LINUX Unplugged - Episode 56: One Packager for All | LUP 56

Episode Date: September 3, 2014

The systemd group has a proposal for universal software management scheme for all Linux distributions. We’ll share the technical details, debate the philosophical impact & explain why it’s all powered by btrfs.Plus some thoughts on the ultimate desktop manager, the true cost of a MacBook, and much more!

Hey, you know, I came across something kind of interesting today. Do any of you guys in the Mumble Room ever struggle with, like, anxiety problems or things like that? Anybody ever have battle with that kind of stuff? I know, like, it's kind of not, like, a normally socially acceptable thing to talk about, but I think a lot of geeks, and probably the geekier you are sometimes, the more likely you could be to have some sort of anxiety problem. Maybe there's social anxiety. Maybe it's just, you know, stress anxiety. I found a really interesting article today.

They kind of put it all in perspective that, you know, it felt like it connected me to one of the OG geeks. And that is Darwin. There's a post over on brainpickings.org that was posted on the 28th of last month. And it goes into detail on Darwin's struggle with anxiety. And there's a couple of interesting things in the article that I found. So here is a quote from Darwin that he wrote in 1865. He says, on two occasions prolonged for months. Vomiting preceded by shivering, hysterical crying, a dying cessation or half faint and copious pallid urine. I'm not even sure what that is.

He says, now I have vomiting and every passage of flatulence preceded by a ringing of my ears. He says, I have nervousness when E leaves me. And E refers to his wife. He would be stricken with horrible anxiety. He goes on to say that this was in a chronicle of Darwin's problems. It says, Darwin was frustrated that dozens of physicians, beginning with his own father, had failed to cure him.

By the time he wrote to Dr. Chapman, Darwin had spent most of the past three decades, during which time he struggled to write on the origin of species because of his anxiety. He was bound to his household based on his diaries and letters. It's fair to say he spent a full third of his daytime hours since the age of 28

either vomiting or lying in bed because of crippling anxiety. You know, you hear a lot about, you know, like you hear a lot about like these historical figures, but you don't ever hear about this part, do you? You never hear too much about the background. And their own personal struggles like this. Oh yeah. A lot of people have struggles with anxiety. I, for one, am somebody who suffers from anxiety disorder.

Not otherwise specified. I've actually been diagnosed. And, yeah, I can totally relate to that. Do I believe in Darwin's philosophies? No. But that's just because of all sorts of other factors. But in terms of struggling with anxiety, I totally get that. Yeah.

And I think it's something that a lot of intelligent people are probably afflicted by a lot, too, because your mind maybe wanders more than it should or faster than you can control it. Yes. In some cases. I have described that over and over again to friends and family is that, like, at night, my mind will just take all sorts of different paths, which is why I have to read before I go to bed. Otherwise, I don't sleep. You ought to do what I do. Yeah. Just call up Aaron Saigo and then hang up after he answers the phone.

That's what makes me feel better. Huh. He's in here with us. Hey there, Aaron. Hey, I always wondered if that was you. Yeah. I thought maybe you'd recognize the breathing, but it's fine. I hoped. I'm glad you could join us today.

Congrats on the new project, by the way. That's very cool. Very cool. Oh, thank you. Yeah, I suppose you're Congrats on the new project, by the way. That's very cool. Very cool. Oh, thank you. Yeah, I suppose you're referring to the Luminosity reboot? Yes. Welcome to Linux Unplugged, your weekly Linux talk show that always recommends building your own online photo backup. My name is Chris.

And my name is Matt. Hey there, Matt. Happy post-Labor Day to you. Are you all rested up and ready to go? We got a big show today. Absolutely. Joining us in just a little bit will be Aaron Saigo. He's launching a new project that I'm pretty excited about. And he's also been talking a lot about convergence. What is the desktop? What are laptops? Where does mobile fit into all of this? Sort of really doing a good meta-analysis of a lot of things we've been talking about lately.

