LINUX Unplugged - Episode 167: Livepatch Bait & Switch | LUP 167

Episode Date: October 19, 2016

Canonical is not first to the live patching game, but they could have the best take on it. VeraCrypt, the successor to TrueCrypt, audit results are out & KDE shares their long term plans for the Plasma Desktop.Then we bust some brewing Linux FUD and misconceptions & ponder the role of Free Software in a world that doesn't care.

All of my thanks. This saved the show today. Otherwise, we'd be still waiting to get started. We'd be later than we already are. No way. This is my best open source feature of the day. G-Edit, I didn't even realize this, has this brilliant, brilliant feature that allows you to cancel a huge, huge file opening. So I was, I accidentally, we're going to play some clips from Brian Control.

We're going to bust some myths. And I accidentally spelled the.mkv file.mk. That's not the same thing. No. Everybody knows that's the Mortal Kombat file and that's not what you want. So I double clicked it and gedit starts to open up because of course gedit is my default editor for Mortal Kombat files.

Right. Naturally. And it... My whole system just kind of goes... I'm like, what happens? Everything slows down. G-Edit sort of slowly loads up. I'm like, what's happening? And then a little blue window of hope slides down.

And it shows a loading bar with a cancel button. And so, yeah, I was able to click this. And somehow, not only did they manage to put in a cancel button, but they're properly threading the application so it actually registered. Exactly. That's where it's like, finally. Yeah, so it canceled, and boom, my system came back immediately instead of opening up a multi-megabyte.

These are the things we want to be able to tell Windows users and be like, this is why you need Linux. I was like, thank you, GN. Thank you for being awesome in subtle ways. That, today, was my favorite open source feature. This is Linux Unplugged, episode 167 for October 18th, 2016. Welcome to Linux Unplugged, your weekly Linux talk show that's busting myths and thud this week. My name is Chris. My name is Wes.

El Wes. This show, I'll be honest with you, if I was new to podcasting, I'd have butterflies in my tummy about how big the show is this week. You just wouldn't know what to do with yourself. Now that I have my good friend alcohol, no, I'm kidding. Actually, it's just a really great show. And I can't wait for us to get into it. We have legitimately multiple breaking news stories at the top

of the show that we'll get into. So some great project updates. We've got a really great picture from the KDE project about their long-term vision for the Plasma desktop. Then we're going to actually get into some fascinating kernel discussion. Yeah, I know. I know. What? That is a thing. And it's going to be really interesting because we'll segue into some recent criticism that Linux has been coming under that sounds really reasonable and would probably make you think, if you heard this story, you'd probably think, oh shit, what is Linux doing?

This is embarrassing. So we're going to bust these myths in this week's episode. Then, at the end of the show, we're going to get real with you guys. Let's talk about the role of free software in a world that just doesn't care. People just don't care. Microsoft, Apple, they're doing good enough. These are the names people know. Let them control the computing platform.

Why bother with it? It seems to be going fine. They know what they're doing. My corporate masters know what's best. Why should I care? So we're going to talk about the role of free software in a world that, at least for the most part, seems to have that opinion. So before we get started, there's one order of business that we must attend to, and that's welcoming in our virtual log.

Time-appropriate greetings, Mumble Room. Hello. Time-appropriate greetings. Hello. Oh, I love that sound. It is. It's almost like music, except for not at all like music,

because it's not music. No, better than music. That's our Mumble Room. So let's start with, are you guys ready for me to just, should I just get right into all of like the, the quote unquote breaking news? You know, I have to do it. I can't, I can't stop myself.

This is CNN breaking news. Oh man, it feels good every time I have an excuse to play that. I got to admit it. So let's start with the lighter of the breaking news and we'll do a little follow up and then we'll get into some of the stuff that I'm really jazzed about. But we've got, the first story is just sort of the breaking news, and we'll do a little follow-up, and then we'll get into some of the stuff that I'm really jazzed about. But we've got, the first story is just sort of the lighter one. We have the name for Ubuntu 17.04. It's going to be dubbed, you know what it is, Wes?

Zesty Zapus. Oh, you do? Look at you. Yep. And it's going to launch on April 2017. That's coming from- Have you looked at the Zapus? It's actually pretty darn cute.

Oh, is it? I have. The pictures I saw online made it look like a disgusting rat. Well, yeah, but it's more like a cute little mouse-like creature. No, that's a squirrel. Oh, you're right. I don't know. I guess we have to wait until 2017 to find out.

Well, I'm sure they'll go for the cuter one. That's for sure. So that's sort of the lighter of the breaking news. What do you think of that name? Z's probably pretty hard. You know what? I mean, there's a lot of choices. There's fun things like zebras.

That's the obvious choice, right? I mean, everyone thinks you're going to go for zebra, so they don't want to seem obvious. They've had a lot of kind of good, interesting selections before. I actually think Canonical has broke release naming to a point where I just don't care. Like when you, you know, they got these crazy names and Fedora tried to go their own crazy one with the meaty hot dog or whatever it was for a while. But this, this, this shenanigans that's been going on now since the 1204 series or whatever. I mean, for a little while, like they made them sound like they were sturdy and sound. But now they're just like.

Now it's just an animal. Now they're just cray cray. And I think Mark's just delighting in making us all say these words. I honestly believe that's what's happening. And so now I'm like, yeah, this is broken for me. Thankfully they have the nice little numbers that everyone uses anyway. So I always have to like double check and be like, wow, which one? Because it's like on the mirrors, it's always by the name.

And I'm like, uh. Yeah. Oh, yeah. Absolutely. Especially after it's been a couple of years. I just can't. Mamaroon, what do you think of the name? Anybody have thoughts on naming in general? Do you care at all? Not really.

Yeah. Fair enough. Yeah, I don't. You're just going to need something to put on the tag. The number is what matters. It's the number for the version that really matters. Okay. Give the room a moment to do any addressing or criticisms of our – while we're on the Ubuntu topic of our review in episode 439 of the Linux Action Show that came out on Sunday. We reviewed Ubuntu 16.10.

Did you get a chance to see it? Yes, I did. Do you have any criticisms or follow-up or any thoughts about our review? You know, I'm interested. Honestly, I think you guys are the ones reading the release notes and kind of the hype after the immediacy of the release. It really seems like they're coming hard on the cloud data center scale. Those are the things that I saw that stood out to me that I was excited for. I like LexD. I like that kind of stuff.

So I'm interested. And I feel like the desktop, I didn't hear as much about, and I haven't tried it myself yet. So I'm kind of waiting to see. Both of your opinions are interesting, but I don't know as much about and I haven't tried it myself yet. So I'm kind of waiting to see. Both of your opinions are interesting, but I don't know where I fall yet. Yeah, I think a lot of people in the audience are waiting. Anybody in the mom room by chance have 1610? I'm going to bet no, but anybody have it installed? Yep.

Oh, okay. One guy. Excellent. All right. How's your experience been so far? Okay, two people. All right. Okay. First guy, how has your experience been so far? Okay, two people. All right. Okay.

First guy, how has your experience been so far? It's been fine. Some things are broken, but nothing major. Okay. What do you say some things like? I had some problems with the video monitors, but it's an NVIDIA issue. And I had it with 16.04, and I had it before that with 15.10. So this is like Linux release problems, yeah. Who's number two?

TechMav. Mr. Mav, how's your experience been so far? For the most part, pretty smooth. I actually have it on my C720, and I also have it on my desktop right now. I've actually switched back to Ubuntu from Arch for a little bit for stability's sake and just to streamline my workflow. Chilling out for a bit. And so far it's working well.

I will say 4.8 and in a bunch of user lands, that's pretty sweet. Yeah, yeah. 4.8.0, though, which we'll get more on in a minute. But let's go to our next breaking news story. I can't help myself. I'm sorry. That's too much fun. It is.

You've watched a lot of CNN. I've watched a lot of news recently for actually a few years. This story is probably the most news relevant, although not the only breaking news story. But probably one we all care a lot about. Are you familiar with VeriCrypt? That's the forked TrueCrypt, right? Right. When the project current maintainer, the old maintainer said, no, they probably shouldn't use it anymore.

Yeah, I'm out. VeriCrypt is a fork of TrueCrypt. And one of the best things about VeriCrypt is it's able to mount your old TrueCrypt volumes. Oh, right, because they – right. Yeah, that's kind of a big deal. So VeriCrypt has been undergoing an audit. And today, as we record just a few hours ago, the public release of the results has landed. And here's the quick and dirty about VeriCrypt.

VeriCrypt 1.18 and its bootloaders were evaluated. This release included a number of new features including non-Western developed encryption options and a bootloader that supports UEFI. Oh, well, thank goodness. Freaking UEFI. that supports UEFI. Oh, well, thank goodness. Freaking UEFI. Quick Labs, the people that did the audit,

found eight critical vulnerabilities, three medium vulnerabilities, and 15 low or informational vulnerabilities slash concerns. Now, in the security auditing industry, that's like best practice stuff and consider this or could lead to... So the public disclosure of these vulnerabilities coincides with the release of VeriCrypt 1.19,

which fixes the vast majority of the high priority concerns. So if you are using VeriCrypt and you get on version 1.19, you're pretty much protected from all of this stuff. Yeah, that's nice. Yeah, really nice. And I think, I don't think we should let that detail pass by. Right. The responsiveness of the project to the audit, right?

That's the other half of the benefit of having an audit. Absolutely. It shows a strong project with an intent to keep their end users protected and it's pulling on the core strengths of open source. Those are three things behind a security product like this that matter a lot to me. So version 1.19, which fixes the vast majority of these high-priority concerns, is out. Some of these issues have not been fixed due to high complexity for the proposed fixes. But workarounds have been presented in documentation for VeriCryptoReview.

And here's another takeaway I have for you. If you use the ghost cipher for encryption, its implementation was fundamentally unsafe. And they're ripping it out. So they're going to leave in the ability to decrypt it for now. But new volumes cannot be created with this cipher. So upgrade to VeriCrypt

1.19 and move your data to a new volume if you use the ghost cipher. You know, I'll be interested to see kind of the further reviews of this and like once 1.19 is out, what people will see as they say it because, you know, some of these, I'm not a cryptographer,

some of this is like, you know, you kind of have to interpret. It's a little disappointing to see that the ghost stuff as well as... It's ghost, not ghost. I'm not sure. Oh, no, you're right. There's no H.

