing a philosophy

that we need more apps and fewer distros for years and a whole bunch of other things about how application development should work and the kinds of developers and the kinds of applications I'd like to see on our desktops. And Cabin is an attempt to answer some but not all of those questions. So to give you an example, think about your phone. I would say about half of the applications I use on my phone are single shot apps. They're not big, complicated apps which do a whole bunch of things. They are small, single function things to grab an image and crop it or to grab an image and put a speech bubble with some text on top of it. Little helpful tools. Yes, exactly.

Now, ignoring the command line on the Linux desktop to a first approximation, we have none of them. We do not have tools like that. If you want to take an image and put a speech bubble with some words on top of it, you have to open up GNU Imp, the GIMP, right? Which is a terrible thing. Maybe you want to open up Krita or something like that instead, whatever. But the point is, these are full-fat, fully complicated applications.

And building such things is hard. Using such things is hard using such things is hard and frankly it's one of the reasons why we've got 10 000 applications and android's got 10 million and so from my perspective the issue is not that people don't have ideas for these applications. It's that it's really difficult to build them because our application development story is fragmented. It's complicated. It's full of holy wars going back 20 years that no one even remembers the original reasons for really yeah what i wanted was a way for someone who's not a particularly experienced programmer but this is

someone who's on who's using an lx desktop right so they're already using those desktop they have an idea for a small application that they would like um those of you out there who are linux desktop users and are programmers so especially if you're web um developers and are and programmers you will probably have the experience of people you know asking you to build tiny little applications to do things so i've built applications for people to calculate the angle that they should tilt the handlebars at on their on their new road bike in order to um best fit their frame and i've built uh tiny little applications to count down to the next star wars film that they could put on their home screen. Stuff like this that I think real people have this idea for applications all the time

and they should be able to follow the whole app process through from I have had an idea to other people are now happily using that application. Well, a couple of observations. It seems that, I mean, the reason we call them even apps now is this idea that they're smaller applications. It's not a full application. can actually be quite profitable for some developers. But when I also kind of look at the broader picture of what you're sort of proposing here, I kind of look over at the elementary OS guys and I go, you know who does have apps kind of like that is elementary OS because they have created

a consistent and clear story for the developers on how to create applications for their platform, what they should look like, and then how to distribute them to end users, potentially with the possibility of making a profit. And they've gotten some traction. It's not blowing the doors off and changing open source as we know it, but it could be in a way a sort of a test lab proving out what you're proposing here.

Yes, absolutely. The elementary people have exactly the right idea, in my opinion, and they're doing the right thing because most importantly, what they're attempting to do is create a culture which values this kind of thing. And this is why, in my opinion, they've taken a fairly long step away from calling themselves a Linux distribution, exactly because they're not. They don't want you to be running applications that run on Linux generally on elementary. They don't want you to be running applications that run on Linux generally on elementary. They don't want you to be running elementary apps on other distributions. When you show up and say, hey, your desktop environment, how do I run it on something else?

They'll say, well, suppose like this, but we don't really want you to do that because they want to build a culture where everything fits together. And that's really important. And valuing this kind of application i think is a big thing what cabin attempts to do is slightly different it is like the elementary model in that elementary have nominated the way you should build applications handed down on stone tablets you should use use their libraries, their programming language. This is how you should publish them, so on and so forth. And that's the right thing to do, frankly.

But I think Valor's too hard for the kind of applications, the kind of programmers that Cabin is targeting. It's at least a big ask. Yeah. The elementary people have the disadvantage, by comparison with Cabin, that they would like you, if you want to sit down and build a full-fat, fully complicated application on elementary, they would absolutely welcome that.

And therefore, their development environment and their development libraries and their development process has to be able to encompass that kind of application. Cabin utterly does not. us that kind of application. Cabin utterly does not. Cabin buys itself a bunch of ease of use by intentionally sacrificing breadth of scope. You cannot and would not ever be able to build anything more than these naughty little applications in Cabin. And that's fine. Is there a risk of creating, to put it harshly, trash apps that sort of clutter things up? Is there a risk of that, do you think? Or does that sort of sort itself out?