So we'll chat with him in just a bit. I'm excited about that. Plus, I don't know if you caught it. We mentioned earlier, but Linus was at DebConf, the Debian conference. And I guess they had it in Portland. So Linus showed up. You know, he just threw on his socks. Then he threw on his sandals, in that order, like a man.

Went down there and talked to the DebConf folks about what was going on. And I love it. I haven't even seen the whole video yet, but the first five minutes of the video are like I gotta play this in the show. Because he goes into a rant and he perfectly summarizes kind of what's wrong with distributions. And it turns out kind of around the same time, the SystemD cabal, they call that themselves, I'm not making that up, the system decabal, they call that themselves. I'm not making that up. The system decabal thinks they have a solution to some of Linus's complaints. And also, if we have a little time, we'll see.

I'm not quite sure, but if we have a little time, we'll also discuss Fedora's future package manager, what it can do and what it can't do quite yet. But what you should be seeing starting in Fedora 21 and why DNF might be a big deal. But first, Matt, as is tradition on the Linux Unplugged show, it is time for emails. Dave writes in on a pre-show discussion we had a week or so ago

about systemd. He says, hey guys, I've been listening to Linux podcasts for a few months. I enjoy them very much. Your discussion of systemd seemed to come down to innovation versus stability.

I've seen this tension throughout my 30-year career in the Unix arena and in the past 20 developing a couple of big commercial Unix operating systems. Customers love stability, and many will play stability over innovation to a large degree. Innovation in these operating systems has to come from something completely new that doesn't disturb the old stable features. in these operating systems has to come from something completely new that doesn't disturb the old stable features. Or, at defined points, where

an OS vendor can warn the customers there could be breakage ahead. What we're seeing play out in the Linux arena is a maturing of the operating system and a growing group of users valuing stability over innovation, particularly when innovation is disruptive. Fortunately, with Linux, the different distributions

are free to introduce these innovations in ways that make sense for them. E.g., I wouldn't expect Red Hat to put SystemD into Red Hat Enterprise Linux for a while, but other distributions might choose to jump right in. And I don't know if I totally, completely agree that it comes down to the camp that likes SystemD is for innovation and the camp that doesn't like SystemD is for stability, because um, I can't speak to that. Yeah. It to me implies like, um, the people that don't want system D are curmudgeons and they're like, they're afraid to change. I wouldn't go that far, but I think it's assuming a lot in both directions and I wouldn't necessarily go there cause I don't know what their opinions are. I agree. I think it is going too far. Uh, and I, I think too far. And I think that sometimes it's not so much about avoiding change as it is just a philosophy

disagreement. And that's what makes it – That I would agree with. It's a multi-spectrum thing, I suppose. It's some of people that are curmudgeons about change. It's some people who don't like the design. It's some people who don't like the people behind the design. It's some people who question the political reasons.

It's all of these things that make it into one big issue. Then that's why it won't die. If it was just one or two things, people could rationally debate and argue it and shut that down. It's not just one thing, though. Jeremy writes in. He says, my thought on Linux desktop managers. Get ready for this, Matt.

Are you ready? He says, hello, all. Thanks for the Jupyter Broadcasting netcasts. And he says, not just netcasts, but resources as well. Now, I'm thinking, I'm wondering if... It's for people who like to mess with computers. I'm wondering if he's a Leo fan. He says, but he likes the netcast.

You caught that with netcast, huh? He says, I recently wrote a little blurb slash rant on my thoughts of Linux desktop managers. And he links us to it. I'll read a highlight from a second. But the TLDR is, I avoid all desktop managers and I run my own lean environment, which makes my life easier. And here's what he does, Matt. Get ready for this. So he wrote on his blog, he says, I've always felt that at some point someone out there would develop a manager that would not try to be an all-encompassing ecosystem,

but rather maybe just something that attempted to just simply be a powerful desktop manager. That time, though, never came. It seems that every iteration of every major open source desktop environment, the wheel is being reinvented again from scratch, and the idea of how things should look, feel, and behave kept changing unnecessarily. They rarely provided any improvements that I felt were positive. So, if I'm not using a desktop manager, what do I use? Well, first of all, I remove the default window manager and replace it with a highly customizable

window manager called Sawfish. Sawfish is very customizable, with themes, keyboard shortcuts, customizable menus, and scripts. It focuses on being a great window manager and doesn't attempt to be anything else. It's easily exceeded my expectations of what a window manager could be, so