I'm just being a spaz. Thanks. It is nice to see a lot of the issues are, you know, a lot of things they're fixing here. Like a lot of issues pointed out are in the bootloader. Those ones I'm less concerned with. So we'll see how they continue to be responsive, see how much of the critical issues they fix. We'll see. That was my takeaway, too, is the bootloader was the source of a lot of the issues.

And that's not functionality I used from TrueCrypt and not functionality I would use from VeriCrypt. And I don't know the long-term sustainability of this project, but it is nice to see active development. I do think cross-platform encryption is a worthy goal. Personally, like Lux or other Linux-based encryption tools meet my needs, so I haven't used

TrueCrypt or VeriCrypt for many years, but it's a project, or at least the idea of which I would support. I did the same thing. I moved to Lux after the truecrypt back. I just said, done with that, going with the built-in solution. Honestly, Lux on an external USB volume has worked

remarkably well for me. Lux has essentially become the default when you check encrypt my home folder on installers. You're almost always using Lux. So it's just kind of been the one i've i've landed on uh before we go on because this is some some pretty cool stuff and if anybody has any thoughts on it i don't uh i don't um want to miss your input so since we have uh several new people in the mumble room aka the virtual lug today i just want to let you guys know that if you want to get my attention, just tag me in the chat room with mum, M-U-S, so

do like Chris Lass, mum, colon, and then what it is that you'd like to get in on the show. And I will call on you as we go along. Just that way. Keeps things organized. Yeah. Yeah. So that's because I know we got a good mix of long-termers and newbies in there. So that's – you know what?

Good on the Veracrit project for not only working with this audit but again, shipping 1.19. And this is how this works, right? The auditors find issues. They work with the project on a like a gag order type basis to give them a certain amount of time typically to fix up the problems. a gag order type basis to give them a certain amount of time typically to fix up the problems. But if you watch TechSnap, you know that sometimes this goes differently and the projects or the companies don't meet the deadline set by the auditors and then the auditors go public while the software isn't fixed. And it's a real catch-22 because it puts a lot of pressure on the developers to actually deliver a patch,

but it also gives attackers a heads-up to start going after end users that are unpatched. And the fundamental issue here is even once you release an update, there's a massive lag time often for users to actually— Users actually get all those updates, right? So that's still an issue even here. There are guaranteed way more people on VeriCrypt 1.18 than there are now on 1.19. So there's a lot of users that are exposed. So there's a window of time that attackers can take advantage of this. I will say looking at this, the PDF they made as a report of this, it's pretty readable.

So it might be something fun to check out even if you're not super interested in cryptography. Nice. Nice. That'll be linked. So if you're not super interested in cryptography. Nice, nice. That'll be linked. So if you go to the story in the show notes, is that linked? Yeah, there it is right there at the bottom of the article. Cool. Good stuff, good stuff. And thanks to the VeriCrypt project for taking on TrueCrypt

because that felt like such a loss there for a little bit. It did. So it was really cool. All right, Wes, I want to talk about something that's new and great from Linux Academy, our first sponsor this week on the Unplugged program. Go to linuxacademy.com slash unplugged and take advantage of their new seven-day free trial. Oh! Now, I've been a member of Linux Academy for a long time, so I can tell you it's worth it.

But this gives you a chance to try it out yourself. Kick it. See what happens. Poke it. There's zero risk. Prod at it. Literally zero risk. Bang on it. See what Linux Academy is all about. Take a look at their courseware. Take a. There's zero risk. Prod at it. Literally zero risk. Bang on it.

See what Linux Academy is all about. Take a look at their courseware. Take a look at their feature set. You'll experience something I think is unique because Linux Academy itself is unique. They're built by people who are really passionate about Linux. People who were developers, educators, system administrators, Linux enthusiasts. They came together and they created the Linux Academy platform. And I'm not making this up

because I spoke to the owner of Linux Academy. I mean, we've known each other for years now, but even before they became a sponsor, we had an email thread about how he watched JB, about doing podcasts, and things like that. There's a great synergy. It was one of these situations where he and I kind of landed on a conversation thread where we got very honest.

And I said, God, you know, maybe I should have thought about doing this because it's really brilliant. It's a great way to advocate and promote the Linux platform while delivering in a content-driven way because all of this is with content. They have labs and self-paced courses that you can go through. They have course schedulers that work with how much time you have available. They have learning paths that give you content planned by instructors for specific career tracks. I think it's also a great business model

just to show, you know, it's like they're working a lot with all these open source tools and platforms. And yet they're also providing a very good service that, you know, it's affordable. You'll get a lot out of it. I think it's clever too, like that seven day free trial.

Think about how much you can learn in seven days. You're going to get hooked and you're going to want more. And now they have external profiles, too. So if you want to just sort of prove to either your employer or a potential employer all of the courseware you've taken, they've got it. They've got external profiles. They have a great community stacked full of Jupyter Broadcasting members. And check out their new quick starts.

About two-hour courses. You get in there, and it's like – actually, it's about two hours, and it can be up to like six or eight different courses that you take in that time. So it's a really great compact way to get a burst of information and knowledge right into your brain organ. Bang them. LinuxAcademy.com slash unplugged. That's where you go to support this show and sign up for a free seven-day trial. Thanks, Linux Academy, for sponsoring the Unplugged program. Now, I go to the virtual log to ask anybody in there, anybody a Plasma desktop user?

Anybody running Plasma 5.8 or a recent version? I know Tamal is, but he's not in there today. Nope. He's our big KDE guy. He really, you know, he's a— there today. Nope. He's our big KDE guy. He really, you know, he's a... He represents. He does. He's there fierce, running the neon.

Yeah. Cutting edge. But we do have a little bit of a picture for those of you that might be interested. Perhaps people listening at home might be a Plasma desktop user. Well, we have a little snapshot of what's coming up. They say their general direction point towards professional use cases, which is Hey, that's just

That's music to my ears, Wes. It really is. I love that. We want Plasma to be a solid tool, a reliable workhorse that gets out of the way. Allowing the user to get the job done quickly and elegantly.

We want it to be faster and better quality than the competition. I like all those things. These are great goals. Now here's how it's all going to break down. Our plan is to move from four to three yearly releases in 2017 and 2018, which we think strike a nice balance between our pace of development and the stabilization periods around that. Our discussion of the release schedule resulted in a following plan. I'm going to underscore here. This is probably all subject to change. Well, here's kind of what they're thinking.

Plasma 5.9, the 31st of January 2017. So this is all 2017. Plasma 5.10 in May. Plasma 5.11 in September. So you kind of get. Plasma 510 in May. Plasma 511 in September. So you kind of get an idea of the cadence there. Plasma 512 in December. Plasma 513 in April 2018. Plasma 514 LTS August of 2018.

So that's our next LTS in August of 2018. So this is interesting. So they're going from four to three yearly releases. And so you can see it stretching out. This is interesting. So they're going from four to three yearly releases. And so you can see it stretching out. But what I think that means is sort of a stabilization of the landing spot for Plasma desktop users.

I think this is a really, really good kind of rough plan, even if they can't nail it exactly. And it's fascinating kind of being on the outside, getting to look in with this much transparency into like their inner working. They say that they're going to keep working on the breeze theme. They're going to double down on its refinements and existing details. They're going to work on icons across the whole UI and better theming for GTK. They have a feature backlog that they plan to get on, including perhaps support for the global menu, similar to the way Unity or macOS do it.

Hey, hey. What do you think of that? You know, I don't use it myself, but KDE is already so customizable. I can see why, you know, if Plasma had this feature as well, I think that'd be great. Also, probably not a surprise to us, they're working more on Wayland. KWin brings almost feature-complete Wayland display server now, which is, remember how you have to write your own Wayland display server? Wow, what a task for these desktop environments, just to go from being a client to have to be the server,

which they say already works in most use cases but hasn't seen a lot of real-world testing and lacks certain features that X11 users expect right now or ones they want to offer on modern hardware. So I like that too. And, of course, they're going to continue to work on Plasma Mobile. But, you know, typically we could easily say, oh, no, another project that's focusing on mobile. I actually feel like being three or four down on the list of things to do is just about the right spot. It's something you want to keep working on. It's something you don't want to lose too much ground on.

Right. But it probably shouldn't be the primary focus of your project, at least at this stage, until something ships that makes it work. That makes it make sense, right. Yeah. So I think, again, I feel like they've got this. ships that makes it that makes it make sense right yeah so i think i again i feel like they've got this online services they plan uh improvements and integration of online services and dependency handling for assets installed from the store they have a developer recruitment outreach they plan to double down on as well go help them out i yeah right yeah i really think that this project is

firing all cylinders right now doesn't it feel like it does it It does. It really does. Especially with that Wayland stuff. I'm excited to see that. It feels like, I don't know, maybe this time next year we'll be talking about how great plasma is on Wayland. Well, 5.8 is a solid release. I mean, I still have it. No, I don't have it on any systems anymore. That's right. I did a full cord install for Ubuntu 16.10. And that was a serious process.

Before we move on, TechMav, I wanted to circle back. You had more to say about the 16.10 deployment. Go ahead. TechMav? TechMav? If you're TechMav, you should... Hello? No?

Arlos. You know, if I could do voices, I would just pretend to be one of them. Oh, man, you could be an entire mumble room. That is worth thinking about. I won't say a thing. Yeah. There we go. Hello, Tech Matt.

Control, not shift. So, yeah, you wanted to loop back to the 1610 stuff? Because we're about to talk a little more Ubuntu, so this would sort of be a good spot to follow up. Oh, no. and more Ubuntu, so this would sort of be a good spot to follow up. Oh, no, you were asking about 5.8 for KDE, and I switched. I was on 5.8 Plasma before I went to the 16th end of Ubuntu. Okay.

