LINUX Unplugged - Episode 145: BuzzwordFS | LUP 145

Episode Date: May 18, 2016

You're insecure unless you're running one of Greg’s Kernels & we think he’s right! Plus openSUSE chairman Richard Brown stops by to follow up on not shipping ZFS in openSUSE which leads to a passionate discussion. And the simple thing we could all be doing to improve open source, but maybe we’re all feeling a little too entitled!

We should probably get started here in just a second, but I thought this was really awesome. This was tweeted to me. It's the hybrid tube amp for the Raspberry Pi. It's on Kickstarter. They had a goal of $20,000. They've already raised $89,229 with 20 days left to go. 648 backers. It is a monster tube amp that you slap onto the top like a hat to your Raspberry Pi.

Look at that thing, Wes. That is neat. Now, I know audio from production standpoints, like through the regular interfaces or uses the GPIO pins to like- Oh, neat. Oh, yeah, yeah, yeah. So you bring that sweet tube sound to the Raspberry Pi, it says. And I'm not a really like super high-end audio guy, but I could kind of tell you a

little bit about it. This crazy thing has a PCM5102A DAC that drives a single 12 AU7 tube gain stage with a solid-state Class A output. It's designed to drive headphones from 32 to 300 ohms quite easily. In other words to say, pretty nice piece of kit. Yeah. And the pricing is getting up there now because of kit. And the pricing, you know, is getting up there now because a lot of the different slots, but right now you could get them for $150. I would never do something like this.

But here's what I freaking love about this. First of all, super high-end, crazy, weird gear being built for the Raspberry Pi. That's next-level stuff, man. That's next-level stuff. I think this is super neat. I'm not going to back it, though. I do like the idea that it's almost the size of just a little USB DAC,

but it's also running Linux, and do whatever you want on it. Boom! Raspberry Pi not included. This is Linux Unplugged, episode 145 for May 17th, 2016. Oh, welcome to Linux Unplugged, your weekly Linux podcast that's detecting more hops than it thinks it should. My name is Chris. My name is Wes. Hey there, Wes. You know, I am really looking forward to today's episode.

A couple of fun things happened on Sunday's show, and one of those things we're going to rip out and expand here in the show, because we've gotten a lot of feedback on it. I love it when the community sort of, like, gives us a feedback loop that feeds right into the next show. That's why we're here. We're riding that feedback loop right now, and we'll be covering that later on in the show.

First, though, some interesting updates and tidbits from The Linus. that feedback loop right now and we'll be covering that later on in the show first though some interesting updates and tidbits from the linus actually some really good like serious nice insights from linus about security it's like getting copy with linus right security in the internet of things which he actually had some really sane thoughts on uh later on though greg kh your buddy he says you are insecure unless you're using his kernel and only his kernel. So it's a bold statement, Wes. We'll tell you which kernel that is and why he thinks that. And he's got both Red Hat and SUSE to point the finger at.

He's wagging his finger. And then after all of that, we'll finish it up with a rather kind of urgent security thing going on. On two fronts, one with Mozilla and one with TeamViewer. So if you are considering remote administration on your computer or have TeamViewer in production, like I know Noah does, you definitely want to stay tuned for that. So it's a big show, big, big show. Before we go any further, though, we must bring in that virtual lug.

Time-appropriate greetings, MumbleRoo. Hello. Hello. Hello.umbleRoo. Hello. Hello. Hello. Hello. Hello. That makes me smile. Wow.

Wow, that is a heck of a showing. And I'll tell you what, kind of a surprisingly good tone from all of you. We should consider doing a sing-along after the show. See, Linux users are nice, Chris. I told you. Yeah, although one voice that will only be joining us from time to time. We should consider doing a sing-along after the show. See, Linux users are nice, Chris. I told you. Yeah. Although one voice that will only be joining us from time to time. I secretly use Arch Linux.

Popey, he's not in there, I don't think, because he's down in Tejas for OSCON, which is going on right now. Boy, OSCON, kind of a rough schedule timing because they're scheduled right up against Google I.O. next week. Or tomorrow, I should say. So that's a tough one. All right. So before we go any further, Wes, I did mention the hops. I got to talk about it. I got to just say, I got to talk about this for a second.

The Triple Seven from Scuttlebutt. This is a Belgian-style ale that was literally brewed and bottled 20 minutes from the studio, right on the waterfront. It's gorgeous. This is a nice one. It's a dark gold, slight copper kind of color, with an aroma that's got delicate cloves, banana esters, whatever the hell that is. I do get the cloves, though. It's kind of spicy.

Look at this. That's exactly what – yeah. And also hints of white pear. Damn. I also did not realize it was 9% when I bought it. Yeah. It's a stronger one. It's got very smooth and delicate and it finishes dry.

That's true, though. With a hint of tartness, which I'm not going to lie, I'm definitely picking up on. And the first time I read that, I thought it said a hint of fartness. It does not say that. And I have not even finished the first one. So it's going to be a good show, Wes. It's going to be a good show. One of the things that I was playing on the live stream, unless it's copywritten because

then I wasn't playing it. But if it's not copywritten, one of the things I was playing on the live stream, unless it's copywritten because then I wasn't playing it. But if it's not copywritten, one of the things I was playing on the live stream before we started the show was a little sit-down that Linus did at the Internet of Things conference, which is actually kind of a big deal on the front right there. The fact that Linus is even at an Internet of Things conference says something, right? Yeah, definitely. And so he's there. Oh, whoa, whoa, hi there, hi there.

You know what I like is when I accidentally scroll when I have my mouse up on the tabs. He's at this Internet of Things conference. And he had a whole range of things to say. And the whole video is embedded right there in that link in the show notes. So if you want to get that, you can listen to the whole thing. It's like 30 minutes long. Pretty low key, but it's a good listen. Yeah, yeah.

He said, so Internet of Things was obviously the main topic at this conference, and the top interviewer was Linus himself. And he said, maybe you won't see when asked about – I'll give a little context here since we're not going to play the clip. When asked about what's Linux's competitive advantage on those really far end, like, you know, sensor devices, the ones that are running other embedded OSes now. And he's like, I don't really know if there is a competitive advantage for Linux there. But he does think that Linux will be the central hub that all those things connect back to. Yeah, I think part of his point is that there's, sure, there

will be small devices that don't need Linux, but they will all be talking to other smarter devices that you will need the kind of capability that Linux provides. Okay. So then he got all Zen master when talking about security. He was, you know, the problem was, and he even, the interviewer brought up the fact that Bruce Schneier wrote an article saying the Internet of Things is unpatchable.

It's busted. It's busted by design. It is not patchable. We are screwed with the Internet of Things. And so, Linus was asked, do you basically agree with Bruce Schneier? And he said,

you know what? I don't worry about it. I don't worry about it because there's not a lot we can do. It's unpatchable. It's just a fact of life. But that's why he thinks hubs running Linux

will be even more important. Your thoughts, Wes. Your thoughts. Is it unpatchable, Wes? I don't know about unpatchable, but I think he's right. Yes, it is. It is pretty much unpatchable. There are ways to deal with it,

but clearly no one is interested in pursuing them and there's no money being spent doing it. So the problem is you have two scenarios. First problem is you either have the up to the user, which then it never gets done. Yeah, absolutely. Or you force the update on the user, and then you have a problem. Like we were chatting in our production chat this morning. Angela's grandmother has an old computer that had, by default, the express settings were turned on and recommended updates were turned on.

And this morning, her computer automatically installed Windows 10, rebooted, and then just fails to boot. Like it just fails at boot up. And now, you know, her computer for her, as a woman who I think is maybe in her 80s or near it, her computer is essentially ruined now. Just doesn't work. So automatic updates, obviously, is not a way to go either because a company with billions of dollars and literally tens of thousands of developers can't get it right. Billions of dollars and literally tens of thousands of developers can't get it right.

I would hope that these smaller platforms with more rigidly controlled hardware will be easier to target for automatic updates, especially if you don't allow any or very much user configuration to make it a more varied platform. But still, you have a very good point, and it's no good if your product goes down. You like it, or I mean, I'm sorry, North Ranger, you think corporate device builders, device builders, like your hues out there, they think what Linus is saying is music to their ears? Define what you're saying.

Explain. Yeah. So, I mean, they love Linux because it makes their jobs easier to bring new devices to the market quickly. But ultimately, they would rather you throw that device away in two years and buy a new one. So who cares if they get updates for a long, long time? So for the device makers out there that want to sell you a new device, Linus not caring about security or not caring about continuous, plays right into their narrative.

MiniMC, I like your point, too, about perhaps this is where Ubuntu Snappy Core or whatever could come in and maybe provide a solution that's actually manageable, right? Yeah, as far as I heard, this was the bright version of Ubuntu to being able with Snappy to upgrade the operating system of these devices. They're able with Snappy to upgrade the operating system of these devices. And supposedly, excuse me, supposedly, Canonical is working with Hawaii and a few other vendors out there, and they're going to be demoing SnappyCore on a device at OSCON this week. Oh, really?

So we should probably have more news by last, I would think. That's awesome. So, yeah, maybe they will save all of us. You sound so hopeful. Speaking of Linus, Linux 4.6 came out today, was it? Let's see, yesterday. No, no, two days ago. And it's got a couple of big things in it. USB 3.1 super speed is 10 gigabit support.