I switched to Sawfish about 10 years ago and never looked back. Anybody in the mumble room using Sawfish? Never tried it over here. Nobody in the mumble room is using it either, Matt. No, no. Been around for a while. Sounds cool. I mean, it sounds compelling, but I just

I don't know. I'd have to. Makes me want to try it out though. Yeah, I'd be curious to see what you could do with it. And I like that he's stuck with it for 10 years. Yeah, that really says something. Quick question, wasn't Sawfish the window manager in older GNOME releases? Yeah, I was wondering about that too actually.

I wasn't sure. I didn't want to say anything. You know, I'm googling it right now. Sawfish. I'm going to search for it right now. Window. Yeah, look at this. I think it was the 1.x times. Yeah. Yeah, that sounds right. Here's a screenshot of it actually on Wikipedia.

Boy, that's pretty hardcore. Wow. Yeah, well, I mean you could... I think it was... Wasn't it extensible with Lisp or something? Yeah, well, I mean, you could... I think it was more power to him. Wasn't it extensible with Lisp or something? Yeah, and he mentions, I think that was the scripting aspect to it, and he also mentions that all the key bindings are totally customizable too,

which is kind of cool. I like that too. That's really funny, though, looking at that. Huh. Yeah, I don't know. I mean, a little conf file massaging, that's fine, but I don't really want to get into scripting something, so I don't know. Yeah, I'd have to kind of weigh that out. Yeah, I don't know. I mean, you know, a little comp file massaging, that's fine, but I don't really want to get into scripting something. So I don't know. Yeah, I'd have to kind of weigh that out.

I actually have a droplet up there that I've installed X on. And you're going to laugh at me, audience, but you've got to understand the way I work. I even installed Chrome on my DigitalOcean droplet. I SSH in. I turn on compression and X11 forwarding. And then I can manage all of my server-side stuff just using local host. And I don't have to expose any of that stuff to the web. And that's just another layer of security for me.

So for me, because I'm always on an X11-based system, well, for now at least, I can SSH in. You guys know about SSH and X11-40. You use compression, it runs like a champ, and you can start Chrome up. And to be honest, I've actually had more success with Firefox. I have both of them installed. And Opera, too. I have Opera installed. Hey, man, I can't help it. You know, Firefox runs really well under X11-40. Yeah, it does. It does work quite well, as a matter of fact. So there's a lot of options you could do with a DigitalOcean drop. But go over to DigitalOcean.com and arm yourself with the knowledge of Unplugged September. That's the promo code you're going to want to keep in mind as I tell you about DigitalOcean because it's going to get you a $10 credit.

So what is DigitalOcean? It's a simple cloud hosting provider dedicated to offering the most intuitive and easy way to spin up a cloud server. It's a simple cloud hosting provider dedicated to offering the most intuitive and easy way to spin up a cloud server. And users can create a cloud server in under 55 seconds or 25 seconds if you're our audience. And the pricing plans are incredible. They start only $5 a month for 512 megabytes of RAM. That gets you also a 20 gigabyte hard drive, one blazing fast CPU, and a terabyte of transfer. DigitalOcean also has data center locations in New York, San Francisco, Singapore, Amsterdam, and London.

Their interface is simple. And their control panel is crazy intuitive. But on top of all of that, they have an amazing, straightforward, simple API that you could just write your own scripts against or take advantage of any of the community applications that are out there. There's plenty of them that you can get your hands on. They're really nice, too, because once you get a system spun up, you still have the option to do one-click deployment of applications. You're like, oh, that's interesting. Yeah, they're going to use Docker to do some of that stuff. How cool is that? They're using KVM for the virtualization, and they're sitting on top of those SSD drives, and they work their butts off to make sure that they're at some of the best data centers in the world. They just spun up a new one with IPv6 in New York. They got that new one in London.