That depends on your definition of trash. And frankly, yeah, it would probably be a problem if we had a million applications and a bunch of them rubbish. But having a million applications would be a good problem to have on our desktop. At that point, yeah, there's a filtering issue there, which belongs to the app stores, the Snap Store or whatever, about how you filter good applications from bad. And it's a problem that iOS is dealing with,

and that Android is dealing with, and that Windows is dealing with, and it's the web's dealing with it. The reason we haven't had to deal with that problem is because we have no apps to filter that's absolutely true yeah so what cabin attempts to do is provide a programming environment that someone relatively new to programming can understand but without hiding the idea that applications are built with code so it's not a snaps and jigsaw pieces together thing i'm being terribly mean to things like Scratch here, which you can actually build really cool stuff with. But I didn't want to hide the idea

that this is still fundamentally programming. Right? If you're not interested in programming, you shouldn't be trying to build applications. You can't hide that away forever. And anytime you do try and hide it away, what happens is people discover it unpleasantly halfway through and then suddenly don't understand things. So you don't want to pull the rug out from under people when they're

attempting to develop new applications but cabin apps are built so that they're essentially compiled into python and gtk and it doesn't hide that so the idea is if you ever get to the point where you're hitting the limit of what could be built in a cabin application, you've then learned enough about programming that you can take a peek under the covers and go, ah, what this is is a simplified interface to a more powerful thing. So now I'm ready to learn the more powerful thing. Yeah, and that's often what gives people a chance to start and begin to build something. So Stuart, you and Popey have put this out to the world. What would success in a scenario like this kind of look like? Because realistically, it must be some form of adoption or inspiration. I would like someone to pick it up and prototype

what we envisaged it to look like with some guidance you know we'd be happy to help but you know we're both busy and it's not something we could commit full time to but we want we didn't want the the proposal to die in um a file we uploaded to a contest that we didn't succeed at we didn't want it to just die there so we want we thought it had worth and so success would be if someone you know started hacking on something if they had a bit of time and they had some time to hack on it and you know get it bootstrapped you know with guidance we'd love to like help as a normal open source project like anything else and you know future success would be people created little apps with it that would be a measure of

success for sure sure yeah absolutely there are a half dozen ideas that cabin embodies in my opinion one of them is um the thing about uh helping people with ideas transition them into applications without throwing the world of programming at them. A second one is that it's a stepping stone to more complicated things when you're ready for them. A third one is the idea, if you look at the cabin specification, you'll see that you've got your application and your code on show at all times. Changing one changes the other in lockstep, in concert, which is an idea from Brett Victor. And I think that's a really important way to develop things. And people don't currently have it.

Fourth idea, building an application is considerably more than just writing the source code. Making sure that your application has a description, an icon, a tagline, a title, that it's packaged up and uploaded to a software store, that it's available in that software store and you can push new versions to it, that the community can gather around the application and feed feedback back to you. All of that stuff is just as important as the source code and it's shamefully slighted by most of our development environments at the moment unless you're using some kind of immensely complicated uh full bore ide like android studio is or xcode is q creator does some of this but not a lot of it and a lot of that is because those environments

have to be able to cope with, A, building the most complicated applications on Earth, including themselves, and B, to be swappable out in the kinds of applications they can build. Cabin's allowed you to say, this is the kind of application it builds, this is where it publishes it,

we make that easy, and you can't switch it out for publishing different kinds of applications. And from my point of view, if someone were to be inspired by those ideas, but build something quite different to cabin, but it still met all those requirements, I consider that very successful. Obviously, the way I think you should meet those requirements is cabin. Why respect it? Right. Of course. But that would be a benchmark of success, right? The idea getting propagated. Yes. The idea that people should be able to build an application, get it into the hands of other users who can then enjoy it, build a community around it, feedback to you about it,

the community around it, feed back to you about it, learn maybe how to start programming applications themselves. That to me is a thing that we should be aiming for on our desktop. We're not currently doing it. So if someone is inspired to do that, I'm happy. I also think another key aspect of this is the concept of redefining what is an application worth time to develop and deploy on Linux. And I look at my usage on and off over the good old years, and I used to have plasmoids loaded up on my Plasma desktop, which would show me different information and pull in different feeds, which are essentially tiny little QML applications. And same with GK Realm.

And of course, there was other systems before then where I've had small little usefulML applications. And same with GK Realm. And of course, there was other systems before than where I've had small little useful contained applications. And there's just never been something that's broader and higher level for it. And I really think it does have some interesting ideas that I'd love to maybe even just see people discuss

at the upcoming Linux Application Summit. That could be a good place for something like this to get talked about, you know. So is there anything really further after this, guys? Or or is this kind of it's out there for the world now and people are welcome to come see it the goal was to get the ideas out there because we as as uh as probably says we spent a while putting this together and it seemed a shame to just have it quietly die on a hard drive somewhere could make for for a good talk. Yeah, if someone wanted to talk about it. But yeah, it's not something I could afford

to devote a bunch of my free time to because I've already got a zillion free time projects. If someone wants to come in with a checkbook and pay me to build it, I'll listen because then it's paid time, which I've got more of. But yeah, I think from my point of view, it's out there. And if someone were to want to pick it up, I'm sure we'd be happy to talk to them about it.