And what were your thoughts? I mean, the Plasma, that was actually the best experience of Plasma I've had in a long time. It's very smooth. And I like how they are really focusing on integrating Breeze across the board. Yeah. They've done really good at that. And also, it was pretty cool to see the Android Connect stuff built in.

Did you get to play with that at all? Yeah. Very cool. I did some, yes. Yeah. And it works great with my Motorola. Yeah, that is super nice, and it really feels like the future when you're getting there. All right.

Thank you, TechMav, for the follow-up. Now, while we're talking about Ubuntu stuff – This is CNN Breaking News. I can't. I'm sorry. I'm sorry. It's okay. It's just –

I'm not saying anything. How many times do we have, like, three or four really big stories break in the Linux world? Right. And thank God they're just not saying anything. How many times do we have like three or four really big stories break in the Linux world? Right. And thank God they're just not about politics. So this is – when you hear about this at first, you're going to say, oh, well, great. Red Hat is already doing this. Suze Enterprise is already doing this. What's the big news?

Ubuntu, once again, doing their own thing. Actually, all of that is wrong. But you're probably familiar with the concept of hot fixing or hot patching or live patching a running Linux kernel. Red Hat has their thing to do it. SUSE has their thing to do it. The K-Patch and K-Splice respectively, I believe. There's that Oracle proprietary one. It's a big issue in the enterprise, especially when you're running VMs or when you got a bunch of containers, you got a host machine that maybe has 30 containers on it or even five containers.

To update that running kernel and get it secure, you've got to restart all of those subsystems as well. It's just – it often just goes undone. And then you're left vulnerable. And so Red Hat worked really hard on their solution for this. They've really – they put a lot of time, a lot of engineers in it. worked really hard on their solution for this. They've put a lot of time, a lot of engineers in it. They're very, very proud of it. And very shortly after that, Seuss also said,

well, hey, we've been working on this. We're very proud of this. And just behind the scenes, all of these companies now have reached out from their PR people to tell me all about it, to tell me all of the different things, to make sure that I know everything about it. So they're all very – each of these companies have been very – yeah, very wanting to make sure they get the message out there.

But I actually think Canonical has the best implementation. They might be the last of the party. But sometimes that's a good thing, right? And this is something that you listening at home with your one computer or many computers can use. So you don't need an enterprise plan. You don't have to be a Red Hat enterprise user or a SUSE enterprise user. If you have Ubuntu on your laptop or your desktop, it just has to be 1604 on x86 and you can do this. So this is the ability to patch critical security vulnerabilities live in your kernel without rebooting.

And this is what they're rolling out today. It's called the Canonical Live Patch Service. Ubuntu 16.04 LTS's 4.4 kernel includes an important new security capability in Ubuntu, the ability to modify the running Linux kernel code. Think about it. That's like the code in RAM. This is amazing. Without rebooting through a mechanism called kernel

live patch. Today, Canonical has publicly launched the canonical live patch service. An authenticated, encrypted, signed stream of Linux live patches that apply to 64-bit Intel AMD architectures of the Ubuntu 16.04

LTS release. Have to be running the kernel 4.4 generic, I think, or something. Can't be running your own patch version. Addressing the highest and most critical security vulnerabilities without requiring a reboot for it to take effect. Now, again, SUSE and Red Hat, they're okay. This is nothing new. Canonical is not the first to think of this.

Why should I care? They're doing a couple of things really interesting. First of all, crazy easy to set up. You do have to create an account on their thing, but it's really, it's a really interesting way. It is a clever take

on setting up something as complicated as kernel live patching. I'll just put it that way. So I want to cover a couple of interesting things about it and a couple of limitations.

The live patch service is available for generic and low latency flavors of the 64 bit kernel. Oh, that's nice. Yeah. So you've got to have x86-64 and AMD-64 builds of Ubuntu 16.04 LTS's kernel, which is 4.4. Canonical live patches

work on 16.04 LTS servers and desktops on physical machines, virtual machines, and in the cloud. The safety, security, and stability firmly depends on unmodified Ubuntu kernels and network access because you've got to connect to the service. The ups, this is where they start to hint why I think they have the superior solution here. This is the first hint of it. The upstream Linux live patch functionality.

Upstream. Live patch functionality is currently limited to 64-bit x86 architecture at this time. However, IBM is working on support for Power 8 and System 390 mainframes. Well, I guess they want that uptime there. And there's also active uptime development, upstream, I'm sorry, development on ARM64. Hey, that's cool. Yeah, that could be big. So that's where this is really different, is this is an upstream thing that's already in kernel – starting with kernel 4.4.

Every live patch by Canonical is rigorously tested in Canonical's in-house continuous integration, continuous delivery system. They do a quality assurance, which tests hundreds of combinations of live patch kernels, hardware and physical machines, and virtual machines. But again, you are pushing out updates directly to people's kernels. So this has got to be... You have to take that pretty seriously. This is a whole other level. They aren't willing to reboot, so they are taking this pretty seriously, right? I mean, if they're going to trust your patches.

And how they do a little additional testing is interesting. And I'm going to get to that, but let me just tell you about this. The canonical live patch service is intended to address the high and critical severity Linux kernel security vulnerabilities as identified by the Ubuntu security notices and the CVE databases. If you watch TechSnap, we talk about this stuff all the time. And so there's a certain level of CVE vulnerabilities, and when they reach that, they'll do live patches. There are some limitations to the kernel's live patching technology. Some Linux kernel code paths cannot be safely patched while running. They say they'll do their best to supply canonical live patches for high and critical vulnerabilities in a timely fashion whenever possible.

There may be occasions when traditional kernel upgrades and reboots might still be necessary, and they will communicate that through notifications. So – One question. Yes. Do they have a catchy song advertising this service? Because I don't – I mean I don't know if I can deal if they don't. It's funny because when they get to the comparison, that's one of the – OK. I'll get there.

I'll get there. But this is the details I think that applies most to people listening that aren't in the enterprise. So if you're a home user or a small business, this is the stuff here you're going to care about. You do need a subscription, possibly. And you definitely need an account. The canonical live patch service provides a secure encrypted authenticated connection, which seems pretty necessary when you're live patching your kernel. I'll just take any old bitstream. That's fine.

Throw it in the ring. Connect over HTTP. We're good. To ensure that only properly signed live patch kernel modules and most importantly the right modules are delivered directly to your system with extreme quality and high testing, they do have a subscription service if you have a lot of systems. But Canonical is providing the live patch service to community users of Ubuntu at no charge for up to three machines, desktop servers, virtual machines, and cloud instances. Well, damn, I got way more droplets than that. But if you got up to three machines, you can play with this live patching. That's your house.

Or that's your three critical servers. Or, you know, maybe my three droplets I never want to reboot, but I want to keep them secure. That actually might work. The other thing is maybe you do it all. You know, you're using their fancy little XD thing. You get three big droplets and you run containers on those. Oh, man. Yeah.

And here's the other thing that I think is fascinating about the free mechanism. Because this is also how Canonical is sussing out if these patches are going to blow your system up and this is i mean you interpret as you want but they say a randomly chosen subset of the free users of canonical's live patch will receive their canonical live patches slightly earlier than the rest of the free users or the subscription users as a lightweight canary testing mechanism benefiting all canonical live patch users. So the free users, some of them will randomly get selected to

occasionally receive the updates before the other users. That's going to be you every time. You know it, Chris. That's you every time. You know, to be fair, this is pretty common. This is how Microsoft rollouts major Windows updates. Once they get enough data back that the Canary live patches apply safely, all canonical live patch users receive their updates. So that's –

Do your part. Yeah. Freeloader. Yeah. Freeloader. Okay. And then, of course, the big question, what about the source? What about the source?

The source code of live patch modules is found here, and we have it linked in the show notes. They got that all published. The source of the Canonical LivePatch client is part of Canonical's landscape system management product and is commercial software. I see. Now, I don't think anyone in the enterprise gives one shit about that. No, they don't care. They have enterprise agreements to protect them. Refer to our topic enterprise gives one shit about that. No, they don't care. They have enterprise agreements to protect them. Refer to our topic later in the show about that.

Now, how does this compare to the stuff that Oracle is doing, to the stuff Red Hat is doing, to the stuff SUSE is doing? The concepts are largely the same. But the technical implementations and commercial terms are very different. Oracle's Ksplice uses its own technology. It's not upstream. Red Hat Enterprise Linux and SUSE currently use their own homegrown Kpatch and Kgraft implementations, respectively. Canonical's live patching uses the upstream Linux kernel live patching technology. That's awesome.

Ksplice is free but unsupported for Ubuntu desktops and only available for Oracle Linux and Red Hat Enterprise Linux servers with an Oracle Linux Premiere support license which is $2,300 a year a node. Of course it is. $2,300 a node. It's a little unclear how to subscribe to the Red Hat Enterprise kernel live

patching but it appears that you need to first be a Red Hat Enterprise Linux customer and then enroll in their special interest group through your technical account manager Wow. And then the SUSE live patching is available as an add-on, but to SUSE Linux Enterprise Server 12, But to SUSE Linux Enterprise Server 12, where you have to get a priority support subscription at $1,500 a note a year. But you do get that free music video. Hey, yeah. So that's got to be worth something.

I mean, you're obviously relevant if you've got a music video. I mean, look, everybody, they're relevant. Then, Canonical. They still owe a lot of money to the people that made that video. That's where the subscription price comes from. Yeah, they've got to pay that out. So when you do decide, if you decide to subscribe to the Canonical LivePatch, if you've got more than three systems, you're paying $150 a node per year.

Well, okay. Well, that's more reasonable. Yeah. And presumably you have to care about these nodes enough that you want a LivePatch, right? If you have six systems, you could probably go make two accounts. I'm not saying, but you're just saying. Yeah. So this seems to be a pretty huge release from Canonical,

even though a lot of people will dismiss it as, ah, that's been done before. But I don't think it's been done like this. I really think what they did is they waited for the upstream code to just have it. Right. They didn't want to invent their own system. Yeah. Which probably means that anybody that's running 4.4 could probably invent their own method of doing this. But I think this is a pretty competitive offering. I also think it's not – at least while it's competitive, see how it used.