They're growing like crazy. Tons of you out there in the community have been trying out Digital Ocean because it's worked so many for you. That's why they invested in those SSD drives, to get you that performance. That's why they use the amazing hardware, all wrapped up in this incredible interface. They recognize that it takes all of that together to make an amazing product. Go over to DigitalOcean.com and use the promo code

UNPLUGSEPTEMBER. You get a $10 credit. You can try out the $5 rig that I've been getting a heck of a lot of use out of for two months for absolutely free. DigitalOcean.com and the promo code Unplugged September. And a really big thank you to DigitalOcean for sponsoring the Linux

Unplugged show. Love my D.O. rublets. I got multiple now. It's like phones up in here with me, Matt. It's like phones. Yeah, I think you've got me beat. I had like four of them at the peak. I dropped back to three. I'm actually looking at doing a fourth one again. Oh, what are you running on there? Are you doing WordPress?

I'm actually doing more WordPress than anything. I'm still rocking my BitTorrent sync. It's a big thing for me. I do a lot of data back and forth. Good for you. But WordPress, it's the best WordPress experience I've ever had. And I've had a lot of them. Same here. It's really, really smooth.

Super fast. Love the setup. Plus, you control it from top to bottom. My wife, Angela, has been talking about how she kind of wants to get back into blogging more regularly. And one of the things we're thinking about doing is we're looking at all the different options, but definitely one of them on the list is just set up a WordPress site on a DigitalOcean droplet. And she can manage it even. It's really straightforward. It's kind of awesome. I would recommend it. It's even. It's really straightforward. It's kind of awesome.

I would recommend it. It's fast. It's just DIY, but it's also easy. It's really great. Yeah, absolutely. Okay, Matt, so we've got a couple more emails we're going to get into. This next one came from Ryan, and this is a topic I love. First of all, Ryan wants to know, just to clarify off the top, all he has to do to be a guest on the show is take a shower with me,

because apparently when Eric was here last week, we implied that Eric took a shower with me, and so therefore he was allowed on the top. All he has to do to be a guest on the show is take a shower with me, because apparently when Eric was here last week, we implied that Eric took a shower with me, and so therefore he was allowed on the show. I don't know. So, Ryan, just to clarify, that only works under two circumstances. Number one, you're a lady. Number two,

if you're not a lady, I'm really lonely that day. And then if you get me on the right day, you take a shower with me, and I do have a shower here in the studio, I'll probably let you on air. I mean, you know, I'm a nice guy after all. I'm dumbfounded that that actually made it. So that happens. Hey, man, just telling it like it is. And Ryan wanted to know.

I'm just here to help the people with the information they require. So he says he's been a listener for a while now. And he wanted to get to his actual question, not just the shower question. He says, I'm relatively new to Linux. I tried to switch my five-year-old desktop last September, but I had to switch it back to Windows for both game-related and school-related reasons. I recently bought a laptop with explicit intent of installing Ubuntu 14.04. It took about eight hours because it came pre-installed with Windows 8. Don't get me started on that OS or UEFI.

But I got it loaded and it works beautifully. With the next six to eight months, I'm looking to get a refurbished desktop with a faster processor, more RAM slots, and the MMOs I want for it aren't friendly with Linux. So unfortunately, Windows 7 for that one. Once I get the second desktop, I'll be looking to repurpose my current desktop as a local file server for cohabitating my girlfriend and my files with a little backup action. She has a Windows-based PC, so we have a mixed OS environment. When I

repurpose it, I plan on using Linux for this new machine. What would be your top choices for distros to use for a home file server? I'm going to set this up in VirtualBox on my laptop in the meantime, so I can play around with it before I actually go into it and put it in production. He says, thanks for everything, and I hope we had a great Labor Day.