And I suspect Alan would agree. 100%, yes. Yeah, and I just like us thinking about this stuff. If nothing else, it's a good headspace for us to be in. Thanks for coming on on your Sunday, gentlemen, and chatting with us about it. It's an interesting idea, and of course we'll have links to it all in the show notes at linuxunplugged.com

slash 371. 371. Nice. I want to say thank you to CloudGuru for sponsoring this episode of Unplugged. You know, we talk about one of the best ways to learn is by actually doing. The CloudGuru now includes Cloud Playground for Azure, AWS, or Google Cloud Platform Sandbox. As you get on there, you log in, you have a system, and it's on their dime, not yours.

I like that. Mm-hmm, mm-hmm. And it's a great way to learn by actually doing it. Of course, then there's a bunch of other resources you can fall back on as well. It's just one click to a fast, fresh, disposable cloud environment at your fingertips. There's really no better way to learn, is there? And then you have the confidence to actually go test or even better, do the job in the real world.

Just go to cloudguru.com. And thanks to CloudGuru for sponsoring the Unplugged program. How about some feedback, Mr. Payne? So last week we had our bug-a-thon after the show and pretty happy with the way it went, some lessons learned, but overall, we got great tips on how to make a bug report

that was useful for developers. That's awesome. I earned my first Fedora badge. Sure did. That's pretty cool. And then also there was just a lot of good information about just general troubleshooting throughout. And there was a couple of times we thought we had found ourselves quite the bug,

but then, you know, reproducing the process and working through it, we kind of worked it out. Sometimes it's just UI inconsistencies. It was really a lot of fun. It was. And learned some stuff. So I decided to make it available to our Unplugged Core contributors. You just have to go sign in.

So if you go to unpluggedcore.com, there's a little sign-in button there. They make it really small because it's the sign-up page. But if you click the sign-in button, then you go to your downloads area. The whole thing's available for download. And I actually think it's worthwhile, even if you're not a Fedora user or interested in ButterFS, to hear the guys talk about what they need in a bug report and the processes you've got to go through and some of the little nuances that Carl and Neil went into particularly when we actually came to the point of, okay, I'm about to create this bug report. Here's my questions.

Oh, well, for that, you need to know little edge cases. You shined a lot of light on the intricacies of keeping track of all this stuff and how you actually make it actionable. And test methodology. So it's available for the core contributors right now. And I think maybe in the future, we'll probably release it in the main feed. Like if we get sick or something, need a week off.

But in the meantime, it's available for the core contributors right now at unpluggedcore.com. And the other thing that's really cool about the Bugathon is we're going to keep the channel going in Matrix. So developers are welcome to join us in there to seek people to help them test their projects. If that is a benefit to your project, if you are understaffed when it comes to testing,

we'll have people that are able and willing in our Bugathon Matrix room. But additionally, what was kind of neat is even people who couldn't attend live ended up doing it on their own. So I've been hearing from people throughout the week about them testing it. And that I really like. I didn't really think about that, but that's a neat knock-on effect, is even the download audience who didn't attend live, could never make it live, can still help us with this. And so that's something I think I want to definitely take advantage of in the future for the next project we do this for.

I think we're going to do it again. Can I also just say, it seems like Fedora 33 is shaping up to be quite a release. I mean, hardly any bugs at all. So far. I mean, we really, you know, we really put it through its paces. I did a live conversion to ButterFS. Oh, that was fun. Which was successful, but did teach us a couple of things that we'll try to cover in our review in the future.