Out-of-memory task killer has been improved. Orange FS has been included. Yeah, that's an interesting one. Yeah, a distributed file system. And the Batman protocol got an update, which I'm always a fan of, tracking Batman. It's actually, there's just actually a metric S ton of updates. I mean, we don't very often mention kernel versions when they get updated on the show. But every now and then there's a doozy.

Or as Linus puts it, a weasel, which is explained in that video. Every now and then there's a weasel that comes out that looks like a real good one. Unless you're Greg KH. And then he says if you're running anything but his. You know, also in here is like the USB 3.1 support. Yeah, yeah. That'll be important. Yeah, so that's called the Super Speed Plus protocol, which is the 10 gigabit speed for OOPSP, which is really great.

Yeah, go ahead. protocol, which is the 10 gigabit speed for OOPS, which is really great. Yeah, go ahead. So in here, actually, it's because Linus has mentioned many times before that he doesn't look at security like the security experts usually care. He thinks it's okay and good that people care about it, but he personally doesn't even see it as a big threat. His software will change and accommodate depending on the scenario. And security is usually much more about the work environment than necessarily the technical feat of the things. And security is usually much more about the work environment than necessarily the technical feat of the things. He says that he's concerned.

While Greg, on the other hand, he tends to listen a little bit more on the security experts because he handles the stable kernel. So his concerns are also a little bit more in line with that vision of what's secure. He was actually just recently interviewed, and he kind of outlines his thoughts on it. So I'm actually, I have that story in here. We'll get to that in just a second.

We'll cover that. In fact, why don't we, let's just talk about it right now. Wait, wait. So before we do that, just really, really briefly, I will mention our friends over at DigitalOcean. Head over to DigitalOcean.com and use our promo code D-O, unplugged, all one word, lowercase, to get yourself a great deal. DigitalOcean is Linux over to DigitalOcean.com and use our promo code DEO, unplugged, all one word, lowercase, to get yourself a great deal.

DigitalOcean is Linux rigs up in the cloud. Now, we all know the cloud means somebody else's server. That's legit. Let's be honest. That's what it means. That's why you've got to make sure they're damn good servers. Like with data centers all over the world where you want them. Have you seen those pics?

Oh, dude. DigitalOcean's data centers are beautiful. Yeah, they are. Seriously, as somebody who used to rack up servers, I look at that and I go, wow. They're better than me. They're way better than me, dude. They have great infrastructure. Linux throughout the entire stack.

It is SSDs for all of the disk I.O. Great CPUs, tier one bandwidth, $5 a month as a starting price. You can get started in less than 55 seconds. You get 512 megabytes of RAM, a gigabyte of – a terabyte of transfer. There we go. Those are the units. I know. There's so much.

I mean it's hard to account for. I can't believe this. Like I've had half this beer and I think it's already hitting my head. I tell you what. Because, you know. I think DigitalOcean is just such an amazing deal. You're not ready for it. Well, let's be honest.

I've upgraded recently to the $10 machine. Oh, I use the $10 too. It's nice. It is. In fact, the pricing structure. You can do whatever you want on a $10 rig. And really what I have done is I started at $5 and I've gone up to $10. And that's where you get the one gigabyte of memories at the $10 rig.

And I love that because then I get 30 gigabytes of SSD, two terabytes of transfer. But they say their most popular plan is the one that has two cores. I don't know. You know, one of the things I really like about DigitalOcean is their super nice dashboard makes it very clear to monitor your CPU usage. And in the early days of my DigitalOcean usage, I think I was coming at this from a perspective, I need way more RAM, I need way more CPU, and I was overbuilding my machines. And I would go look at the charts and go, all these things are kind of just sitting here.

And so I started to realize that when you have unbelievably fast bandwidth, and when you have all SSDs for the entire stack, and you're running a low, small install of Linux, it doesn't have the graphics stack, it doesn't have any of that crap, you can get away with, I mean, it feels like you're stealing it. It's so cheap. It's so great. And so, for me, the $10 plan has really been the sweet spot. They say the

$21 is their most popular. The $20 plan is their most popular. But $5 and $10 has been solid for me. And you know what? Whenever we need a new rig... See, for me, I like the $40 for build machines. Really? You were talking about, like, oh, DKMS, you've got to compile the ZFS module. But what,

are you doing it hourly, or do you run them all the time? You can just, with the API, spin up a new machine, run your script, grab the compiled module, shut it down. Yeah, that is something else we're doing more now. It's so nice. We are spinning machines up and down for shows now that we don't need running when we're not live.

It makes it very economical. It's only there when you need it. They're honey badger bad. That's why they created a really good down for shows now that we don't need running when we're not live. It makes it very economical. It's only there when you need it. They're honey badger bad. That's why they created a really good API for all of that. They also have great documentation. They have one-click installations of apps and all of the OSs. Like core OSs. We'll be talking more about the opinion of— What's that one on the bottom right right there, Chris?

Previsited? Is that how you say it? I think that's based on Slackware. I'm not sure. Yeah, I'm not sure. I think it's based on Slackware. They also have Debian, Fedora, CentOS, and Ubuntu. You can check it all out.

They have one-click applications of just a single app or the entire stack. It is sweet. DigitalOcean.com. Check them out and use our promo code D-O-Unplugged. Wow, I'm glad you brought more than one of these beers because I think I'm going to have another one. You know what I should do? Okay, just before the next ad, I'm going to finish this beer. And we'll just see how I do.

We'll see how I do. We'll just see how we do. Okay, so you heard Mr. Devlin begin the conversation. Let's pick that right back up. Linux can't keep you safe if you don't update it. Pretty, pretty obvious coming from Greg Cage, but it goes further than that. So there is a Core OS Fest in Berlin. Man, OSCON, guys, I've been going to OSCON for years.

You guys are great, but I'll tell you what. You're up against Google I.O. and Core OS Fest in Berlin. That kind of sounds fun. And Greg's over at Core OS Fest because he's involved with that project, like I was just saying we'd be talking about in a moment. And if you're not familiar, we have interviewed him in the past. He is the Linux kernel developer and maintainer of the stable branch, or sometimes called the LTS branch,

as well as the USB subsystem in the Linux kernel and many other important subsystems that people just frankly didn't want to do. So good guy Greg picked him up. And he was saying, for the last 15 years in the Linux kernel community, they've been following a rule to fix things as soon as possible. The Linux community fixes bugs and then pushes them out so the vendors can then push them to their users. He talked about, for example, a bug that they fixed

and then pushed the new release and got it out there. But then three years later, someone realized that there's a security bug. You could go to a local root user and you could go away. You could just go run around. Turns out Red Hat and Seuss had to go back and fix all of their old stuff. We had a very bad history of keeping bugs alive for a long time. Somebody did a check of it. And most known live bugs in the last five years in these systems. There are things that people know and they know how to exploit out there. They're not closed.

And that's a problem for infrastructure. Now, it's kind of a bit of a ramble. But what he's saying is if you're following a mainline secure patched kernel, you're going to be safe. But if you're using like a Red Hat Enterprise kernel, they're not necessarily taking all of the fixes. Following all the hard work that Greg does to put those patches into there. And you assume they are, but he's saying they're not.

In fact, a lot of the vulnerabilities in the last five years have been ones we already had patches for. They just haven't been back. Oh, that's rough. He says Android is another example. What does he say? We have a very bad history of keeping bugs alive for a long time. Yeah. That's not what Linux should be known for.

Yeah, so he points out Android also has issues. Yeah. That's not what Linux should be known for. Yeah, so he points out Android also has issues. He says, basically, direct quote, your machine is insecure unless you're running MyKernel, or based on MyKernel, or at least based on another one. If you're not taking these fixes, then it's insecure.

He has a very clear message for companies and communities building products around Linux. You have to be able to update your machines. You have to be able to provide a system where your machines are updated and you constantly take advantage of it. Man, that is so true. That is so true. The LTSs and

the Red Hat Enterprise Linuxes out there are great and super stable platforms for enterprises to target, but if they're not pulling in the fixes that Greg's putting in, then we are not taking advantage of one of the fundamental features of an open source operating system. Well said.

It's like a commercial model applied to something that's a moving target. And he's building these LTS kernels that aren't updated as frequently as the main kernel. In fact, it's mini. Right, they should be a pretty easy target, hopefully. And they're updated and they're maintained for years. Derek Delvin, I wanted to have you pick it back up. Go ahead. it hopefully and they're updated and they're maintained for years derrick delvin i wanted to have you pick it back up go ahead basically i think this is the equivalent of greg's uh fu

to nvidia after lena said uh basically and maybe spur attention to the issue from the distros by having the pressure now i hope that that's the intention there So you're saying this is his middle finger to the overall industry? Basically. In a way, what's the effect it's going to come out? He's not winning anymore for people using his kernel. Basically, he's just saying he started up in the patches in my work. I've been working on this. Yeah, and I guess the thing about Greg's kernel too is it's not updated as frequently as the mainline kernel.

So there is a lot less cycle there you have to follow. So you're on a Pricity right now. Yes. If you check for updates, do you have a kernel 4.6? It came out two days ago. I'm just curious about that because I think Arch got it pretty quick. Because I think Arch got it pretty quick. Apricity, Wes, if I recall, it's Arch-based, pretty limited customization.

Reminds us a lot of Anagross. It does, yeah. You've stuck with it for a couple of weeks now. How come? Well, it's been working well. I'm not seeing. Maybe you already have it. Maybe I do. No, 4.5.4-1.