But I can definitely see in the future it getting thrown over the open source fence. And then here's the other thing that I think it shows. It shows why Ubuntu is so much more popular on the cloud type VPS instances than SUSE or Red Hat Enterprise Linux. Would you think about it? It's almost a joke to suggest that you go buy a VPS and run SUSE or Red Hat Enterprise. Maybe CentOS for sure. Yeah, totally. But SUSE Enterprise Linux or Red Hat Enterprise Linux,

which are built for servers designed to run on 24-7 workloads, it's a joke to think you'd go get a VPS with one of those on there. You could find them, but nobody does it. No. And why is that? If you look at those prices that we just talked about, if you decide to go with this on your production VPS with Ubuntu on there, you likely did not, A, pay for the version of Ubuntu that's on there,

and B, you may not even have a support subscription. And C, if you decide to get this live patching, it's $150 versus $1,500 or $1,300. Really, so Red Hat, you'd have to get the $2,300 subscription from Oracle and the $1,300 subscription from Red Hat. And with SUSE, you have to get the $1,500 subscription and it only works on Linux Enterprise Server 12.

Jeez. And my, who knows, but my gut tells me that likely the next LTS will also support this. So they'll do this for every LTS. And now you look at a system here that's sitting in production, I start thinking to myself, well, I wouldn't mind that. I don't like rebooting these ever. And $150, that's really not that much, especially if it's only a couple.

And hey, you get three free. Yeah. Pretty neat. Anybody in the Mumble room want to jump in before we move on? I think – does that wrap up our breaking? I think that's all the breaking news. All the breaking news. Yeah, so that's, that is it. Oh, actually, no, it's not.

There is one more thing. There is one more, and I won't do the soundboard because I've been obnoxious with it. Scraper, you have a, let's see, scrappy paw? I'm not even sure what. Scrap jaw? Scrap jaw. You have a comment about reboots go ahead yeah you guys being in a studio that you only broadcast from you know once uh you know a day i know you have live things that are running throughout 24 7 but wouldn't you shut down unused machines you know you would think um the big issue is with HDMI video capture and God, I can't wait till we move away from HDMI.

There is a display port everywhere. I hope there is a there is like a sequence you have to go through for your capture and your output to see each other. And it has to be timed just right and when you have multiple systems so we have we have three doing hdmi in and like we have like four or five hdmi outs and they're just they're just a catastrophe to work with and so we generally what we do is we run the systems in like a a low power mode where the display is still sending a signal but but like the screens are off, the hard drives are spun down or something like that. But we leave them running 24-7 just because the schedule of doing daily productions and sometimes two shows a day is tight enough. There's not quite enough room for troubleshooting that because sometimes troubleshooting that can take like 45 minutes to an hour.

And there are a lot of times where you're showing up here and then 10 minutes and you need to be on the air. Every now and then. Yeah, that does tend to happen from time to time. Most of the time I'm here before a show, but there have definitely been times where there was one time for Tech Talk today where, thank God for Rikai because I don't know how he knew I wasn't just not coming in, but somehow his beard spider sense was tingling, and he knew that I was on the road, but was not going to

make it on time. And I legitimately don't know how he knew, but he knew. Because it could have been, I guess because I don't ever just not show up. I guess that's how he knew. I guess he knew because he knows. I always make it. And so, he's like, but he must have looked at the clock and said, there's no way he's going to get. So he came down to the studio and got stuff going, and I literally

walked in the door and sat down and started the show. And that was like the tightest it had ever been. And it is. Wow. And if there's any outage right there, I spend the first half hour of the show troubleshooting that. And then I'm all. And you're grumpy. I am wrecked.

I'm wrecked by the end of that. Yep. I'm wrecked. So, yeah. And you don't want to start a show all wrecked. No. Okay. So we do have one more breaking news.

Okay. This is CNN breaking news. This is something I think I'll be using in the near future. Me too. This is really cool. It's a new tool that lets you easily install Ubuntu Touch OS on your mobile device. It's a tool developed by Marius Quebec. Marius?

Marius. What do you think of that last name? Quebec? Yeah, I don't know. Okay. And it's called the Magic-Device-Tool. Magic-Device-Tool. The first stable version, Magic-Device-Tool 1.0, is now available to everyone and promises

a very simple and easy to use batch tool for installing Canonical's Ubuntu Touch as well as Android, CyanogenMod, or Phoenix OS. So it's not just an Ubuntu Touch tool. You'll be able to replace your mobile operating system on your device with any of the following. Cyanogen, okay, I just said all that. I don't know why they repeat that. Oh, oh, oh, there is one caveat I wanted you to know about, though. They don't support the versions of images that have the Google Apps.

Okay. So here's the devices, though. This is what matters. The BQ Aquarius E4.5, the BQ Aquarius E5 HD, BQ Aquarius M10, BQ Aquarius M10 FHD, et cetera.

So the US MX Pro, Mewis Pro 5. Now we're starting to cook with gas, though. The Nexus 4, the Nexus 5, the Nexus 7 2013 Wi-Fi, the Nexus 7 2013 LTE, the Nexus 7 2012 3G, the Nexus 10, the OnePlus, and the Fairphone 2. May not work.

You don't think so? Well, that's what that says. Oh, does it? Just at the bottom there. The Fairphone 2. That's okay. I don't think that's probably. No, it's not a big deal.

The Nexus 5, though. That's a big deal. Also, the Nexus 7 7 i've always wanted to do it on that but i just don't have the time and this thing it also it can do backups here of your machine it can do restorations it can unlock or lock the bootloader and it also supports installing the twrp recovery image this is amazing yep you know i've seen some similar tools for windows that i think it's think it's a nice niche for people where it's like, well, maybe you flash your phone or need to do this like once, maybe a year. And so if you're not like ready to jump down in the ADB shell or any of that, here's a handy tool that you can use to do that. Or honestly, for me, it's not that I haven't used ADB, but I haven't used ADB in so long that I would spend enough time Googling around and reading XDA forms that I just don't want to spend that time.

This thing already knows how to do it. Because what I really want is the end result. I don't really care. You're not developing for Android. You're not making it. Yeah. Yeah, I'm that guy. I'm that guy.

I admit it. When it comes to this kind of stuff, I'm that guy. But I think this is a super positive development, and it's a tool I'll be watching for a while. I think just as soon as my Pixel gets here, I'll be using this and trying out a bunch of us on the Nexus 5. You got the Pixel! I did. I did. Which one?

I got the smaller one. Yeah, good call. I mean, it just seemed—I don't have huge hands. It's all about you keep it in your pocket. That's what it's about. I got a 4.7-inch phone now, and I'm all about that. It's in my pocket right now. I didn't even know I had to feel it.

I had to touch my pocket. I've literally been sitting here for how long? I didn't even know that phone's in my pocket. That's what I'm talking about. See, the 6P, it's a brick, dude. Totally. So that's interesting. Did you get the 128?

I did. I did. I don't know if I'll need it or not, but I plan to have the phone for a while. When did you say you're getting it? because I went for the 128 I don't get it until like mid-November so a month or so that's rough man

mid-November huh? jeez yeah have you read any of the reviews that came out today? yeah they're positive I think it's the Android phone to get right now if you're going to get an Android phone. And I need a new phone.

Yeah, you do. It seems nice enough. Not only do you have a Nexus 5, but your phone is busted the F up, dude. It is mad. Well, I've been waiting for this. Have you scratched yourself with that yet? Thankfully, no. But it is dangerous.

It is definitely dangerous. That's probably just because you don't make many calls. It's made by LG, but it may explode. You also are constantly charging that thing, I've noticed. That's true. And that's one of the things that come back on the early reviews. By the way, User Air 7 just came out where we gave you our full take on the Pixel if you are interested, as well as the absolute very best way to get

great audio on Linux. We talk about that and a few other tidbits that I think are actually, I think it was a really good episode of User Air 7. I'm excited to listen to it. User Air 7 is out where we do talk about the Pixel, so I cannot wait to try it out. I'm kind of on the fence myself. I think I'm going to let things go for a while. I'm good for a bit.

When you're in a position, you know, if you have a good phone that you like, it's not necessarily worthwhile. Also, I am personally, until I am proven otherwise, I am an avid believer in optical image stabilization. I know that they have incredible post-processing, but physically having a lens that floats, I think, is unbeatable. And so when pictures are a big part of it, I know it's got a great camera and a great sensor, and so they say, and the reviews do seem to back that up, but it does lack actual optical image stabilization.

That, to me, in 2016, nearly 2017, it's going to be your phone in 2017, that's a bit of a... That's a bit of a bummer. I guess we'll see. We can do some photo shoots here in the studio. Yeah. I mean, so far the reviews have been pretty good.

I've loved it. And the video, they say, really is impressive. So, yeah. You know what? Why don't we take a moment and why don't we thank our sponsor, Ting? Go over to Ting. Actually, you know what? I can't believe it.

I almost forgot to say you can put Linux in your – Linux. Linux.Ting.com. Linux.Ting.com is not only a good way to support the show, but it will give you $25 off a device. Or if you bring a device like a Pixel, it will give you $25 in service credit. I would really consider this if you're going to get the Pixel 2 because Ting's no contract, no determination fee, and you only pay for what you use. That's really the secret sauce to Ting is if you can be kind of clever, if you can download your podcasts over Wi-Fi or what I do on Spotify.

I have Spotify on my phone, but I download the tracks on Wi-Fi and I play them offline because I like the super crazy extreme quality anyways. There's just like really four or five things you can do, and you can save so much money with Ting. Yes. They just take your minutes, your messages, and your megabytes, and they add them all up. And whatever it is, at the end of the month, that's all you have to pay. Ting keeps rates simple. We don't make you pick a plan. Instead, you just use your phone as you normally would.

How much you use determines how much you pay each month. You can have as many devices as you want on one account. That's good because when you use more, you pay less per minute, message, or megabyte of data. Your usage plus $6 per active device on your account plus taxes is your monthly bill. Simple. That's what we mean when we say mobile. That makes sense. They also have incredible customer service and a fantastic dashboard. You get to speak to a real human being when you call customer service. You're not playing that game. And if you do end up with something like the Pixel from the Play Store, you can get a Ting SIM card for $9.