So I have a few suggestions. Matt, I'm sure you have a few. Anybody in the Mumba room have a home file server distro recommendation for our emailer, Ryan, here? I use OpenSUSE. Okay. It's not a Linux distro, but I can't speak of FreeMAS high enough. I run an OpenMedia phone.

Yeah, something's up, guys. No, go ahead. I was just seeing what else I was giving you guys a chance to say. So we've got OpenEvolve, FreeMAS. Oh, sorry. What else? What else do we use? Well, I just use SFTP into an OpenSUSE server.

OpenSUSE. And that's how I get my stuff done. Matt, do you think, I mean, come on. Nobody's going to say Ubuntu LTS, right? Come on, right? But Matt, that makes a good one. I think Ubuntu is good. I think anything Debian-based is certainly worthwhile.

There you go. That's kind of my go-to thing. I think really moving more toward a Debian situation is just a comfortable place to be. I like that. I think even if it's not Ubuntu, I think Debian's good. I also think OpenSUSE, like the Evergreen release, would be quite good. Wouldn't recommend Fedora. And if you're in the Red Hat camp, check out CentOS.

That's also a pretty solid option to go. in the Red Hat camp, check out CentOS. That's also a pretty solid option to go. But like Wimpy says, OpenMediaVault, also a great opportunity to look into. And one we're going to do a review on soon, here on the show. Maybe we should do that next week. Those of you doing the OpenMediaVault, would you be ready

to do the review next week? Yep. So we'll do on Linux Unplugged next week, unless something changes, we'll have our review of OpenMediaVault, Ryan, so that might help you as well. Okay, so we'll do on Linux Unplugged next week, unless something changes, we'll have a review of Open Media Vault, Ryan, so that might help you as well. Okay, last but not least, Sean writes in with the real cost of a MacBook. He says, I know I'm a little behind on this comment, but I just started looking at Linux within the past year,

and I came across Linux Action Show a few months ago, and he's been catching up. I have a MacBook from the spring of 2008. It runs better than Windows was my thought when I bought it. Now my thought is I can only run up to 10.7 and that's as far as I can go. In fact, he's stuck at 10.68 right now because as some of you may or may not know, after a little while, Apple kind of discontinues support for these older MacBooks and they don't really allow you to install the newer OSs on them anymore after a point. He says, I tried running Linux distros on a VM in the MacBook, but it's slow and sketchy. So I know I'm stuck with it for a while,

but I got a hold of a couple of old desktops, and I've been running Ubuntu 14.4, and I want to set one up as a home server as well. So I've downloaded CentOS. Well, there you go. All that to say the inability to move forward without shelling out major cash for new hardware has made me look back towards Linux.

Also getting back into computers, and I'm in the process of getting an A+. And looking at a Linux 101 course. So he says, keep up the great work. Boy, isn't that true, though, about the MacBooks? One of the things is you could throw Linux on there and probably get a couple more years out of that, right? I think that's really it,

because I think that really when you're in their own ecosystem, they set the rules, they set the pace, and you're limited to whatever they want to do with you. I, for example, have an old PowerPC Mac Mini that runs Linux. And the reason why is because that's all you're going to do with it short of using it as a doorstop. It doesn't have any benefit in the OS X universe. Very good. Very good. All right.

10 universe. Very good. Very good. All right. Well, we're having some connection problems with Aaron, so we'll see if he's able to make it back on. Oh, okay. I hope he can. If not, we'll still give him a good plug for his project he's working on because it's pretty

exciting. So what we'll do in the meantime, as the guys work that out on the server end, is I'll tell you a little bit about our next sponsor, and that's Linux Academy. And Linux Academy is a great sponsor for the Linux Unplugged show because it's an opportunity for anyone in the audience that's learned on their own, or if you're like me, self-taught on the job, you'd be surprised going back

and taking some of these fundamentals, how much little bits of details they'll fill in. And in my case, at least, I won't speak for you. In my case, I learned, oh, I could have been saving myself quite a bit of time if I had learned this the right way. You ever had that? Now you're like, oh, I guess I could have done this a lot easier. I've been doing this the hard way the whole time. I have that at least once a week.