Okay. So you just checked, you just did a package update? And I have not yet looked what's, they give you one custom repo. I have not looked at everything that's installed there. I don't know if my kernel's from there or not. So I have, yeah, I have Linux 5, I have Linux 4.5.4 now. Okay. So I'm just curious overall, Apricity seems like, so you had something, you had what, did you say? Do you remember which kernel? 4.5.4. Okay. So I'm just curious, overall, Apricity seems like...

So you had something... You had what, did you say? Do you remember which kernel? 454. Okay, yeah. So it seems pretty up-to-date. Not quite 4.6 yet. Right. But what are your couple of weeks in thoughts now with Apricity? Just like as a super quick follow-up review. Yeah, I'm enjoying it. I've had Ubuntu 16.04 on here. There's Linux Mint on this laptop. Apricity...

Antigros has been on here in the past. And I've always enjoyed Antigros. So Apricity seems like a little bit of a competition. They have a kind of like stripped down, or at least so they say, version of GNOME. So it's pretty lightweight. They don't install all of the GNOME apps or all the dependencies. In fact, when you sat down on the laptop at first,

I was like, is that elementary OS? Yeah, you know, it's pretty clean style. It seems to run well. And I have access to Arch, so it's familiar if I need to install things, all the configuration is the same as upstream. And they come with the TLP power saving tools

by default, which in my experience have done a very good job. Oh, it also has the nice power line terminal. Oh yeah, that is cool. That is a cool feature. BI and Bash and ZSH. So Appricity gets the Westpain nod of approval. I don't know if I would install it.

I'm building a nice new desktop. I don't know that I would install it as my be-all, end-all. But as a laptop I use for the show and a few other things, it hasn't been bad. Oh, nice. It might also be my, you know, it's nice to have an Arch graphical install just as like something where I can use Gparted if I need to or Rescue CD. So, I can see Opricity being that as well. Yeah, cool. There you go, a little quick update on Pricity and if you guys are looking for something that does stay fairly

current but is interesting, it might be worth a look. I kind of want to try it on a machine. I do not like this. We have been covering different iterations of the FBI and their attack on Tor and Tor developers now for a couple of weeks. And this feels awful.

So the FBI has a case where they used Tor, an exploit against Tor, which now appears to be potentially an exploit against Firefox, to get this guy. And Mozilla wants the FBI to reveal the details of the exploit ahead of the trial because if the FBI refuses to do so, which so far it has, then it would be revealed publicly in the case, which would mean that anyone could just jump on an actual in the wild exploit right now. That's bad, bad stuff. It's an interesting argument. Yeah. Especially since if it is going to be in the public domain or the public record anyway, I wonder if they have – what their case even just internally is for not doing it. It also ties into like obviously the FBI knew about it before and is willfully deciding

that it better serves the public interest for them to keep it secret than to share it with the public. So where does that fall now? So does anybody in the Mumba room have a good argument for why Mozilla needs access to this information before the trial since it's all open source anyways? Go ahead if anybody does. I have one. Yeah. In the end, having more information will be useful if not only to construct a good case. Well, I guess. So why hold it so close? So is the FBI – I guess here's another way to ask the question. To me, it seems obvious. Mozilla wants this information so they can get a patch out there as fast as possible for its public information. It's in the public domain.

possible for its public information. It's in the public domain. The FBI, though, why are they not revealing it to Mozilla privately? Couldn't they do it like with an NDA or something like that? Like, why? What's the motivation there? That I don't understand. Right. At least let them patch so that they can have something ready to ship when it is released publicly. Seems like... Well, but then you are giving preferential treatment, right? As a public entity, in a way, this comes a lot as if you start being okay with private talks, the transparency that they've already been accused so much of, you know, will further damage their reputation. And so...

Yeah. Yeah. To me, it seems like the FBI goes on and on all day long about how they want companies, private companies to work with them to save us all from terrorism. But then as soon as private companies want to protect citizens from cyber attacks, the FBI don't have time for you. It's just kind of a little – it sucks. I'm glad Colonel Linux just made it here because I actually put this story in the updates just for him. There's been a sudden increase in TeamViewer account compromises, even TeamViewer accounts that have long, complicated passwords that have only been used on that TeamViewer account. I guess there has been a wave of TeamViewer hacks. Mr. Colonel Linux, are you aware that there's two-factor authentication for that TeamViewer?

I was, although I don't use the account feature of TeamViewer. I think that is primarily reserved for companies that use it as their primary source of remote administration. And the thing is, for me, it's a way to help mom with their computer. It's a way to help dad when he needs something and I'm away or a couple friends. So if I have a TeamViewer account, I certainly don't use it. And if you sign in, then you can save all the computers you connected to. I'm asking for one-time passwords every time i'm doing that right so i'm getting the team i'm getting the partner id and i'm getting the password i'm fixing the problem and then i'm done with the problem solved all right then my uh my next story was also put in here for you really showed up just

at the right time it's kind of magical i plan these things so the question has been asked has the ubuntu 1604 network manager been fixed? And I started digging around in the bug, and it is super depressing. First of all, the answer is no. The bug has not been fixed. And second of all, the answer over the Launchpad bug was to go file a GNOME bug. And then when you go read the GNOME bug, the gnome bug completely misses the point. They start thinking it's like some sort of DNS resolve issue. Everything is lost in

translation. Kernel Linux, I know this has been sort of the number one, I think, the number one bug you've been struggling with in deployments. Have you found a workaround for this particular issue? This is the one where you wake up from sleep and all of a sudden network manager's just gone. Oh, I've had that problem too. Yeah. Yeah, so let's be clear. There is no deployments. Noah has not done any deployments with 16.04

because that is a show-stopping issue. In fact, so much so that I don't even use it on my laptop. I can't because I need to be able to access a network when I bring my laptop out of sleep. And restarting every three or four times just isn't an option. If it has been fixed or somebody has done anything about it, I haven't seen it.

I might have a workaround for you. Okay, that would be good. I got that problem on my Ubuntu laptop as well. I had to turn off and turn on the cart with ifconfig to get it to work. I've had to reboot to get it to come back. So it looks like the command line utility nmtui-connect you should be able to come back. So it looks like the command line utility nmtui-connect

you should be able to use that. You were also mentioning talking about using a systemd what was it? Systemd networkd. Tell me about systemd networkd, Wes. You will need, obviously then, some other thing to manage the wireless side because it'll just handle the link side, not the wireless.

But if maybe what I was hoping you would say is say you were just going to connect to Wi-Fi and it was the same side, not the wireless. But if maybe what, okay, so what I was hoping you would say is, say you were just going to connect to Wi-Fi and it was the same network, nothing had changed, that you could just use that to bring the link up and maybe it would just connect to the same Wi-Fi network. Yeah, you'd probably need WK supplicant, but if you're doing the same network, then absolutely, that's a good case.

So, system D, network D. Okay, that could be another way to get the interface to come up, if you're just connected to the same Wi-Fi. If you have simple needs, yeah, or Conman is another popular. This is a real bummer, though. This is a major issue that we have gotten a lot of people that have written in about, too. It's just, oh, man. It really is unfortunate that this has sort of been punted around.

Blaster, you had something you wanted to add. Yeah, go ahead. Yeah, so when we ran into this issue on the XPS at Linux Fest Northwest, actually, Paige busted out Network Manager CLI and fixed it that way. Yeah, so there's another potential. That might be another workaround for you at Kernel Linux that isn't too painful to use. Yeah, so I'm just going to throw this out there. You know what another workaround is?

Just use 14.04 because everything still works in it. Ha, ha. Super, yeah. Yeah okay okay sure you can also distance yourself further from greg's kernel uh-huh yeah i yeah exactly yeah i'm just to me that is just such not an answer it's just it's it's not even an option for for me i'm sure it is for some people but in 2016 to be told to install 1404 that an option for me. I'm sure it is for some people, but in 2016, to be told to install 14.04, that's rough for me. That is rough.

That is rough. Especially on something where wireless is relevant, where I probably want newer packages. Yeah, okay. Preach it, Wes. Preach it, Wes. Preach it, Wes. Especially if you have yourself a brand new laptop. Quite literally, I don't know, but somebody could write in and go look at the Oryx Pro specs

and see if there's features that aren't being lit up by that old kernel. Oh, that would be fun. Yeah, I wonder if like the internet. They'll send it right to kernel Linux. Those new later generation. What's that thing got in it, kernel Linux? A sixth generation Intel? Which one now?

The Oryx. The Oryx is sixth generation, yep. Yeah, I wonder if, although they do do hardware enablement in 14.04, so it's possible. Yeah, it's possible. That's true. All right. So the grand experiment continues. I am now on the second beer of the show.

You know, I have not had a beer or any other drink since the last show, so I think maybe that's why it's hitting me so hard. Oh, wow. Yeah, because it's just, so anyways. I cannot claim that, folks. So let's see if we can make it through the ting spot. Are you ready? You want to see if we can do this?

Let's do it. Let's just see if we make it. Now, everybody, everybody right now, stop what you're doing. Are you driving? Doesn't matter. Pull over to the side of the road. Or just do it while you're driving. That's totally fine.

Wait, don't do that. Don't do that. Don't do that. Don't do that. Hold on. Hold on. Hold on. I lost my... We need our ambiance. Yeah, I have do that. Don't do that. Don't do that. Hold on. I lost my...