You just put it in there. You activate it. There's no contract, no determination fee. Try it out. See how it works for you. Why not give it a go? In fact, I don't know. I haven't looked at the Pixel, but I bet you could choose CDMA or GSM because Ting's got both networks. And if the Pixel's

out of your price range, because, geez, it is an expensive phone. Yes. I bought two. Did you really? Oh, for the lady? The wife's getting one, too.

Yeah. Yeah. Yeah. Boy, I mean, and that's nice for charging and all that kind of stuff. That's a huge thing. All USB-C.

All the consistent USB-C households. Yes. Yeah. I can understand. The future. Yeah. If can understand. The future. Yeah. If you want to save some cash, Ting has a semi-recent blog post from the 11th about five wise smartphone choices that are under $200.

So if you still want to get a smartphone that has some functionality, that's going to get some updates, maybe has a nice camera, they've got a post about that too. So that's super nice because you can either bring your device or grab a device directly from Ting. It's unlocked. They just make it so easy. Linux.Tting.com. And a big thank you to Ting for sponsoring the Unplugged program.

So do you remember a couple of weeks ago there were some headlines where Linux – it went essentially like this. Linus admits that Linux 4.8 had buggy crap. And he raged against it saying there was no effing excuse to knowingly kill a running colonel. And he talks about how it was his bad for taking some patches from Andrew right before the 4.8 release. Do you remember all of this, Wes? Does this ring a bell? Yeah, I believe we touched on it here on this show just a little bit. So what I want to do is I want to take a moment and I want to go a little bit deeper and talk about what they're actually discussing here, because this turns out to be a topic that is now giving a lot of Linux critics some serious fodder.

And if you heard this, you would think, oh, geez, if you're not familiar with intimate details, you might think, geez, what the hell is Linus doing? You know, this is Linus. He's raging out once again, and he's acting out of emotion. Bad press for the Linux kernel. Yeah, and it's, I don't know a lot about this, Wes, but I got some clips I want to play. You've had some time to think about it. It's about this bug

underscore on that led to the Linux kernel stopping. And we've got a post linked in the show notes by Samsung's open source group that say that they are working, this is something they're working on. They say it's a top priority for Linus to change this bug. What are we talking about, Wes?

What is bug on? What are we talking about? Bug on is a kernel function or macro that at least in some ways is used when you see a kernel bug where you're in a position where and this is something they talk about in this thread as we'll talk about, you know, when the kernel is in a position where there's no meaningful error recovery. Right now there's a current trend where it gets used a lot by kernel

developers in development. So it's kind of like in other terms, it's similar to like an assert statement where you're basically saying like, well at this point, everything that I'm doing, this should be true. If it's not true, then it's violating one of my assumptions. Going forward, well, at this point, everything that I'm doing, this should be true. If it's not true, then it's violating one of my assumptions going forward. It doesn't make sense to continue because, you know, my axioms have been violated. It gets used a lot in development when you haven't written error code, you don't know what error code to write yet you're exploring.

Basically, this trigger is like, oh, well, my new feature clearly didn't work. Crash the kernel. We get a stack trace. You can begin debugging. Sometimes this does then go through into the real kernel, and it does have a place. And then part of why I think this thread is interesting is it's kind of you get to see the discussion of, like,

when should this be used? When do we crash the kernel? And so it's sort of a discussion around should it be easy to crash the kernel? Because really, you know, and Linus is, I think, in a good way, concerned about, I mean, it's not like stupid, bragging uptime concerns,

but there's lots of places like, do you remember the bad press during when Windows 10 rolled out when there was like hospitals in Africa or other places where their machines were rebooted? Absolutely. Problems were caused.

There are situations not in the data center where you don't want to go down or you want the admin to be able to choose to go down. Yeah. And so, you know, they're very concerned about if there's anything we can do to avoid a situation. You know, by all means, if you're the file system code and you can't, you know, you're going to write damaged data to the drive, stop what you're doing. But it doesn't also, in that case, it might not mean that you have to crash the whole kernel.

So this is what's fascinating about this particular issue because what you just said there seems absolutely reasonable to me and seems like a likely use case scenario for Linux's number one deployment, which is the server. However, Linux critics have taken this and spun this as, well, this is Linus, sick and tired of bugs crashing his buggy kernel, and they're just paper machining over the problem, and they're just suppressing the issue. And if you watch this week's episode of BSD Now, a popular guest, Brian Cantrell, was on, and they talked about this very issue. Brian is – I love every episode of BSD Now he comes in. It really is. Because he is such a great Linux critic because he really sticks it.

I mean he's so outspoken. I love watching him. He's very opinionated. He's often right on the money. He's a terribly smart guy. Yeah, and so he tears in to this particular issue on this week's episode of BSU Now. I fully recommend you watch the entire interview. We're going to play like two minutes from a much longer interview.

like two minutes from a much longer interview. But this is him talking about them switching over from this bug on to something less dramatic that doesn't stop the kernel. And this is what's becoming the sort of external community take on this particular issue. And music. You need to be like have all sharp objects away from you. They are going to use like I know said can fix this bug. Yes. No, no, no., said can fix this bug. No, no, no, said

did not fix this bug. So they're going to change every bug on to worn on. And it's like, there, I fixed it. It's like, okay. And so now, if you actually have a Linux kernel programmer that has, I mean, the poor right-thinking individual that is living in the ghetto of Linux just trying to do the right thing and has an actual legit bug on, if this happens, please panic the system, that is now just going to be turned into a warn-on, and now you've actually introduced bugs effectively in systems or certainly made any bugs that do exist in the subsystems much more difficult to debug. It's like, how dare you? Come on. I mean, aren't you old enough to, I mean, like Torvalds and I are about the same age. Like, do you still have this many hormones

running through you that you think this is a good idea? I mean, don't you realize that, I mean, aren't we getting older and wiser? But, you know, whatever. Hey, worn on with ButterFSs what could possibly oh god twitch i do love the i've got the and the whole like the the it's okay system d will recover it you know there's a there's a you know chris there comes a point where it's just like you're clearly you're just piling on here you're just i'm to – I'm sorry. Like can I actually make him pass out? I mean are you two playing a game where you're going to see if you can actually – I think we have a betting game on the side here. It's funny to me.

How long before we can make Cantrell stroke out live on the show? Exactly. Like is it normal? My vision is kind of fading right now. So he's so frustrated with Linux that he's about to stroke out. And so you could hear a couple of assumptions that he's working from. And it's not just Brian. Brian is just sort of representing an opinion that is forming.

And I wanted to address a couple of assumptions that I think he's working from when he makes these statements. when he makes these statements, you might have noticed that in his world, Linus the man is irrational and makes code commits based on his emotions, or as he put it, his hormones, not rational code commits, which I think is probably not the correct assumption to make. But if you, if you believe that to be a fact, you could see how he's built a foundation of an assumption there. Another assumption that kind of crops up in his, and you'll hear another example here, is everyone working on Linux basically kind of crops up, and you'll hear another example here, is everyone working on Linux basically kind of just doing it their own way. They're doing it for fun. And it kind of ignores the fact that the vast majority of commits to Linux are from business

enterprise developers who are getting paid by their company. They're serious people. And I want to, before we completely respond, I want to talk about this particular aspect, the way this kind of gets spun and becomes the reality of the situation. And he continues to talk about sort of the pervasive bad thinking that's in Linux that led to things like ButterFS. And I want to talk about that too. Before we go to that, did you want to interject on his take on the bug on thing? Go to that. Did you want to interject on his take on the bug on thing? Well, I think him highlighting the said point there, they do mention that that is mentioned by a contributor. But what Linus says, that's not that implementation that they would go with. And then there's some debate about how it's used.

But I think what's really missed is just that the culture has not been using bug on. And in this case, it's VM bug on as a like, oh, no, we you know, like we messed up the CPU stack and we need to crash. So it's not it hasn't been used in this situation. So and what Linus is really concerned about here is like minimizing that. Right. And so like the issue is, is that once something went out at scale in production, it got used in a different way than they intended. And this one in particular crept by in like an RC8 patch level, right? So it's like it wasn't during the regular merge window. It didn't get all the regular testing it should have.

And like it's okay, right? Yes. There are times to crash. And he admitted it. It's probably his mistake. And I think what he's just saying is that it's really it's a discussion now about like what are the semantics of crashing the kernel. We should probably make it so that the developer like it doesn't get confused with how you debug. It doesn't get confused with –

If you had heard that criticism on its surface from an intelligent person like Brian, I would think, God, Linux is really falling apart. Right. I mean Brian has great credentials. He's a super smart guy. He's a CTO of Joint. And I think his argument is right in that like, no,, you should, I mean, you should have asserts, right? You should have, I think in Solaris he says it's verified. You should have these things that say, like, you know, if the kernel memory algorithms are working correctly, this is this.

And if not, we need to crash. And Linux does have those. And this is a conversation about where do you set that level and what are the semantics of crashing. I also, I mean, I love all, I love alan and chris and and i think brian's great too but i love it when they when they poke fun at system d because i don't feel like i don't even need to respond to that because what i'm going to do is is like a tiger i'm going to i'm going to sit back and watch my prey and i'm going to watch them take shots at system d for the next couple of years and mark my words five years from now, BSD will be fully rolling out their systemd equivalent.

They won't – don't call it that. They will be – they'll be sure not to call it that. But the reality is systemd, despite what we might like to think in the Linux community, not a new idea. No. No, it's not. Sun was working on – is it SMF? SMF, yeah. Yeah.