A couple of going indie updates, I figure. Keep everybody in the loop on what's going on with JB now. You'll see a new logo rolling out to the shows and to the website, a little updated logo. I don't know if this is our permanent logo or if this is like get us to Linux Fest. I haven't really decided yet. I'm sitting with it and I kind of just like the simplicity of it. So if you haven't seen it yet, it'll be showing up like in the MP3 art and on the feeds and on the website. So you can check that out. Lots of work going on behind the scenes to build up the pipeline for the membership feeds. Thank you specifically to

you and Drew for working super hard on that. I really appreciate it. So that's been getting rolled out. We're still making tweaks to some of that and working out a pipeline. You know, a lot of times we start manual processes. A lot of times they fall on Wes's shoulders. So I really appreciate you doing the picking up the slack there. And then we develop them into an automated process. Exactly. You know, so first the human, the human process, and then it's like, you learn it, right? You understand what has to be done and then you just replace it

with bits of automation. Right. Otherwise we wouldn't, I mean, we've got to answer a lot of questions of how are we actually doing this? Yeah. I really love that you said Wes and me when really my whole thing with it is,

hey, Wes, are you handling the ad free feed this week? That's like the extent of it. Yeah. You can find it on our NextCloud. Yeah, go get the file. You do it. Go here, yeah. And so the other thing I want to be clear about that is it's limited ads.

There is a couple of short ads in there that are contractually obligated, but we're going to start cutting out the membership plug itself from that version of the show, and future ad contracts won't be included in that show uh right now it's it's it's probably technically slightly a shorter show too i'd imagine i guess but it's uh it's coming along and it's really been a team effort and it's been also kind of like i think we're all kind of motivated because there's been a great community response it's been really awesome to see um and then we're also still taking your ideas for linux prepper which is in the matrix. Be prepared in layers

and what are those layers? What if your cloud service goes out? What if your main rig goes out? Just can't boot that day. What if you got a bug out? How do you prep as a Linux user? Join us in the new Matrix room and check for the Linux Prepper room in particular and

we'll get a conversation that'll eventually turn into show content. See, if we were thinking ahead, we would have had a conversation that will eventually turn into show content. See, if we were thinking ahead, we would have had a Linux-powered smoke filtration system set up already. No kidding. No kidding. All right. So, Five Creative, you think that's how you say it? Five?

Five Creative? Writes in about dev IDEs. He says, hey, guys, long-time listener, first-time collaborator. Dirty collaborator. I thought it might be a cool thing to do something similar to the Podcatcher Playoff, but for IDEs and editors. Oh!

Nobody has an opinion on editors, right? Of course not. He says, I've been a long-time user of Komodo IDE, namely for its integrated SSH support. I've gone all into Linux, and I no longer require that feature with built-in SSH support in the OS. I just received my Pinebook Pro and thought it would be great to ask around for a lightweight editor.

I know of many, but which ones do people love? Thanks for the hard work you put into the podcast. And he also says that Ubuntu podcast is great. It is. Doesn't actually say that, but we do think that. And that Bad Voltage podcast, too. He says that in there, too. He just put that in there. I shouldn't give him edit access.

I know. Weird, right? So does anybody have a text editor that's lightweight? You know, we often on this show will joke about Nano being the one and true text editor, but there's others out there. Does anybody have one they want to give a little love to? I use four of them on a daily basis. Wait, wait.

Four. All right. Walk me through this. I'm going to brace myself here. Okay. So I use Vim primarily. As one does. As one does because the church of Vim is awesome. But sometimes I need a graphical text editor because, you know, sometimes I like to use a scroll wheel of my mouse, you know, use like the standard keyboard shortcuts and all that. Like a gentleman. So I use mouse pad because that's installed on all my systems.

On my workstation i have vs code and and another one i've been playing around with called cocoon which is like vim but different cocoon okay and don't think i didn't just notice that you just cash drop that you're an xfce user i totally caught that uh no actually i'm not using a xfce really oh but you're using mousepad what kind of animal are you a savage animal that. That's the kind. Makes them match. And Mousepad is actually a pretty nice lightweight GUI editor. Yeah, it's pretty nice lightweight. It does not actually pull in the XFCE dependencies.

So it's actually pretty well standalone. All right. So I don't see why you wouldn't use Mousepad outside of XFCE. That's fair. Yeah, definitely one of the Chromium-based ones isn't going to be a list, isn't going to be a candidate for this one, is it? Anybody else have a... Oh, that graphical one is tricky.

Well, you know, some folks really like Sublime text. You know, there's a stalwart contingent there. Me, for a start. Yeah, me too. Another vote for Sublime. Okay. Sublime's wonderful, although I've got terribly unassuming requirements

for the text editor I spend half my life in. It basically has to start up quickly. It has to have multiple cursors, syntax highlighting, and it must never, ever, ever, ever lose a file. Even files I've never saved, never given a name to, if I start typing into a file, and then I yank the power socket out of the back of my machine and then plug it back in again, it will still be there in Sublime when I start typing into a file and then I yank the power socket out of the back of my machine and then plug it back in again, it will still be there in Sublime when I start up. And that's why I can't switch away to it,

switch away to anything else because nothing else will guarantee that. So I'm now used to doing that. Half the files I've got open in Sublime, I've got, I think, six Sublime windows open at the moment, each of which have got about 10 or 15 tabs in them. And half of those have never been saved. And they're still there across reboots.