We need our ambiance. Yeah, I have it set to keep playing with. When the loop stops... Oh, there we go. Here we go. Boom. There we go. Okay, as I was saying, don't actually do this while you're driving, but if you're doing anything but driving right now, do me a favor and go to linux.ting.com.

Go over there to support the show and learn more about the Ting service. Not only will you get a discount, which is great, Ting's going to give you some money if you sign up, either off your first device or

just towards your account if you've got a compatible device. Oh, and by the way, you might have a compatible device, because you know what, Wes?

They got GSM and CDMA. Pretty much everything works. It's hard to find it. You have to go out of your way to buy a phone that won't work on Shoot, you know what doesn't work on CDMA or GSM?

This bell. There, I'm done with the loop. Nope, analog only. Yeah, analog only. There's no SIM slot for this bell. But just about every other device that does have a SIM slot will probably work over at Ting. Go there, check them out. They have no contracts, no other termination fee, and you only pay for what you use.

It's flat $6 for just your line, plus anything Uncle Sam's going to take. And then your usage. Way simpler, way more straightforward, no crazy gimmicks. Plus, they have really good customer service. When you call Ting, you actually talk to a real person, which not everybody believes, Wes. Some people think it's not true. Thanks for calling Ting. This is Isabel speaking.

How can I help you? Hello? Thanks for calling Ting. This is Isabel speaking. How can I help you? Hello? Thanks for calling Ting. This is Isabel speaking. How can I... Hey, this is Isabel Ting. For service in English, press 1. Para el servicio en español, presione 2. To talk about phones, press 1. To talk about something else, press 2.

To speak to an operator, press 0. Please hold while we direct your call to the first available agent. Your time is important to us. Thank you for your patience. Amazing. Your time is important to us. Thank you for your patience. Ba, ba-da, ba, ba-da, ba-da. Amazing. Ba-da, ba-da, ba-da. Linux.ting.com. Go there.

You can talk to real human beings. They've got great dashboards. They're also doing a giveaway of the Galaxy S7, which I have a feeling after tomorrow's I.O. keynote is going to be a VR powerhouse. But you can find out more about that at their blog. Just start there by going to Linux.ting.com. Linux you can find out more about that at their blog. Just start there by going to linux.ting.com. Linux.ting.com! I made it to that ad, too. Well, except for I burped at the end.

Linux.ting.com. Well, see what happens when you bring in 9% whatever it is. A show happens. A show about Linux. This was supposed to be a Windows program. So there was a conversation that we had towards the end of last week's Linux Action Show about entitlement. And we took on the conversation of part magic. And Colonel Linux, would you just super quickly recap sort of your position on that for people that didn't catch last but assume most of them did?

So essentially software called part of magic, great piece of software, was available for free for a long, long time. Company decides that they're bringing significant value into the enterprise and they could charge a couple bucks for it and a bunch of people complained about it. So now they have dropped the price and I'm upset. You're upset that they dropped the price. I'm upset that a company that works very hard and diligent and has a proven track record in the marketplace for making quality, valuable software cannot charge a small nominal fee without the internet erupting into constant whining and complaining about something they feel entitled to. And not only are they too cheap to pay for the software, then on top of that, they are too lazy to go torrent it, which is what they would do for proprietary software. And if that's not bad enough, then they have to go write blogs and complain and whine about it.

There's a lot in there. Oh, boy. I'm glad you're here on the Linux Unplugged couch this week, Noah, because when I hear entitlement, I'm not sure I'm hearing – I don't think of the same thing. When I hear entitlement, I think of people who believe they deserve something even if that's not necessarily true, if the value exchange isn't necessarily there for that. Do you follow what I'm saying? Yes, I believe so. And I feel like the situation with part magic is more like you gave me something, then you

changed it, and then I freaked the F out. It's also complicated by the nature of markets in that. I mean you have to – if those are your user base, you have to – maybe they lower the price because that's what the market is demanding of them. Yeah, OK. Bringing morality into that can be tricky, whereas in the open source world, I think it's a different picture. All right, Colonel Lennox, what do you think? I think that the project that when it started back as a free project, when they were giving it away for free, was a vastly different tool and had a vastly different place in the IT sphere as it does today. I fundamentally could not do certain aspects of my job without part of Magic. And the thing that I was using it for the other week was I had a client, had a laptop.

They were sending it in to the manufacturer to get some work done, and they had sensitive information that had to be wiped off of it. Well, new SSDs, if you're not familiar with it, are all encrypted by default. And you can simply change that encryption key on the hard disk, which essentially, for layman's terms, erases the SSD instantaneously. And I wanted to access that secure erase function. And the easiest way to do that is with the tools built into part of magic. And I as a person who have used part of magic numerous times before, and I'm fully capable of torrenting it, and have actually paid for it in the past a couple of times more than once, I still paid to download it again, because I want to continue supporting that project. Are you familiar with d-ban and does that

not meet your needs necessarily d-ban will write zeros to every block right but with a solid state you actually have more space on the drive than the computer is aware of and what will happen is where leveling will kick in and it will start writing to yeah so that actually won't work so what what you need to do is the hard drive is encrypted by default even if you don't know it is it is and so if you change that encryption key on the hard disk, then what happens is essentially the data might still be there, but it becomes useless. Effectively, it's erased. And that's a feature built in at the hardware level of the SSD. Darren Devon, go ahead.

So basically, I think there is actually a benefit of people using these routes versus the other routes. What will be? They will pick up the source code and basically just lower the cost and make it freely available. It actually is a good argument against those that usually say if you put it under a free license, people will just take it out and make it freely available. People actually took the effort of saying, okay, we're thinking that you're going to want somewhat that they consider unreasonable. I don't know what the figures are, but if people that are consuming the product are looking at it and then saying, all right, you're increasing the cost and now it's a problem.

It's apparently – if there was a cost before, it means that the problem is not with the existence of the cost. It's with the increase. You want to keep support, right? But, Dar, if I pay for Part of Magic per the terms of open source, I can give you a copy of it. I can upload it onto my site and let you download it for free. I can upload it to Torrent. So the only thing you're paying for is, A, to support the company and the product that they have, but, B, to use the convenient little download link that they have on their site. Sure.

OK. So let's – I want to give Wimpy a chance to jump in here because I want to talk to him specifically from the standpoint – and I'd actually like to hear from – actually, maybe I'll start with Richard. Richard, are you on mic right now? Because I'd like to actually get your perspective too if you're – I am indeed. Are you on mic right now? Because I'd like to actually get your perspective too if you're – I am indeed.

When you hear entitlement, is this matching up with what your mental model of entitlement in the open source community is? Because I have one, two, three, four, something like that, links in the show notes. People who have blogged over the years now about how entitled the open source community is. over the years now about how entitled the open source community is. And I actually have started to feel some of it myself as somebody who creates something free and puts it out there. And at times – like there are at times things that feel like they cross the line of what is expected of me as a content producer. And sometimes that's just me being grumpy. And sometimes it does feel like it's crossing the line.

So, Richard, when you hear entitlement, what does it mean to you and the state of entitlement in the open source community? Yeah, I mean, I think you've pretty much covered, you know, what it is. I mean, you know, I would simply define it as, you know, well, users demanding without being prepared to actually help fix the issue. And I'm not saying they need to do it themselves, but, you know, there's plenty of people who want lots of things for nothing and are not willing to give back anything for it. Now, my perspective on entitlement is with my open Sousa hat firmly on, and this is open Sousa, notza not suza because it's a very different dynamic um is i mean you know i run a project where for example we don't take money and

we're not we're like we're anti-interested in it i mean it's more hassle than it's worth and it doesn't actually solve any problems i'd much rather take a contribution from somebody than any cash. And so with the whole parted magic thing, which is a totally different project, but I struggle with some of these projects where money is thrown around like it's the magical solver and supporter of everything. Because I don't believe that. I've seen far too many open source projects where a simple contribution be it code or advocacy or something is worth

millions more than my 10 bucks hmm uh wimpy do you ever struggle with a sense of entitlement in the community especially since there is people that are contributing on patreon and things like that so i've got a tangent gentle point about um entitlement which i'll come back to later okay perfect but here's a kind of entitlement talking about a bug that affects you and saying it's widespread and it's affecting lots of people and you have lots of feedback to say this is a big issue. And then 10 minutes later, I go off and find the bug reports for that bug. And I find three bug reports about Wi-Fi not working after suspend. And in one bug report, eight people have said this bug affects them.

In the other bug report, one person has reported it. No one else has said it affects them. And in the third bug report, one person has reported it. Nobody else says it affects them. Yet this is apparently a widespread issue, and I agree it's a widespread issue because I've also heard about this anecdotally as well. This is so far my exact observation of this particular network manager issue. Yeah. Now, the entitlement that people feel here is that this bug should be fixed. Right. The very least you can do is just go and click on the bit that says this bug affects me just to add to the count,

because this is what brings it to the attention of the people that are going to resolve these issues. If you're prepared to invest a little bit more then actually provide some you know details some a reproducible um test case for example which some people have done in this bug report so and i was reading another blog earlier where somebody was whining about how they've had this bug with a particular ethernet controller not working and it hasn't worked since 13-0 something or other. And they cite their own blog going back over five years, but at no point in any of their blog posts is there a link to a bug that they filed or contributed to.