And of course, Apple had LaunchD. I remember using LaunchD to configure services before SystemD was even a thing. It was like, hey, this is pretty sweet. I believe there are two, if not three projects in the BSD community already to replicate SystemD-like functionality. It's just none of them are necessarily mainstream. It's coming. So just sit back as a tiger and watch your prey as they poke fun at systemd now, and then when they will roll it out and all of the logical backflips they'll have to do to explain

why it's nothing like systemd. But let's go on about the pervasive bad thinking in Linux. This is a common BSD culture projection on the Linux community that Brian manages to put in really great summed up words. And so this is a great point to sort of look at something that I think is a common misconception. No, they actually won't. And the other thing I think that is actively happening to Linux, I think is kind of the peril that they run, that right thinking individuals are not attracted to that culture. I don't think he means like politically right. I think he means right brain. They're talking about right brained people earlier in this. So right brain versus left brain. The thing is, he's saying is that Linux doesn't attract the right type of critical thinkers. Right thinking

individuals are not attracted to that culture. And so people, even if you're interested in file systems, are you going to go in year 22, are you going to go work on ButterFS? Are you going to go work on CFS? If you are interested in operating systems, are you going to go to a culture that is replacing bug-ons with worn-ons? Or are you going to go to the BSDs

or to Illumos or what have you, where you've got cultures that are not just more innovative, but also more rigorous in their thinking? So cultures are subjective for the eye of the beholder. But if I'm 22 years old and I want to get a career in this field, I'm probably going to go with the platform that has the demonstrably larger deployment in the server space. That's what I'm trying to say. I mean, think about it.

If you want an employable skill, it's obvious. The choice is obvious. So I kind of disagree with the premise to begin with. But his core message is the way the culture is, the professionalism, that's more appealing in BSD. And people that think that way – Or a Lumos. Yeah, or a Lumos, which is his baby. Interested in operating systems, are you going to go to a culture that is replacing bug-ons with worn-ons?

Or are you going to go to the BSDs or to Illumos or what have you where you've got cultures that are not just more innovative but also more rigorous in their thinking? And I think that in an all-open-source world, there's total freedom of choice for technologists and the differentiators become almost the cultures and these cultures become self-perpetuating. And so I think there's a danger that you end up with butterfesses over and over and over again. And I mean, I just wonder how many of these is it going to take before people kind of like connect the dots that. OK. Let's talk about that point.

How many times are we going to screw something up before we admit that sometimes we just we think that there's bad ideas and we should just move on? I think he's kind of right. You still have ButterFS defenders out there who are going to argue with this very moment in this show. Yep. Even though it's so obvious to everybody else watching from like a distance that if you're an enterprise and you have serious data needs, you're not going to deploy ButterFS. That's what that system has worked out to be for. Yeah. And it has enabled an entire product category for some companies out there. I will say it is interesting though just given Facebook's use of it.

It probably has a comparable number of deployments now. It's just that it's not for data reliability. You know, what are we going off with that Facebook thing though? I mean I'm not – if you're right, if you're right, then that probably is a massive deployment. But here's what I've heard. I've heard several things. I've heard that that story is old and out of date. Sure.

I've also heard that those systems run on RAM. The storage is RAM based. They're RAM disks and they wipe them every 24 hours. I can see that as well. I'm not trying to argue for ButterFS as a

data storage solution. But that was always our go-to about ButterFS. Facebook. We had that. That was our go-to. But I don't our go-to about Butterfest. Like, whoa, Facebook. True. We had that. We had that. That was our go-to. But I don't actually know how much that holds up. And I mean, you don't see a huge number of patches these days or that much improvement. But here's where I really think he starts to make a better point.

He really kind of nails it when he talks about the fundamental issue that I always had with ButterFS and I couldn't put into words as well as Brian does here is ButterFS should have been production ready from the beginning. It didn't have to be feature complete, but it should have been production ready. Cultures become self-perpetuating. And so I think there's a danger that you end up with butterfesses over and over and over again. And I just wonder how many of these is it going to take before people kind of like connect the dots that there's actually – there are broader differences than simply the technology. What befuddled me was how they didn't catch this sooner because the test case they did for it of create the RAID Z and like DD zeros into one of the drives and then run a scrub. Right.

FreeBSD handbook has every new ZFS user do that on purpose to watch, to show them how a scrub works and repairs the data. Like our newbie guide has every user do that exact experiment to prove that ZFS works and no one had ever done it with ButterFS?

Ever? What? Right. Well, and I think that this is the idea. Well, you wouldn't do it because it's not ready for production yet. It's like that. It would be mean to do that because it's not. It's like, no.

That's just setting us up for failure. It doesn't understand what software is born ready for production or not. And the software that is, especially software in the data path, you don't make it ready for production later. It's like security. You don't retrofit it onto an extant system. You either design it for production use or you don't. And it's not something that you asmetotically approach.

And yes, there will be bugs in the system. But the bugs in a system that is actually designed for production use, that is actually a salvageable system. Bugs in a system that was never actually designed for production use or thought that production use was something that would be done later or by the little people, it is not actually tenable. And with ZFS, people don't realize this, that ZFS was designed to be used in production from day zero. So it's just that it didn't have features in the beginning. I think that's a pretty fair point, actually. And I think the lesson learned from ButterFS is a change in our perspective. Let's not

wait for something to be production ready. It's just, it's not 1997 anymore. We're in a different world now when it comes to this stuff. Like it or not, people are pretty production. Linux is used, right? Yeah, and it's happening on people's

laptop, it's happening on people's phones, and it's happening on people's servers. If you want to follow more, episode 163 of the BSC Now program, The Return of the Cantrell. That's a great episode. It was actually your suggestion because you had watched and said, we should talk about this. So thanks for the suggestion because I actually think it's good to talk about our weaknesses. Oh, yeah, absolutely. And it's good to be completely honest. But at the same time, we do need to call it when people are maybe spreading a little bit of FUD, especially around that bug on thing.

That's just FUD. If you follow it closely enough, that's just straight up FUD. You want to criticize us as a community about ButterFS and say that maybe the culture of Linux allows projects like ButterFS to succeed? Okay, maybe. But that just sounds like a really active, vibrant, huge worldwide community that's just many factors larger than the BSD community. Right. Well, and ignores other subsystem communities as well as, you know,

I mean it's not like we haven't had successful production file systems before and it's not that we won't have them again. So we need the realistic criticism as well as an up-to-date assessment of the question. And it's not like you can't get ZFS right now. The reality is Linux is such a natural force in the technology industry that if you want your product to be relevant on the server, it has to work on Linux. And that even is true for ZFS. and that even is true for ZFS.

But it was a really good interview and there's a lot of good stuff they talk about and I hope to have some of this very discussion with Brian in person at MeetBSD. Oh, hey. Yeah, I think he's going to be there. We're going to hear more about this. The whole BSD Now crew, Alan's going to be down there,

Chris Moore's going to be down there, producer Q5Sys is going to be down there. Chris Moore is going to be down there. Producer Q5 Sis is going to be down there. I mean, like it's a full BSD Now crew plus Brian's going to be there and I'm going to be there. It's November 11th through the 12th. Here's what I decided to do. I have not mentioned this yet on any shows, but I've gotten a few emails about it. I've got a couple of tweets about it. People are like, hey, can we meet up while you're in the San Francisco area?

And we really had this tight. Hadiyah and I, we planned to be there for the 11th and the 12th, and then the other time was spent driving from Washington and back. And it was her idea. She said, you know what? Why don't we just – let's just take a couple extra days. I'll take a couple of days off from work. Maybe Noah can fill in on a show and let's spend a little time down there and let's, let's meet up with some folks. So, uh, I will probably be also be in, in the area the following Monday or Tuesday. And what I plan to do is I'll get down there. I'll get me

BSD, get my head around that. And then I'm going to start tweeting out what my plans are. If you're in the San Francisco area and want to meet up, email me chris at jupiterbroadcasting.com or tweet me even better at chrisles. Let's start talking about it. That sounds awesome. Because what I would like to do is maybe rent a car or maybe Ben will let me borrow his car an extra day maybe and

drive into San Francisco and just hang out and maybe people can show me some cool places and maybe we'll shoot a little rover log and all that kind of stuff. I think it could be a lot of fun. So I'm going to take a couple extra days. You know, we have a lot of viewers. We have thousands and thousands of viewers in the California area, specifically in the San Francisco area.

Totally. And so if you're in the area. If you've ever wanted to see Chris last in person. And maybe, maybe show me your favorite place to eat. Maybe I'll do like a tour. I'll go Alden Brown. I'll just like, I'll go around and drink like coffees from all the different places and have donuts and just get like all kinds of grub on and then record it all for the overlock.

That could be fun. And I think, I think I'm going to be there Monday or Tuesday. I'm going to take one of those days and use it for a meetup and one of those days to play the tourist thing and just like take a day off and actually enjoy the area. So I'm hoping to have two extra days in San Francisco and I want to use one of them to spend it with the audience and one of it to do the tourist thing. So I don't know which ones yet, but it will be Monday and Tuesday and I'll say more. I'll probably on my Twitter account at Chris Elias.

But if you're going to be at MeetBSD, I'd love to see you. Or if you're in the Berkeley area and want to meet up, let's do it. Meetbsd.com, by the way, if... I think... I'm going to tell you, if you're going to go to one BSD event in 2016, I think this is the one. It really seems like it. Now, there's some international ones that are pretty badass, too, but

if you're going to go to one in the U.S. this year, there's probably other great ones coming up, but if you're going to go to one, I think that might be it. BSD returns home. That's what they – Pretty cool. Berkeley, yeah. Yeah, the Berkeley Systems Distribution, right? All right. So let's talk about our next sponsor, DigitalOcean.

Go over to DigitalOcean.com and use our very special promo code of absolute power and knowledge, DO Unplugged. It's one word. It's lowercase. You go create an account. You go in their dashboard. You apply it as your payment method. Now you're not putting a credit card in there. You're not putting your PayPal's in there or whatever's.

They've got your back. You're going in there. You're putting in DOM plugged. You're getting a $10 credit. Actually, if you already have an account too, and you've not used our promo code, you can also get a credit. What a deal. I know. DigitalOcean is super cool to let you write.

They've already got you. They didn't have to let you do that. No, not at all. It's really cool. And so that kind of gives you an insight to what a badass company they are. They're a really great simple cloud hosting provider dedicated to offering the most intuitive and easy way for you to spin up your own super fast system on their infrastructure.