They're actually across upgrading the operating system. It's magic. Magic Sublime Voodoo. I love it. It's the best. Just use Sublime for everything. It's brilliant. I have to say, I wrote years of consulting notes in Sublime,

and it never, ever failed me. And it always relaunched with my notes right where I left them, even massive multi-month documents. I do not use it anymore because I've just switched over to VS Code. I'm one of those people. But I love it, and it is definitely of the graphical, nice high-end text editors. It's definitely one of those people, but I love it. And it is definitely of the graphical, nice, high-end text editors. It's definitely one of the quickest.

It's not Chromium-based, and it's pretty fast. It's been around for quite some time, too. It's not going anywhere. Battle-tested. Casual Chemist, which is one of my favorite names, writes in. This is funny because you can tell where people are at in the back catalog because when they get to the next Cloud episode we did, they go, holy crap, I've got to write these guys.

What are they doing? Don't worry, we're working on it. But he says, hello, listening for a few years, although I'm a few months behind. Yep, as he told you. He said, I've just heard of 362 on Linux Unplugged, the hidden cost of NextCloud. Yep, that's totally fair casual chemists.

A lot of people are still getting there. I'm also up to episode 21 of Self Hosted. He loves that show, apparently. Awesome. Good job. A new episode just came out where I compare sync thing to episode 21 of Self Hosted. He loves that show, apparently. Awesome. Good job. A new episode just came out where I compare Sync Thing to NextCloud, which was something we were going to do in the show, but we actually ended up cutting it for time,

and then I had more experience with it, and so I'm trying to convince my co-host Alex to give Sync Thing a go. You're working there. I'm working it. Yeah, I'm working it. So Casual Chemist goes on. He says, I was surprised that you all pay $350 a month. That's $350 a month for NextCloud storage. I was surprised, but then when you broke it says, I was surprised that y'all pay $350 a month, that's $350, a month for NextCloud storage.

I was surprised, but then when you broke it down, I was less shocked. However, have you considered using Backblaze or Wasabi instead of DigitalOcean's object storage? They're both S3 compatible. That's true. And then he gives the price currently of Backblaze. And he says, by the way, I want to give a shout out to my favorite backup solution, the love of my life. Boy, that's somebody who loves a backup solution. Well, when it has all your data in its hands.

A Borg backup, which we've talked a little bit on the show before. And it just recently stumbled across a project called Borgomatic, which uses some YAML config to automate Borg backups. There you go. What's one more YAML file? That's a winner. Yeah, so we've kind of taken a different approach. This has been a common suggestion. It's just essentially swapping out the backend cloud storage. We're kind of thinking about doing more of a primary storage on-site here at the studio kind of thing, because

we do have a NAS here, and then keeping the raw editing files up on the cloud. So that way they're fast and easy to move around and quick for remote host upload and quick for remote editors to download. But keeping that tighter than we have in the past, you know, maybe keeping it close to like 50 gigs or something, right? Right. Just managing the hot stuff that, you know, needs to be up there and then anything else, we can move it to long-term storage. Yeah. And we'll probably in this process, we're probably going to migrate the server over to the node. We're probably going to move the domain over and then we're going to have to migrate a bunch of the data down to the studio. He is.

Yeah. And then we might take advantage, maybe, maybe not, of federating the two NextCloud instances. Although I'm kind of tempted to keep them separate. You know, one's like an archive and one is current working projects. And we're using them quite differently. Yeah, exactly. One is so we can retrieve something as an asset that we might need a year down the road. And one of them is for collaborating with co-hosts and remote hosts and editors.

They're kind of two different systems anyways. Right, and it's just going to be pulling in the one direction. So we're going to kind of, we're thinking, cut it down that line. And that's where the bulk of the storage is going to be now is at the studio, which means maybe, you know, maybe I end up needing to buy disks sooner for the NAS down here than I planned. But that still will be a little bit

out. We'll be really, you know, getting our use out of the old large server. I know. And it has survived yet another summer where there's been no air conditioning in that garage. And I'm not joking, this year, the vent for the water heater broke.