So this is a kind of entitlement that really rubs me up the wrong way, because somehow by yelling into the void of the internet, you are somehow alerting people to the fact that you're unhappy that something needs fixing, but you're doing absolutely nothing to solve the problem. Right, yelling into the void. That frustrates me. Well, it seems to be also sort of not taking advantage

of a huge feature of using an open source product where the bug tracker is actually out in the open. Just as an example, when we rolled out the new OBS production rig that we covered in Sunday's Linux Action Show, when we booted up and realized that Antigross wasn't building the ZFS module before it created the init image, yeah, the nice thing we could do there is we could file a bug about that. And so that way, down the road, when we do updates, hopefully we don't have an issue. But you're right.

It seems like even the most basic, even the most basic, not even having to create the bug, just saying it even impacts you, is too much to ask. Maybe it's because you have to create an account or something stupid like that. I know, but it's like the least you can do. And also, we've got these open bug trackers in all of the distribution projects. Everyone has them. And try and do that with Microsoft or Apple

or Android even, because frankly, if you look at issues on Android, you're just, again, lots of people with the same issue talking on a Google group somewhere and nothing ever getting done about it. Now, Colonel Linux, you hear this all again, lots of people with the same issue talking on a Google group somewhere and nothing ever getting done about it. Now, Colonel Lennox, you hear this all the time. What are your thoughts?

I, you know, I try to be a good community member, file bug reports when they come up. And, you know, certainly I'm not saying that everyone is like this, but there are times when I have had an issue and I go to file a bug report and I provide the best information that I know how to provide, and the assumption is that if that is incorrect or further information is needed, that people could just politely ask for it. And more than once, I've had a very negative experience with the people that respond to bugs being upset about a bug being opened by something for various different reasons. And I think that part of it is an attitude problem that needs to change. When people are taking their time to give you feedback about a project or about software, take that, be respectful and appreciative that somebody is volunteering their time. Yes, you volunteered your time to write that code to begin with, but somebody is essentially

doing troubleshooting for you. And so if you have to work with them a little bit, come alongside them and walk with them and say, this is how I need the information to be provided to me, or this is more information I need. They need to be open and receptive to doing that and not just bite people's head off because they didn't get exactly how they wanted to hear it.

That is true as well. That does happen. That is true. Yeah, but they're really busy too. I mean, you can see why, right? They just don't have time. They just want you to get it that is true as well that does happen that is true yeah but they're really busy too i mean you can see why right they just don't have time they just want you to get it right but you do have to hit on that good like as you know as a developer you you don't want something where it's like yeah hey sometimes i get this thing fix please but well it's no go on no i didn't mean to cut you off but i just responding to chris's point of you know they're busy that's fine but the reality is if that happens and you start to and you continue to become frustrated with people then the natural reaction is and this actually

just happened i had a bug with a bluetooth keyboard and uh i went back just because i don't really care what other people think and i went and posted in the bug report and said this is how i fixed it uh and but then i went back and posted on a form uh which is where the rest of the discussion had been occurring because nobody wanted to participate in the bug report uh yeah that's what you're going to get is you're going to get people that will just go somewhere else and talk about it then uh and that's an that's an interesting point in itself because i think a lot of people have the misconception that discussing bugs in the forum is tantamount to raising a bug report and that somehow developers are going to see discussions in forums and i can with the exception of the ubuntu mate community forums which i'm

active in i can tell you for a fact ubuntu developers and the flavor developers are not looking through the forums for bug reports right there is only one place that developers are looking for bug reports and that's launchpad so if it's not in launchpad it never happened and from the developer's perspective you need that workflow like you'd go insane trying to find bugs like you need one place for your bugs so you can just go down and solve them yeah absolutely i wouldn't be you had a second point you wanted to make though um yes um this is more about um linux compatible hardware and companies like system 76 and intro where and it's a different type of entitlement and i feel that people uh go and buy whatever computer it is they get usually a cost-based purchase. You know, everyone's got a finite amount of spare cash,

so they're choosing a laptop that they can afford at a price point. But then they feel entitled that Linux should run on this computer that they bought that was designed to run Windows. And then some of those people get pissy and shitty when it doesn't work and they have a sense of entitlement. They don't appreciate how much work has gone into getting the support that we do have. I have a lot of thoughts about that too because also one of the things that fills our inboxes, in fact, I was just kind of going back through the history of Glass

just to see when our anniversary is, which is June 10th. And I was looking. You know what our first episode was about? And this is an issue that has literally plagued our show from episode one now for a decade. Do you know what episode one? I mean, it was about a lot of things. It was about the new Ubuntu release. It was about a lot of things.

But like the headline topic. Wireless support? Nope. Although that was a big issue back then. That's a damn lot of things. But like the headline topic. Wireless support? Nope. Although that was a big issue back then. That's a damn good guess, Wes. That's really close. No. Installing Linux on, oh, go ahead, Colonel Linux.

You can say it. The MacBook. Ooh. Yeah. The MacBook. Which is a big one. Right. So I want to talk about that just for a second.

So let's actually dedicate a little bit of time to talking about that. So first, that means we go to Linux Academy. Go to linuxacademy.com slash unplugged. linuxacademy.com slash unplugged where you can get their new discount, which is awesome because this of Linux to teach you all of the ins and outs about Linux and all of the really cool technologies that you can make money off of using Linux. And I'm talking everything, including even all the way out there to Android development, PHP, Python, Ruby, linuxacademy.com slash unplugged. They now have 2,391 self-paced courses. I think for me the scenario-based lab is one of the best things they've ever done.

But if I wasn't such a loner and antisocial, I would say probably it's instructor mentoring. I mean, in terms of value in the service, instructor mentoring is incredible, especially for this kind of courseware. They have graded server exercises, which is dope. Automatically grades your actions. Why not? Why do you have to wait around? Check them out, linuxacademy.com slash unplugged.

And I don't normally mention it, but Mr. Radio there in the mumbles pointed it out, and he's absolutely right. If you go over to linuxacademy.com slash unplugged, just go there, linuxacademy.com slash unplugged. Look around. And then if you're looking for a gig, they are hiring two telecommute positions.

Oh, really? Yeah, you know, this is the great thing. Linux Academy isn't just a great learning resource. They're managed really well. They're part of the community that they service. They have been really, really smart about the way they've structured their company, and they've grown like crazy to make sure their content stays relevant. Even the old stuff.

Plus, I bet if you worked there, they got some good training there. I would just guess. LinuxAcademy.com slash unplugged. That's a double ding plug. LinuxAcademy.com slash unplugged. Okay, Mr. Colonel Linux, I'm going to open the floor to you. As far as folks that are going out there and grabbing hardware, Noah kind of has mastered the art of going into Best Buy and purchasing the Linux machine. But it doesn't always work out, and you've got to have thoughts on this topic.

So I open the floor to you. OK. So are you looking for just general advice or are you looking for past experience? Both. OK. Are you looking for just general advice? Are you looking for past experience? Both. Okay. So you're right. I can say with almost full authority that there isn't a computer manufacturer out there that I have not owned and installed Linux on at some point. There are some that tend to be better than others.

These days, you'd be surprised at how many companies are using the same ODM. And so really, you're buying a different brand, but it's the same thing on the inside. Maybe the chassis is a little bit different. Most computers these days are coming with Intel insides. And, you know, Skylake issues aside, if it says Intel on it, it probably will work right out of the box with Linux, wireless sound, video, all that good stuff. Yeah, you know, I got to disagree. I've bought two ThinkPads in a row over the last few years that came with wireless cards and a Dell that had to have their wireless cards replaced to an Intel. I've never had to replace the wireless card in a ThinkPad.

In fact, it would be difficult to do so because they actually whitelist their wireless cards. Wait, wait, wait a minute. Don't you remember? Didn't we have some issues with that ThinkPad? That's not a ThinkPad. That's an IdeaPad. Totally different, Chris. No, they are.

They are. Okay. All right. Well, so, okay. So the IdeaPad, yes. And the IdeaPad came with Broadcom. So that's going to be the one exception, right? The one thing that... Oh, Broadcom. Not the IdeaPad, but Broadcom in general is going to be the one exception, right? The one thing that...

Not the IdeaPad, but Broadcom in general is going to be the one exception you're going to find when you go into a store. But that's easy enough to do, right? From Windows, you can right-click, look on the device manager, and just kind of take a gander at what's inside that machine before you buy. And if it has a Broadcom chip, just stay away from it. The second thing I look at is if it has a Broadcom chip, maybe it's a super good price. And I'll tell you, actually, they have actually gotten in 16.04, Broadcom is actually much easier to deal with. In the installer, it recognizes you have a Broadcom chip and will just say, plug me into the internet and tell me that you want me to use this driver, and it does everything for you. And I just dealt with that with one of those h those little hps but the uh if it if it has a broadcom chip and i want it to be a super smooth experience

i'll flip the computer over and see if there are screws on the bottom if there's screws on the bottom the the high probability is you can take that bottom panel off and just swap that wi-fi card and as long as you're gentle you won't break any of the pins, then you won't have a problem. It's for people who like to mess with computers. Damn it, Leo. You're right. Those are good tips. So, Wimpy, I want to hear your thoughts in regards to the ThinkPads. Well,

with regards to ThinkPads, I've had to change Wi-Fi cards in ThinkPads, and in many cases, I've had to actually flash a modified BIOS in order for it to accept the Wi-Fi card I want to put in it rather than the select few cards that Lenovo have allowed to be used. And the same goes for hard disks as well. I've even had to do that for hard disks now that said i think i can see one two three four five six think pads from where i'm sitting at the moment and that's not all of them in the

house but i am a changed man and i'm going to tell you about this um when people buy computers that were designed to run windows or mac os 10 but predominantly windows and then try to install linux on it and have issues that's unfortunate and yes people have to get into this business of do they potentially replace the wi-fi chip and all the rest of it because these days it's generally the wi-fi chip a lot of the reasons why people say they have bought a particular laptop is because it's significantly cheaper than a laptop designed to run linux from the likes of entryware system 76 czar reason think penguin and the many others that are out there and i think because open source linux lin Linux distributions are free of cost, there is an assumption that it should be a cheaper laptop because it comes pre-installed with a Linux distribution. And the reason why it is not is because the likes of System76 and Entryware and the others don't have the volume of sales for economies of scale to come into effect.