They're all Linux. They use KVM for the virtualization, SSD for the back-end storage, so everything's super fast. You know it. And you get to pick from all of the distros that you would run on a server, plus free BSD if you're a BSD folk.

And there's some cool things you can do when you start playing with your block storage and ZFS. There is some cool things you can do. But really, you can get started for $5. $5 a month. And if you use our promo code DLNPLUG, you get $10 two months for free. Whoa. You can get started in less than a minute.

You'll get a really nice system. And they have an interface that even if you've never set up a server before, you'll be able to actually. It's so easy. If you've never installed Linux even, you could do it. You really could. But what I think is probably more noteworthy and remarkable about that is, first of all, they've managed to do this in a web application. Second of all, they managed to do it in a browser-compatible way that works, like, on your phone. It works on your Linux box. You don't have to have Java installed.

You don't have to have Flash installed. Yeah, you get a full HTML5 console access to your rig. That's all nice. Yep. But beginners or advanced users that have been setting up systems for decades are going to be happy with this. And then on top of that, if you even know how to poke at Ruby or PHP or Python, you can use their API. Like, if you just know how to poke, and they have tons of great open source applications you can just take advantage of already. That's amazing. It's a

great service. And they're super fast and they got data centers in New York, San Francisco, Singapore, Amsterdam, London, Toronto, Germany and India. So you can be like a boss all over the world. Just use our promo code DOunplugged. I mean, it is so awesome to see how Linux and

KVM can bring the fundamental technologies to a company where they can focus on the things like the API, the interface, the documentation, the community. Finally, somebody took advantage of all of it, and that's DigitalOcean. DigitalOcean.com, just use our promo code DLUnplugged.

And a big thank you to DigitalOcean for sponsoring the Unplugged program. So we were kind of kvutzing about this a little bit ago in the show, and we were talking about how most people in the enterprise don't give a single crap that the canonical live patch client is commercial software or that the back end is managed by Launchpad, which is commercial software. The businesses that are more than 10 people, I've

never met a single one of them that would care unless their business is Linux and open like System76, they care. A lot of the enterprises that's not what they're concerned about. If your business isn't open source and Linux, you don't give any shits. You really don't. And let's be

honest with this. And Mumbleroom, you feel the give any shits. You really don't. And let's be honest with this. And Mumbleroom, you feel the urge to disagree with me. But I thought this post over at naughtycomputer.uk, which is a great domain name, was pretty good. It says, the role of free software in a world that doesn't care. The free software movement is about personal and social liberties, giving the owner and the user of a computer control over it. and social liberties, giving the owner and the user of a computer control over it. But most people don't see the problem with a small number of multinational, mega-corporations having control over everyone's computers. They think Apple and Microsoft know what they're doing, and they do a good job.

So why would I need free software? I can see a lot of people's thoughts like that. Or even worse, they've never even thought of, heard of, or conceptualized a world of free software. They only see a world of apps and commercial software. And I would suspect without – I mean I would find it very hard to dispute the idea that that must be the very vast majority of consumers. I mean, that's a default, right? You need your next program, you buy it off the shelf, you get it for however long it's supported, and then you buy the new one.

My lady friend is a good example. Before she met me, I think she had heard Linux mentioned as the thing that runs her server and she knew a couple of computer people that sometimes talked about Linux, but she didn't know that it was free or that it was open and she didn't really ever even consider that there was such a thing because everything else in our life that we interact with is productized, commercialized. You need a specialty thing. You got to buy that model. It's branded.

It's marketed. So you don't even – your mind doesn't even shift to that mode to understand that there could be something like free software that's created not by a company but by a system of volunteers who are each scratching their own itch. That idea isn't something that you just consider. You have to be introduced to it. And like my lady friend, when I introduced her to it, she said, well, this is obviously the better way to go.

This is unquestionably, even as a, I mean, I'm a small business person with like a couple of people working here and this seems like an obviously good choice for my business. But until you understand and see it, you don't even conceptualize it. But until you understand and see it, you don't even conceptualize it. So what is the role of free software in a world where that is perhaps arguably the reality? Mal Maroum, do you have any thoughts? Yeah, I think the problem is that they don't understand the whole computer thing. It's not the thing that they don't care.

They don't understand. So it's some kind of self-protection to pretend that Microsoft and Apple care for them. Because you trust in the brand. You trust in the name. Because really, if they screwed it up so bad, they'd be out of business. Yeah, but if you don't know – you know how to drive a car, but you don't know what the car is doing. You don't know the whole motor. I don't know how to build the car. No, I couldn't replace my car's computer.

No. So it's the same thing for a computer user. A normal computer user does not know what the computer is doing. He can only consume software and use software, but doesn't know how it works. So it's, yeah. Moserath, your thoughts? Okay, so as you can see in the chat room there, we've been discussing it as well. But a lot of it is off of it's not a matter

of really an active not knowing or an active not caring what they see is as um the previous person said they just see a product they use it and as i was saying to others does it work yes then they use that product they don't know if it's free they don't know if it's not free in terms of free as in speech, not free as in beer, as that is one of the main things that comes up quite often. But in terms of, see, call to wisdom, it's like the sell-free software, commercial software could be free or proprietary, right? It's a matter of does it work? It does.

Okay, I'm good. I'm out. They're not worried about any of the other stuff because they know full well Microsoft doesn't care about them. Apple doesn't care about them. You know what those companies care about? What companies always care about? Your pocketbook. Your wallet.

The moment that you open it up, they are happy to talk to you. You close it up, the conversation ends. You go to something like, say, Cody or or vlc they say we don't need your pocketbook the person says hey i don't need to open up my wallet awesome i'll use this yeah excellent it works excellent i'm out yeah so um this is where i find the absolute brilliant silver lining in this concept. Because so often as users who perhaps are just solely consumers of open source software, maybe you download the latest distribution, you install all the great open source software, maybe you covet online in the comment section about something not working,

but you don't very often or if ever submit a bug perhaps you don't submit a patch upstream you don't necessarily get involved in the mailing list you primarily consume but you do something else too you likely advocate you spread the word about Linux. You show maybe a friend who gets intellectually curious. And it turns out that's fundamental to the survival of free software. And so as regular Linux users, we have an absolute vital role in the entire ecosystem of free software and its continuation and survival. And I don't think it's overselling it to say that this is one of the most important roles in free software is advocacy and education. Now, how you do it, the people you do it to, that could all be argued.

That's the nuance. But there is such an important place for advocacy still, even in 2016, even when Linux has taken over the cloud and networking and mobile. This is still a fundamental issue. And there are people who are not even necessarily tech heads or technologists who can see the benefits. It just has to be a certain person who's intellectually curious about these things or has a motivation to learn more.

And if you, you know, look at Noah. Look how many people he switches to Linux in daily conversation just by like kind of sussing out their needs and yeah. Well, here's how Linux could fix that for you. Here's how it could solve that for you. Instant, right? Like you don't have to wait to talk to a sales rep. You don't have to. It's just

it's there. It's your right to have it. It's – you've always had it. You just didn't know. Also, I suppose this is very self-serving I guess. and open source and BSD and help people learn and discover more about it and make it useful in their daily lives because I really truly believe that advocacy is a crucial part to the free software ecosystem. And if you've ever felt bad about maybe being a leech like that, if you've played that role or you can play that role, you are taking part.

Now, you can do more at any time and I encourage you to do so. But there is something to that, Wes. I think that's pretty cool. And that, right there, brings us to the end of this week's episode of the Unplugged program. If you'd like to join us live, join us over at jblive.tv. Oh, when? What a great question. Oh, come on.

Tuesdays. Well, yeah, of course. Of course, Tuesdays. What time, Tuesdays? What time? You know, that's 2 p.m. PST. Yeah, yeah. You can also get it converted automatically.

You have a magic robot. It just does it for us. Calendar page, jupiterbroadcasting.com slash calendar. You can send us in your thoughts, jupiterbroadcasting.com slash contact. Submit stories and community updates at linuxactionshow.reddit.com. And if you're on the YouTubes and you made it this far, give it a thumbs up. Come on. For us.

Come on. All right, we'll see you back here next Tuesday Thank you. so we're going to start giving these away or selling them i'm not sure these look amazing yeah i think it's a super great idea i think you could probably sell a whole sheet of them i don't think see and i think she's just thinking about giving a few away at a time but i think you could probably sell a whole sheet of them. I don't think, see, I think she's just thinking about giving a few away at a time, but I think you could probably just sell. Yeah, I would. Then I'd just have them for my future keyboards. I don't want that Windows shame. Alright, so I'm trying to cover up my Super Key.

I think I got it. It's a little tiny bit lopsided, but that's fine. That's my fault. My Super Key actually had a circular indent to begin with, so I'm pretty happy. It's not fault. So mine's – my super key actually had a circular indent to begin with. So I'm pretty happy. It's not great, but you see mine actually fits in right in there, doesn't it? Wow.

Yeah, that's nice. So yeah, that's our super key. Yeah, yeah, yeah. Yeah, so how does it look on yours? Actually not too bad. Not too bad. I'd say it looks as good – I'd say it actually looks better than the Ubuntu sticker that's on the Librem 15. Yeah.

So, I wonder how it would look if your keyboard lit up. Oh, your keyboard does light up. Not too bad. It looks fine. Good. I was wondering if it would shine through, but yeah, it looks good. I like it. Yeah, it does have a little bit of a shine, I'd say.

Ah, very good. Very good. So you approve then? Oh man, this is great. Everyone should get them. It's funny because, you know, when you think stickers you approve then? Oh, man, this is great. Everyone should get them. It's funny because, you know, when you think stickers, you think big stuff. But that little tiny Super Key sticker is legit. And I hate seeing that little Windows key.

I mean, I like, you know, most of the keyboards that I have, they don't have them. I know. You can't do anything about most of the laptops. I wish, you know, Microsoft, they really, they really, I wonder how they pulled that off. I would like to know the history of the Windows key and how they worked. What I envision is they nailed a couple of manufacturers, and then the industry just decided to take off with it. It was easier to just have it all under.