So yeah. So here's just a little friendly reminder from someone who's been in this position. Maybe it's time to just check up on things. Go over to linuxacademy.com slash unplugged. That's going to get you the Summer of Learning discount. You should go jump on that right now because it's a 33% off offer. And Linux Academy is offering new content every single week. So you're going to want that sub. So what is Linux Academy? Let me tell you. It's kind of an awesome service created by Linux enthusiasts and educators who really wanted to put something together that Linux users could go to and know they're getting really good content from people who love Linux. They have step-by-step video courses, downloadable comprehensive study guides, and each course, when you need it, comes with its own

server, spins it right up. You can pick from seven plus Linux distributions and they'll automatically adjust the courseware to match the distribution you've picked. So if you're a Red Hat guy and you need to get in on a little bit of Debian, you can take care of that or vice versa. It's really kind of neat. And I've heard from a lot of you out there who've tried this out and say, you know, you mentioned the downloadable comprehensive study guides. You didn't mention that was also video and audio

that happens to work really well as a podcast. Or you forgot to sometimes mention they do live events that you can ask the educator directly when you've been taking courses. I love all of that, but me, I'm a recluse. I sit up in my office late at night when I'm not sleeping very well or when I want to accomplish a me, I'm a recluse. I sit up in my office late at night when I'm not sleeping very well or when I want to accomplish a little something, get a little intellectual

stimulation. I bring up my Linux Academy dashboard. I look at how long a course is going to take me, and then I just go to town on it. And sometimes that's a scenario where I'm building something from beginning to end on S3 using S3, EC2, all of the Amazon web services. Or maybe it's something on a Linux box. It's maybe something using OpenStack because it's not something I've had a lot of in-production experience with, but I still want to talk about it from an educated standpoint. That's what I love about Linux Academy, is they're constantly adding this new stuff that I can continue to test myself, see where I'm at. They give me progress reports, and I can also see if there's a new kind of technology that's coming up that interests me, that sort of excites me, because sometimes you

run the risk when you work on a particular area of getting a little stale. This is my opportunity to prevent that. Go over to linuxacademy.com. Check out their OpenStack Essentials. It's now 100% complete with tons of content. Their AWS-certified SysOps Administrator prep course has over 14 hours of content that you can check out.

I think it's a pretty neat system. linuxacademy.com slash unplugged to get that Summer of Learning discount. Go over there right now. Let them know you appreciate them supporting the Linux Unplugged show. That's linuxacademy.com slash unplugged. And a really big thank you to Linux Academy for sponsoring the Linux Unplugged show. Check them out, guys. They're great.

I don't see Aaron back just yet. So while we work on that, I got another topic. We can move to our next topic because we've actually got way more show than we can fit anyway. So either way, we're totally fine. And the meta story we've been covering is, is it too late for the Linux desktop? We covered John C. Dvorak's trolling in Linux Action Show, right? And then the week before, actually for like the last couple of weeks on Linux Unplugged, we've been talking about sort of the perceived threat that the Mac desktop represents to the success for Linux on the desktop. Well, this past week, Linus Torvalds, your buddy and mine, was over at he actually wasn't just over there.

He lives in Portland, so it's not that far for him to go. But he was at the DebComp 14 in Portland. far for him to go. But he was at the DebComp 14 in Portland. And Linus's style these days is to do more of like a Q&A session than it is so much to go up there and do like a big speech. And so he sat and just kind of right off the top of his talk, started taking questions from the audience. It's kind of cool when he does that too, because, you know, anything goes, right? And it's not like it's this pre-manicured experience where he's going to hit on these bullet points and go out there and give the Linux talking point speech and really pitch Linux to all of the people. He's just a guy going out there answering questions, and I love that about him. So the first question is the one we're going to start with today, and then we'll move into what the system decabal is trying to do to address this.