So they cannot compete on price with the likes of Dell and Lenovo and HP and so on. So if you want to buy a computer and you intend to run Linux on it, save up a little bit longer. Buy your computer from one of those companies that are supporting linux as a first class citizen because the more people that vote with their money what will happen is they will be able to improve their margins and the costs of those computers from those linux supporting companies will come down as their sales volumes increase there's that and you legitimately will probably have a less challenging Linux experience.

You will not be having a conversation on an online user group saying, what Wi-Fi chipset card do I need to get to go in a Lenovo Yoga 3 or whatever it was. Now, Colonel Linux, though, you and I, I know you and I agree with Wimpy, like just totally like you. We I I don't know. I'm I may have been System76's first or second public customer. Like I am I am a diehard in that brand for years. Yeah. So it's over a decade. And but even yet, you and I still on occasion either build the PCs or we go out and grab a think pad. What are your thoughts on that? So it's never a first choice, right? Like you're you and I still on occasion either build the PCs or we go out and grab a ThinkPad. What are your thoughts on that? So it's never a first choice, right? Like you're right. I 100% agree with what everything Wimpy just said about supporting manufacturers that produce computers.

And if there was a System76 store in every single city – You just went away? You're just good? You're good? Oh, because you like to go into the store. Well, I think he found a System76 store. He's distracted now. We can forgive him. You should have seen the sales they had.

And what Wimpy is talking about, if and when people support them enough that they get their margins up, and we get to the point where they can offer a more cost-competitive product, that'll make this argument even less relevant. But at the moment, there are times when, for example, you have three days to switch somebody to Linux, and that person is there, and we need to have them a laptop, and we have to get it here installed and rolled out in 48 hours. You don't have a choice but to go into the store, and I think there is an advantage, an unspoken advantage rather,

of buying from a company that isn't totally invested in the operating system. When you buy a MacBook, Apple needs you to use macOS because that's where the longevity of their product is, right? It's really in the whole experience, not just the hardware. Whereas HP, really at the end of the day, HP, Dell, Lenovo, they don't really care about Windows. They care about Windows to the point that it makes them a profit on selling their hardware and past that, they don't really give a rip. If tomorrow everyone wanted to install Linux, you better believe Dell and HP and Lenovo and all those other companies would be making the most Linux-friendly computers possible

because they want their hardware to sell. And I think there is an advantage, however small. I'd say that's like 2% of the overall battle. The rest of the 98%, I completely agree with Wimpy on just trying to avoid them altogether if you can. So there is also an undeniable fact I completely agree with Wimpy on just trying to avoid them altogether if you can. So there is also an undeniable fact that there is a huge, huge price point at $300 that is extremely attractive to people looking at Chromebooks and stuff like that. Pacers and like – Yeah, yeah.

So Wimpy, you mentioned there's other options out there. What are your thoughts as far as that price point goes? And yeah, maybe you could call them having sort of certain sets of expectations or entitlements, but at the same time, wouldn't we love to welcome these Chromebook buyers or these $300 HP stream buyers with open arms in the Linux community? Your thoughts? Yeah. Well, so the other options, I've talked about the Linux specific manufacturers, other options i've talked about the linux specific uh manufacturers but if somebody wants a laptop from a mainstream vendor in two days to convert somebody to linux hewlett packard have a range of and i'm talking about in europe and north america now rather than the floods and floods of ubuntu laptops that are available in in the far east because they are awash with Ubuntu laptops out there.

It's about 40% of PC sales are pre-installs with Ubuntu. So we're not seeing all of that in Europe and the US. But Hewlett-Packard have a range of low-cost laptops, all pre-installed, running Ubuntu. They, in the UK uk are under 300 pounds i think some of them are just a touch under 200 pounds so they're entry-level chromebook pricing running ubuntu and then dell of course have a range of machines and they don't just have the high-end developer stuff you know the sputnik and the x 13. They also have a line of precision laptops now and their entry level costs are reasonable.

And again, if you buy those computers from those manufacturers pre-installed with Ubuntu or whatever distribution they may have, you are adding a count on the chalkboard to say there is a demand for people to buy computers pre-installed with Linux rather than buy one of their Windows computers and then try and install Linux on it after the fact. I completely, completely agree. I think that's also maybe worth part of the cost increase for those people who can make it work. I wanted to just, now that we've talked about the hardware entitlement aspect, which I think was a fascinating point, Wimpy, I wanted to now go back, if we could, just for a little bit longer to the issue that you brought up rather brilliantly is, and it just so perfectly underscores what is kind of an entitlement issue here. This 16.04 network manager problem is, it's big enough that we hear about it all the time here on the show. is it's big enough that we hear about it all the time here on the show. And so that to me is my barometer.

Been around for a few of these releases, and I can tell you what kind of things I get feedback on and what kind of things I don't per release. And this is one that's showing up on the radar. I can tell you from my metrics it's an issue. So with that said, like Wimpy brilliantly just said, nobody's doing nothing. It's sitting there, and in fact on the known bug tracker, it's gone off in the completely wrong direction.

And what could we do better, Wimpy, as far as an infrastructure to make this work? Rekai in the chat room is saying we need some sort of distributed API that interconnects all the bug tracking systems and allows them to share. Maybe something that uses the blockchain, and there could be coin rewards for bugs. What do you think, Wimpy? What are our options? Yeah, Rikai put in the chat earlier about creating a better bug tracking, bug logging system. And he's right, because even the good ones that exist in some of the open source projects aren't really that great by modern standards in terms of web applications.

And with the work that I do in Debian, filing and closing bugs in Debian, it's like pulling teeth. It's all done with email. It's just, I mean, I absolutely hate having to involve myself with the Debian bug tracking system. It's truly awful. I'm sure that everyone knows how to use it with email, that great but i so rarely use it not an encouraging way to get volunteers to uh spend their free time no and it gets more complex when you know you have a bug raised at say a distribution level say in ubuntu and then you know as you've pointed out that bug then actually is referenced in an upstream bug and you know conversations get lost between the two chains.

I don't know what the solution is, but I agree there's probably a better way out there for somebody cleverer than I am to invent and create and improve open-source bug tracking for everybody. It does seem like we should have a better system. Yeah, go ahead. I've actually brought this issue before, and I think the solution will be better laid in something like the Linux Foundation

to set the common bug tracker type thing. So that you have one account. It's also because the kernel is tracking. Because the usual problem is you are dealing with multiple entities, and the distributions deal differently. But what they have in common is the foundation. So I have bug tracker accounts on probably 15 or 20 different open source projects. And I'm going to be honest, and I probably almost will regret saying this on air, but Listen close, Phil.

It just gets to the point now when I go to a bug tracker where I don't have an account, I bail. I just, like, here's what happens, is like sometimes something crashes, and then like someone comes up and says, would you like to submit a bug? I say yes. You're like a little high of like, yes, I'm going to do it. I'm just going to submit the bug. I'm going to go for it.

Yeah, I wasn't planning on this, but you know what I mean? Say five minutes, it's fine, and then I'll move on. And then I go, and it's like, yes, I'm going to do it. I'm just going to submit the bug. I'm going to go for it. Yeah, I wasn't planning on this. It's like five minutes. It's fine. I'll move on. And then I go and it's like, oh, I don't have – yeah. In fact, as a matter of fact, what happens – you know what I hear in my head at that very moment? I feel like I hear the fail horn on Price is right in the back of my – yeah. And then I just move on. And I feel so bad for doing that. But I'm just – I'm tapped out. I think everyone's been there at times. I just move on. And I feel so bad for doing that. But I'm just tapped out.

I'm tapped out. Like when I go into LastPass, it's ridiculous how many accounts I have for bug trackers. And this is, I hate this crap. So. I mean that. I know. The IBM does have one tool. Report bug.

Yeah. I was going to say maybe this should adopt something like it. I don't know. Well, and, you know, like Rikai points out, this was why I created a GitHub account, like, a couple of years ago. Because, like, just more and more now is on there. And that's nice. But I don't really want to encourage open source projects to have their bug tracker tied to GitHub. Right.

So I just. I mean, it would be better on GitLab or something. Yeah, yeah, that's cool. More open. Yeah, that's fine. Distributed. Yeah, yeah. Even then. GitHub versus GitLab or something. Yeah, that's cool. Yeah, that's fine. Even then.

GitHub versus GitLab is a conversation for another day. Oh, absolutely. We could totally have that conversation. One last conversation. Since Mr. Brown had a chance to make it here, I didn't have anything in the feedback or follow-up because we didn't have much discussion. I thought I sort of threw out the bait out there for a ZFS versus ButterFS conversation. You really did.