Yeah. Oh, look at that. DOS Keyboards sells a Linux keycap bundle. That's awesome. Excellent. Seems to be a lot of people are getting into replacing, getting keyboards where they can individually replace the keycaps, which is none of my keyboards I can do that with. But I don't

like any of my keyboards a whole lot except for my Logitech upstairs. Okay, yeah. But like, that seems a lot like a lot of fiddling. But at the end you end up with something you've crafted, you have great pride in it. You have to have a lot of, you have to have some free time though. That's not something you have a concept of.

Oh, I see. Theoretically, I can do that with my Gateway laptop. Really? It has keycaps on there? Huh. Yeah, I learned that the hard way about six years back when I went into a fit of rage and then

accidentally popped the U-key loose from the keyboard. I had to quickly put it back on there before anybody noticed. We got 10 Fridays left or 9 Fridays left until Christmas, so you guys have enough time to get me a mechanical keyboard for Christmas out there. I kid. I would not.

Actually, I would probably use one, but not in studio. You'd never get to see me using it because it'd be too loud, I would assume. Get one of the quieter ones. But then what's the point? Don't they not feel as good well i mean there's there's many different actuation level you know you can get lots of different feels uh i have a code keyboard at home with the clear keycaps which i like a lot they're not too loud i mean i don't know if they'd be quiet enough for the studio but i bet you could find something and you can still be you know a lot of times they just have nicer construction anyway okay so oh i'll wait or they have you know a deeper so are you are you telling me that

i can my understanding is that if you want the right feeling like you want that old mechanical keyboard feeling from back in the early 90s 80s and 70s or whatever the hell you know almost like clunk clunk clunk i thought in order to get the right feeling, you had to have a certain amount of sound. I thought there was, you can't divorce the two, I thought. So I thought if I, as I begin to compromise on the amount of sound, I also compromise on the feel. Is that true? I mean, I think if you're trying to exactly recreate like the IBM Buckling Spring type switches, then yeah, maybe you do, but I think there's a range of things that people like

and that people appreciate and for different purposes right like a lot of gamers like the cherry mx red style series where it's like a linear press um so i think there's a wide range you might not find one that's a you might find you know like a loud one that you like more but i think you can find something that's better than say that little apple keyboard you've got over there yeah although uh i mean but that's very portable. One thing I'll touch on there. So, first of all, I'm afraid of my own personality when it comes to quieter keyboards. But to the Apple thing,

I'm not a huge fan of them, but I also don't... They're not the worst. I don't hate them. And one thing that if I was an Apple user, I think I would appreciate is, I'm not positive, but I think... I'm not sure if this is true anymore, actually. Maybe it's not true anymore, but for a long time, they had the same keyboard layout across the laptops and their physical keyboards, so you could interchange, I mean, it was that

muscle memory worked across a range of the products, which is a great idea. But going back to the mechanical keyboard thing, and I don't know if that's true anymore for Apple. Here's why I don't want to get quieter keys. This is the type of person I am. I have two modes. I don't care about the quality of this thing, or I care about the quality of it. And if I care about the quality,

and if I don't, what I'll do is I'll first buy the low-level one, then I'll buy the mid-level one, and I'll try to stay there as long as I can, and I will guilt myself. I will internally beat myself up for wanting the better one when I have already bought the low one and the mid one, and I'll fight with myself. And it'll be like this ongoing thing that steals brain CPU for days or months. You're constantly assessing it. And depending on the price, sometimes a year.

And then I finally make the go, and I love it. I love it way, way more. Like, for example, perfect example. I bought the the rover, right? It was a great price. I got it on I forget consignment. So I got a great deal on it. And it was a nice travel trailer.

It was a mid-level travel trailer. And then once I realized that I actually liked living in a tiny space and that I liked the concept of RVing and I liked how it's – I use less. I consume less. I liked all the things about it. And then it was, well, now if I'm going to do this, I'm going to do it right. I don't want to – and I didn't go out and buy like some $300,000 diesel pusher. I mean I have my limits due to budget, but I knew almost within the moment I knew I liked RVing, I knew then immediately I was going to be getting a class A RV at some point, because

I'm not going to, if I know I can do better, and that same thing applies to production. Like, if I'm working on, if I'm editing or I'm putting together something and I know it could be a little better, I'm not going to stop there. I'll go till it's as best as I can get it. And so I'm worried about with the keyboard thing is that if I start caring about that, I will really care about that. And then I'm really in a world of hurt because what do I do? Do I start replacing all my keyboards and then some keyboards don't feel right? So far, what I have allowed myself to do is I have a keyboard that I really like to use

and I would be sad if it went away. But I don't let myself go any further in fear of, have you found this to be true for yourself? Once you started looking into super nice keyboards, did you slip into a keyboard, like a keyboard snob range now? Yeah, I think I was ready to become a keyboard snob. I have not gone as far as replacing everything. I think the next on my list is getting a really nice one for work.

Well, and if you think about it, dude, like, there are a few things in my life that my hands touch more of. And, like, that's one of the things you do that you consider yourself good at, right? Like, you're supposed to be good with computers. It's a huge part of my work. It's a huge part of my hobby. It's, like, it's such a fundamental thing. Oh, God, now I'm starting

to care. And I can also appreciate that, like, it's hard to, like, you have to know, like, all right, well, if I'm going to do this, like, you can't do it half-assed, right? You're going to have to do it well. And that can also appreciate that it's hard to, you have to know if I'm going to do this, you can't do it half-assed, right? You're going to have to do it well, and that takes a lot of research and time, and sometimes it's just you don't have time for it. Okay, so somebody who must have been listening to Coder Radio, because this has been

on my mind, it looks like T3 Leighton in the chat room is asking about, I have a keyboard upstairs, this is my favorite keyboard, and it is a Logitech that runs Linux. And I can't remember if it's like a, it's a gaming keyboard. I think it might be the G15. Let me go look.

And it's an older, no, that's not it. That's not the G15. It's got a flip-up LCD screen. And it's super old. I actually did, I'd actually used it as a runs Linux once on last. Because, I mean, how could I not? The damn thing actually runs Linux with a little flip up screen

and it's supposed to have like these widgets that interact with your computer and there is some Linux software. So here is a version of it. Nope, that isn't it either. But it looks kind of like this with an LCD screen. It's a Logitech. I can't remember.

But it is, it has a great feel to it. It's not all the way crazy. But I really like the position. Here it is. This is it right here. I really like the positioning of all of the keys. I like the – one of the things I really like about it is it has this roller knob right here that's sunken into the keyboard. Is that for volume?

Yeah. That's nice. And then a dedicated mute switch and then logical playback controls that are laid out like they normally would be on a player device with play, pause, and skip tracks right there. Like you'd find on your stereo or whatever. Yeah. So this is the Logitech G19. And I don't even think they – oh, they make newer ones. Interesting. And I don't know about any more, but the version I have, which does run a little warm, has a Linux computer in there to do all the little widget-y stuff.

That's awesome. Yeah. Some ancient old version. Oh, God. I'm sure. I can only imagine. It'd be fun to find out. Let's see.

The new version, oh, they don't even tell me the price because Amazon doesn't sell it directly. But it looks like, oh, you can get it used from a reseller for $100, but new, it's $280, $290. Whoa. Here it is. Yeah. I think I remember spending quite a bit on this. But, you know, looking back at it now, I have had it for a really long time. I mean, I bought the first version years ago, and so I think the investment paid off.

And so I think the investment paid off. And here's the idea, is while you're working on the LCD screen, you're looking at your CPU usage, you're looking at your memory stats. Like it's a little mini top screen. I do like that. That's neat, right? Or temperatures. And of course they have a whole bunch of shitty software that makes it work automatically with Windows. But there were ways eventually to get it to work on Linux. But you know what?

I probably haven't bothered setting that up in two, three years. I could see that really handy if you could just redirect one terminal's output to it. One of the other things you could redirect to it was VLC. So you could redirect VLC playback on the LCD screen. Wow. Yeah, which was so cool for just watching a little video or something.

Also playback, you could see the track information, stuff like that, Which was so cool for just watching a little video or something. Also, playback. You could see the track information and stuff like that. For maybe if your music application supported that. It was nice. Notifications could go down there, too. But again, I don't even know if it works anymore. Because I probably haven't bothered setting it up for two years. Is the version today I just duct tape my phone to the top of your keyboard?

Actually, dude, that's exactly what I use that screen for a lot. It's because the screen rotates up and down. So I have it pivoted up and I lay my cell phone on it. And it's the perfect cell phone holder right there on my keyboard. Oh, man. That's great. That's funny you say that. Well-designed technology.

Hello. May I jump in here? Yeah, yeah. There is indeed a software again, GNOME 15 it's called. Before there was the GNOME 15 daemon, but that's not maintained anymore, so now they use GNOME 15. You can have a look at

GNOME 15 org, but you have to compile it yourself. It's not in the repository of any distribution I used. I gave it a try. I have a Logitech speaker. The Z11 or Z... I don't remember.

Z11, I think. And you can use the GNOME 15 software to display the same information you talk before cpu usage and all that stuff i do like that i do too yeah that's quite cool you'll be you'll be the coolest geek project ever and i wish i could remember stuff when i say this uh would be to have a way to wire out you ever watch star trek voy? Absolutely. Do you remember when they built

the... Are you making a bio-neuro gel pack over there? Good reference, man. No, do you remember when they built the Delta Flyer and they put the physical toggle switches in there that were like the ones from their holodeck program? Yep. And the Captain Protons type stuff. And it was

this 24th century technology that had a... What we would see as touch interface right yeah normally would yeah it had more like a 1920s style physical toggles and stuff like that i would love to take like from a like a old old used parts store something where you can get like voltage meters and stuff and and be able to hook those up to an interface on a linux box that would correspond the uh the little needle to CPU usage. So I could have physical boxes with needles moving that represent

CPU usage. I think that'd be so cool. Or like NixieTube spelling out things in widgets. Oh, God! I'd totally rather actually go NixieTube because then you can actually change up what it is a little more. Yeah. That'd be really cool.