And so the water heater was just venting into the garage. Right at the server. Kind of actually on the same wall as the server, at least. But, you know, it keeps static electricity down. So, I mean, that server not only is put up with like 90 degree. I mean, it's hot in there. And then I pull a car in there that's been driving. Right. So it's got a hot engine and hot metal surface, and then I park it in there. And that's got to raise the ambient temperature. And then the hot water heater, which we don't use

a lot of hot water at the studio, but it still keeps it heated up, right? So that's venting. I mean, it's unbelievable that what that server has survived. You know, we do sell colo here, but you probably don't want it. Yeah, Jupyter Broadcasting's discount hosting. Yeah, our prices are so low. Well, and not to mention the fact that we abused the hell out of it for fun and profit. Yes, that's true, too. Live on the air. Poor, poor server. It really has put up with a lot. All right, let's talk about Cozy. Cozy is a modern audiobook player for your GNU slash Linux desktop. And it's designed to make it super comfy, comfskies and cozy to browse your library. It's sort of like Plex for audiobooks, but it's a local application. Looks pretty sharp

too. It does. I mean, if you're of the GTK persuasion, which I often am, even on my Plasma desktop, I still like GTK app design. You know what? Sue me. Simple, clean. I'm that guy. But it has sleep timers, which is my favorite for audiobooks. And yes, it remembers your playback position. It has offline mode, of course, so you can keep your books locally.

And then there's other things out there too to help with that. I think it's called Open Audible. Are you familiar with that? Have we talked about that before? No, I don't think so. Yeah. I mean, I might have to do a little live Googling on the show, but if you're an Audible customer, check out Open Audible. And then you could use something like Cozy to import all of those. So it's kind of a nice combination. It seems like it's not biting too much off either. You know, it's not this like crazy, complicated ebook managing monster. It's really just like a clean way to read your ebooks and what they have done. You know, it's been well integrated. It's got, you know, media player controls on your desktop. If your

desktop has that, it seems very usable. We'll have a link to that in the show notes. And then we already got a whole bunch of really good, nice picks. Can I, uh, can I give a book suggestion for somebody to listen to on Cozy? Totally. I'd recommend going to check out The Strain. It's kind of a vampire book written by Guillermo del Toro and Chuck Hogan, and it is narrated by none other than Ron Perlman. You know who I've been trying to find books narrated by?

I'm such a dork. Captain Janeway. Kate Mongrew. She's got such a cool voice. Yeah, good voice. You get a good book by her. I found a Voyager book. So that kind of worked. But yeah. Yeah. Good pick. All right. Well, we'll have a link to that in the show notes and we will be back to our regular live time next week, which is noon Pacific, 3 p.m. Eastern. Eastern people time. See you next week. Same bad time, same bad station. And of course, everything we talked

about is at linuxunplugged.com 371. Check out Popey on Ubuntu Podcast. Check out Stuart on Bad Voltage. Thanks to both of you for making it on the show today. You're welcome. You're welcome. Let's see, what about you, Wes? Where can they find you? They can find me at Wes Payne.

I knew it. I thought you might be on Twitter. Yeah. What about you, Drew? Did you do it? Did you do the Twitter? Oh, yeah. I'm on Twitter, at Drew ofne. I knew it. I thought you might be on Twitter. Yeah. What about you, Drew? Did you do it? Did you do the Twitter? Oh, yeah, I'm on Twitter, at Drew of Doom. I did it.

I'm at Chris Lass. The show did it, too. At Linux Unplugged, I think. I think. At Jupiter Signal for the Network. Pretty sure on that one. All right, thanks so much for tuning in to this week's episode.

Sorry we weren't live to see you, but we will see you next week. Maybe we'll see you on Tuesday. Get off my ship. That's my Janeway. Oh, Hadiyah and I watched The Hunt for Red October this weekend. It was the first time she had seen it. It kind of introduced me to a couple of the classics. You know, it didn't actually hold up quite as much for me.

There's always that risk. I was like, oh, good movie. But I wasn't like, before I was like, all right, well, let's get everything ready. We'll get our popcorn. We'll get our wine. Dim the lights. Yep, yep. We're like, let's settle in and prepare.

This is going to be a real, you know, like a guy movie from the 90s. Let's do this. And we put it on. I'm like, you know, it's not bad. But you're watching it like this just doesn't really hold up to scrutiny. Anyways, I don't, it's still fun. Well, and wasn't that Alec Baldwin as Jack Ryan in that one? Yeah, as Jack Ryan, the CIA analyst.

Yep. And there's some stuff he does. He's just kind of like, oh, really? Yeah. But still fun. It's a good romp. And then tonight we kick off Star Trek, the original series. Oh.

Oh.