More baiting than you usually do. Thank you. You noticed. You know, Wes, I like you. And I'll tell you what, not many bites, like at least so far. Now, I've only checked. That was the feedback as of about 9 a.m. So, Mr. Brown, I would love to hear your feedback if you had a chance to see what we

talked about in the new segment oh oh yeah i saw i i'm yeah didn't put anything on on your subreddit because you know not safe for children oh i love it i love it okay okay let me have it let's hear let's see what you think well i've got issues at multiple levels. So let's start with the simple one, just get it out the way, licensing. Until the ZFS licensing gray area is solved, no business that doesn't want to take a gamble can be using it. But can I stop you? So let me stop you there. I want to break this down. But at the same time, there must be a competitive pressure since now Debian, Debian of all distributions.

No, because Debian didn't. Okay. Debian didn't put it in their distribution any more than OpenSUSE has had it in our distribution for years. It's in a contrib repo. But they are the ones doing the groundwork around the quote-unquote workaround. That's what I'm referring to, not even necessarily them shipping the code. Then that moves me on to the next funky thing. So, okay, let's forget about the philosophical stuff and the licensing because, you know, that's another debate anyway.

Let's stick to nice, pragmatic, technical grounds. The Debian workaround is a DKMS, you know, compile-on-your-machine solution. Do you really want to be doing that for a file system? I hate it. No, I hate it. It's already been me once. But I will say anyone in production, you would pre-compile that module. Yeah.

You would have your own repo where that module is compiled and shipped. You would only update the kernel in conjunction with that module. So I don't know if that's an enterprise, like desktop, yes, but an enterprise, I don't know if that's a problem. But then that brings you to think, okay, who do you phone phone who's going to deal with those bugs when you have them i mean just look at the bug statistics on zfs on linux if you if you're nice and generous and only look at the same period as the current open upstream bugs in btrfs there is still more than twice as many bugs upstream in zfs for linux open right now than there is on BTOFS.

Okay, so one thing I want to just touch on really quick before we go too much further. When you mentioned the gray area around the licensing, I totally agree with you. It is a gray area. And I was shocked that Canonical had the Hotspa to to actually – is that Hutzpah? Is that what it is? Hutzpah? Had the Hutzpah to ship it. I tell you what, Richard.

I never thought – I thought it would be another year from now. And I feel like it's a strategy tax a bit perhaps on them down the road. But nonetheless, they did it. road, but nonetheless, they did it. So once they started doing this, it started to have certain flavors of the codec discussions. The same thing around licensing of codecs or licensing of the DVD playback. This feels like sort of a worry of yesteryear that never really panned out. There never really was any major repercussions for shipping codecs. There never really was a problem for using dvd css in the enthusiast in the enthusiast space which of course is is you know

a valid area fight yes it there wasn't and like even you know debbie and reviewed uh patent stuff because you know basically boiling down to we're not making any money so it's no problem for us that's fine but then your whole argument with zfs during las was the market demands well the market is money the market is businesses the market is enterprises those guys can't afford to do that distributions i mean take for example even open suzer we're not exactly the most enterprise friendly distribution we have organizations using us embedded stuff and on their devices because we have no gray area because if you're redistributing that stuff if you're taking but putting it oem on your hardware which is of course is a problem mint had the oem is suddenly then

responsible for that gray area so no one is going to touch that stuff with a barge pole i think you're right and this is the area where you and I probably disagree the strongest. So here's how I see what you're saying is you're essentially making the Internet of Things or small embedded devices or shippable devices argument, which is a pretty solid one. The issue is not only is there a wash of non-Linux OSes already in that space, but there's already a bunch of great Linuxes in that space that already have a great bunch of momentum behind them. At the same time, I would imagine it must be important to the project to gain ground on the desktop, in the DevOps space, in the engineer space, and on the cloud primarily. And when we talk about the market demands and we talk about the larger market as a whole, my main point that I was trying to make really there is ButterFS is screwed as far as a brand. Now, I'm not saying R-Linux is the measure of anything. But if you just look at the thread that was posted four hours ago on R-Linux about ButterFS versus ZFS on the desktop,

there are 19 comments. And literally, Richard, every single one of them is shitting on ButterFS versus ZFS on the desktop. There are 19 comments and literally Richard, every single one of them is shitting on ButterFS. And I'm not saying they're right. I'm saying the brand ship on ButterFS has sailed and nobody is going to hitch their wagon to that. I'm interrupting you for a change because this is

fun. File barriers. 2008. Every single distribution had file barriers disabled because of performance issues apart from susan because you know we care about this stuff that totally negated the the protections of journaling on ext3 ext4 and xfs every single one of those file systems was royally screwed with people losing data and everybody was screaming about it and it was a big issue and OWN did articles about it of how the world is falling apart

and then LVM made it worse because even when they fixed it on the file systems LVM didn't enable that so LVM was screwing up everyone's file systems and yet everybody loves the XT3 and 4 now and everybody likes LVM now because I would not make that up. Those were fundamentally And yet everybody loves the XT3 and 4 now. And everybody likes LVM now. I would not make that up.

Those were fundamentally flawed issues that every distribution apart from one was shipping. Now, BTIFS, yes, has some marketing issues. They are fundamentally flawed. BTIFS, when done properly, like you see in OpenSUSE, is a damn good file system. And if you want a root copy-on-write file system, it's your only choice on Linux. And if you want to boot to Snapshot, it's your only choice on Linux. You've got to try it. You've got to learn it. You've got to understand it. If people aren't ready yet, fine. But running to ZFS is like some savior. When the license is a complete mess, ZFS on Linux is a complete joke right now when you look at the bugs and who's actually fixing them.

You can't put your faith in that. ZFS is a great file system, but not on Linux, not yet. So let me, I want to address a couple of points. The first one is, I think in terms of enterprise storage deployments, I don't think you see a lot of extended for as the main core file system on like NASs and large storage. You'll see a lot of different file systems and some of them are extended for. But a lot of times there's like if they are, there's a very, very large and robust virtual management suite that sits on top of that. Or they're on like large proprietary appliances. Yeah.

Providing like a SAN. I think in the enterprise, I think that's sort of a gray area where ButterFS has basically no foothold. But I agree with your core fundamental point in that ButterFS is worth our time and worth our energy and we shouldn't be running to ZFS. Just using ButterFS on your route just to get snapshots alone, like that will work fine and you get a lot for just doing that. But I think if we zoom out a bit and we go to the meta level of what our conversations

are, what I'm saying is what you are presenting to me is a technical argument on the valid reasons why long-term ButterFS is a better investment of our collective resources. From a legal standpoint, from a technology standpoint, from an owning it standpoint, from an integration standpoint, and even from a feature set standpoint, ButterFS is a better file system for us to invest in. But at the same time, what I feel like you're not hearing is my argument that people are choosing Ubuntu or they're going to choose Debian over OpenSUSE because it simply doesn't have ZFS. And that's where the decision tree is at.

It's not – they're not going beyond that decision point. It's does it have ZFS support? Well, yeah, people – weird. But if ZFS – OK, the ZFS on Debian thing is a joke because ZFS is in OpenSUSE as much as ZFS is in Debian. So fine. There we go. If you want to have – I agree from like a software availability standpoint, but we're talking about they're the ones that did the legal groundwork to make it possible.

No, they didn't do any more legal groundwork than we have, which is basically you don't ship it as a core part of your distribution, which is simple, fine, you know, the user can then choose what the hell they do with their licenses, which is a concept that every distribution follows. And Debian have only just done it now, and yeah, okay, OpenSUSE should have done an announcement, what, two, three

years ago when we did it. I'm not saying that OpenSUSE doesn't have a repo where I can get ZFS and then install the userland utilities. That's not the defining thing here. What seems to be the defining... Debian put it in contrib. It's not in main. It's not in their distribution. It's not a core part of their distro.

It can't go in main. It can't go in main. Just like ZFS can't go in the main OpenSUSE repos. But that's the thing. If you're saying ZFS is in Debian, then ZFS has been in OpenSUSE for just as long. So you don't give any credence to their, what they claim to be is three years of legal work to analyze the licenses

and have a legal decision that they feel makes it safe. You're not giving any credibility to that work there? We did the same work. We came to the same conclusion. We kind of made similar steps. We don't like DKMS as much, so we build it differently. Does it seem like picking on Debian is almost irrelevant when the real cloud dominator is Ubuntu? And I guess what I'm trying to get to is, honestly, it seems like an anti-growth strategy not to have ZFS support. Honestly, it seems like an anti-growth strategy not to have ZFS support.

Okay. Data is sacred. You really think it would be a good thing if OpenSUSE… I'm not making a technical argument. No, no, no. But this is the thing. We have a responsibility as distribution maintainers to make informed decisions for the benefit of our users. ZFS right now on Linux has way too many gray areas. Ignore the licensing ones.

I'm just nervous about the fact that it's basically two guys, LLNL and LLNL, maintaining it with a massive bug backlog, some of which include data loss stuff. It's not in the kernel. It's not keeping up with kernel development that well. I wouldn't want to offer that. If the licensing issues totally went away tomorrow, if Oracle relicensed

it, I would still be nervous now saying, okay, this is a fully supported distribution which you can trust, fully supported file system on my distribution that you can trust. Because I can't trust it yet if this was zfs on freebst or solaris totally different story but this is zfs on linux and it's been less tried and tested and it is less trustworthy at this point and there's way too many scary bugs i see in that bug tracker and i don't see it being

fixed fast enough because it's only a couple of guys because you have the licensing issues, but then we're back in this cyclic organization. So just to be clear on my position there, the only place where I have production data that I deem valuable on ZFS is on a free NAS server. And I think you would be very risky to run ZFS in a large-scale production unless you're doing, like Wes mentioned mentioned pre-building your module and you have safe update process.

Will this all change? I mean, should we really come back to this in a year or two years or the next LTS now that we have that? I don't think so. I think the licensing is always going to make it this weird DKMS workaround. Except on 16.04 where you have it pre-compiled. I guess will that

change this discussion? And will that make ZOL better tested? I'm not saying it has now. I'm just curious. I mean, if this was a different company doing this move, I'd have a totally different opinion. But, you know, file systems are traditionally kernel-y kind of stuff. I mean, they live in the kernel. They play with the kernel. They do all of that.

This is canonical we're talking about. They're not exactly the biggest contributors to the kernel by any measure. So, yes, they do all of that. This is canonical we're talking about. They're not exactly the biggest contributors to the kernel by any measure. So, yes, they've taken some package, they've packaged it up, they've put it in their distribution. Great for them. But if something goes wrong, who are you going to call? Who's going to fix those bugs? Where's

the skills? Where's the knowledge? Where's the trust? So, I will claim that canonical is, as a company that is selling a product as well and trying to put this product out there, they can take responsibility if it comes down to it. They've done it in the past. They've also decided not to do it in certain circumstances. But as it's their choice and they understand that they have to support their decisions,

they will probably take the responsibility. This is the first thing. So if we separate out the legal issues and the technical issues, which I think the legal issues we just have like West is saying we have to stay tuned and see where it goes. But with the technical issues, Richard, it reminds me fundamentally of people who are very intimate with container technology say a lot of these same things about Docker. But yeah, but Docker is the big brand that everybody wants. I mean, you saw Red Hat. They slammed it into 7.0, so that way they could claim Red Hat Enterprise 7.0 had Docker support,

even though it was nowhere near enterprise-grade ready. We have it in sleep, yes. I mean, yes, there's decisions like that you can make, but there has to be a line where you say, okay, we're not going to cross this. We can't do this in a stable and sane way. And I'd say right now, if it was just down to me,

just looking at ZFS on a purely technical level, ZFS and Linux has not proven itself yet. And this is file system data we're talking about. Yeah, I agree. Something in some container which is going to explode anyway because it's a container. I mean, it's meant to die at some point. Right. It's a totally different scenario than this is my data which I want to keep forever.

But seriously, if we want to concern about legal issues, we have to be paying more attention at Google versus Oracle as it goes to APIs. And guess what? Linux copied Linux API. Well, yeah, we'll see. We've been talking about that on the Coder Radio program if you guys do want some discussion on that. So I guess, Richard, I feel like you are right from a playing it safe standpoint in regards to legal risk,

and you're right from a technical standpoint. And I don't discourage OpenSUSE from taking this stance because, if anything, it just adds more pressure to make ButterFS continue to get better and better and better. So from a pure Linux user standpoint, I sit here and I go, sweet, I'm really glad that there is a well-established distribution that is taking this position that is going to forward our locally grown file system. Like, that person inside me? That is value to our ecosystem. I'm like, yeah, tip of the hat to you.

But then, like, the market analyst side of me that looks at the market positions that OpenSUSE is hopefully going to grow into, and I think, oh, geez, this could be a barrier. So I'm really torn on the issue. don't think you are no i'm i'm not i'm i'm seeing too many like when it comes to like okay evolving storage trends like the the tech that everybody's got to be talking about isn't zfs it's saf and and that's a completely different kettle of fish with a completely different use case but that's you know that that's where the low mid-range storage market's moving that's where the money's going that's where you know trust me suza's seeing that you know our storage product was funded

before we even started building it um you know it that's yeah that's where the disruption is really going to be not in not in zfs i mean especially even like if you look at like the cool things in zfs like raid z which isn't raid and everyone goes on about that oh yes it's finally solved the raid hole that everybody worries about and everything yeah then you can never read your data quickly because it only reads at the speed of one disc i mean zfs has lots of cool things in. In theory, it's like tech porn. But in the real world, the world's moving on. There's more interesting things on the horizon.

I'd much rather look at that than... You may be right. You may be wrong, but you may be very right. And I hope you are right about that. And thank you for having such a strong viewpoint for this. Yeah, absolutely. Wimpy, I wanted to give you sort of the follow-up word before we wrap up this segment. I've got nothing new to say to this.

Richard has summarized it all very well, but this hasn't been a recent development within Debian. It was something they've been working on for years, and they actually announced that ZFS was coming to Debian just over a year ago when the last Debian project leader, Neil McGovern, was in. And so that was last April that they announced this was coming. And then a year later, it's landed. So it's not been... And also, they've had discussions with all of the legal experts that they can talk to, and they're confident that they are right to do this.

Would that sort of be the green light, Richard, is if some legal trials happen and this thing actually gets tested and things sort of look like they're shaping up in favor of shipping, you guys would probably reevaluate the position if there was a need to, right? If Oracle relicens ZFS tomorrow, everything I just said changes. Yeah, I suppose so because, I mean,

if they really, which I doubt it, but that is also the position of the Free Software Foundation too and the Software Conservancy. So maybe there is,

so there are some big foundations out there that are at least calling for it. I don't know if Oracle is very motivated to do it, though. I've checked the weather

forecast in hell for tomorrow. It's still looking toasty. Yes, exactly. But, Richard, I really do appreciate you stopping by to chat about it. And, of course, we can always pick it back up in the post show. You know, the Unplugged program is live on the Tuesdays. You've got to be here live.

Over at jblive.tv, jupiterbroadcasting.com, slash calendar to get it converted to your local time zone. You can also join us in that virtual lug and hang out with your local virtual Linux users group. We'll see if Wes's laptop makes it another week without getting reloaded. Probably not. Probably not. But be here next week. Linux Action Show at reddit.com for feedback or just leave a comment wherever you watch this. We really do appreciate that.

And don't forget, you can catch us next week or grab our RSS feed. See you then. Get it out of here. Zombies don't run Linux. I've never installed GNU slash Linux. Richard, you're a good sport. Heck yeah. Richard, you're a good sport. Thank you.

And nice to have somebody who's willing to come in here and voice a well-stated and well-informed opinion, too. So I really do appreciate it. Yeah, excellent. All right, jbtyles.com. Bang for the best. Now we vote! Healthy-looking mumble room today. I love that.

By the way, while I'm on a roll you do know that SUSE Linux Enterprise only uses the LTS kernels now right that's nice good point to make good point to make well said well said Greg's kernel of truth isn't bad the price of freedom

the file system a bug report pain alright you guys go over there and go shake them out I do have a little post show thing I wanted to share with you guys oh drop it down

you know what look at this thread over there speaking of cheap laptops and stuff there is this huge thread submitted 10 hours

ago by read harder blog slash rant oh let's get in yeah about taking a serious look at Chromebooks as the Linux desktop and talking about

loving Chromebooks etc etc but then of course after 17 comments it devolves into using Crouton or other things to get Linux on Chromebooks. This is a hot, hot area. That's a hot market. It really, I mean, they're so cheap. So cheap.

Hot market. Colonel Linux, you've had some good success recently with picking up a cheap laptop and throwing Linux on there. I have, but basically what I found is that Microsoft is subsidizing hardware just like Google is, and the nice thing about that is they don't lock the – Chromebooks do this thing, which a lot of people are aware of, but some people don't know exactly how much of a pain it can be. They lock the bootloader so that there is a signed key for Google,

and if the battery dies on certain models of Chromebooks, what happens is it defaults back to that read-only state with the secure boot enabled. And you can't actually boot back into Linux because it doesn't have the Chrome key. And because Chrome OS is no longer installed, you can't boot into Chrome OS to turn secure boot back off. And so you're just kind of in this limbo that forces you to reinstall Chrome OS and then reinstall your Linux distro.

And so you can fix it on certain models by taking a right screw out, but what I found is that basically, these ODMs have made cheap little computers, like HP makes a cheap little computer and they sell some of them to, Google subsidizes some of them, puts their little goofy right protection in there and sells it with Chrome OS. And Microsoft does the same thing with Windows 10, except when you wipe Windows 10 off, then you're just left with a normal boot.

So this is why I'm going with Wimpy. My position on this, if ever possible, is skip a computer, save up, and get one that runs Linux, that's built to run Linux. Because, I mean, that stuff doesn't sound too bad, but just having to worry about that would drive me crazy! So, let's be clear, though. I mean, there are certain times, though, where you're not skipping one computer, you're skipping two or three, right?

If you're looking at those HP streams, they're $179. Now, what's the cheapest computer I can buy with Linux on it? Yeah, and you know what they were saying in that thread was my computer died, I needed a quick replacement. I hadn't budgeted for a new computer. I totally understand that, too. Well, how much computer can you buy for $179 on eBay?

And that's going to have a decent keyboard. It may be a little bit older, but hey, you can take the backplate off and replace the Wi-Fi card or add more RAM or change the hard drive. Go used. You could never do with a Chromebook. In a pinch, I feel like we didn't mention the Wi-Fi USB 3.0 dongles you can get. Yeah, that's what I do.

You can just buy a better Linux, everything else. I just go USB wireless. Kernel Linux even gave me this tiny little dongle that's the size of a mouse receiver, just the little tiny ones. Nice. Kernel Linux and its tiny dongles. Woo-hoo!