And I think it should be for an interesting discussion so i'll start with linus's a question here that it should be i think he's wrapping up a statement and then the question begins and we'll let him go for a bit it's a great talk i have a really hard time just doing something else because that means i have to switch over all my kids machines my wife's machine and and and that's just painful so that's just painful. So that's why. Okay, so this is a little difficult. I guess we're all, at least for me, I'm kind of wondering very much like I think you are lately, how to get to the year of the Linux desktop, if you will, right?

Wow. And we'd all like it to be the year of the Debian desktop, if possible. And I'm trying to figure out if anything, if you have any insight as to possibly how to get closer or towards that. So, I mean, there's a lot of things that are getting closer to that. And I think I mean, I people technical people don't tend to use Chromebooks. But I think that Chromebooks are the kind of thing that will make the year of the desktop more possible, right? Because once people get used to running their applications basically as a browser, that makes a lot of things much easier. That said, let me go on my first rant of the evening, right? One of the problems desktop has, I mean, ignoring all the purely market and getting free installs

and making it just so that normal people, and by normal people I mean obviously non-technical people, will just buy a machine and it just works. One of the things that none of the distributions have ever gotten right is application packaging, right? And now somebody will say, hey, dpackage is way improved and much better than RPM. And that's not at all what I'm talking about. I'm talking about actual application writers that want to make a package of their application for Linux. And I've seen this firsthand with the other project I've been involved with, which is my Dialog application, right?

We make binaries for Windows and OSX. We basically don't make binaries for Linux. Ouch, that's Linus. Ouch. Why? Because making binaries for Linux desktop applications is a major fucking pain in the ass, right? You don't make binaries for Linux. You make binaries for Fedora 19, Fedora 20.

Maybe there's even like RHEL 5 from 10 years ago. You make binaries for Debian, well actually you don't make binaries for Debian stable because Debian stable has libraries that are so old that anything that was built in the last century doesn't work, right? But you might make binaries available for Debian, whatever the code name is, for unstable. And even that is a major pain because, Christ, we had this small local flame just a couple of days ago. Debian has these rules that you're supposed to use shared libraries,

right? And if you don't use shared libraries, getting your package in is just painful. But using shared libraries is not an option when the libraries are experimental and the libraries

are used by two people and one of them is crazy. So every other day, some ABI breaks, right? So you actually want to just compile one binary and have it work, preferably forever and preferably across all the Linux distributions. And I actually think distributions have done a horribly, horribly bad job. One of the things that I do in the kernel,

and I have to fight this every single release, and I think it's sad. We have one rule in the kernel. There is one rule. We don't break user space. Everything else is kind of a guideline. The whole security thing, it's a guideline that we shouldn't do stupid shit, right? But that's not a hard rule.

People do stupid shit all the time. I don't get that upset. People break user space. I get really, really angry. I mean, this is something that is religious for me. You do not break user space. And even in the kernel, every single release, I have people saying,

okay, I'm changing this ABI because it's cleaning stuff up. And I'm like, no, you're not changing that ABI because I will crush you, right? All right, so I'll link to the rest of the talk. It's available in WebM. If you'd like to watch the whole talk in its entirety, obviously it's a good one. I mean, Linus, that's what he does now.

I love his frankness about these matters. And I think it's so telling that Linus does not package his own application for the Linux desktop. He makes binaries for Windows and Mac OS X. Tell me that doesn't sting. That's just, wow. I mean, that's really telling. I think we could debate all day long if the system is broken or not or if it's technically feasible or not.

But at the end of the day, when the creator of the frickin' Linux kernel won't package his own software, I think that's an indication of a failure at some point in the system, right? I mean, I think it's – and I don't think like a small failure, right? I don't think it's like a small failure. I think it's kind of like a fundamental failure. Am I wrong? It seems like a big deal. It was a major pain. I mean, at the end of the day, I mean, the fact that someone comically earlier asked, hey, when's the year of the desktop coming? Well, there you go. I think these are things that still have to be overcome.

Big time, right? I mean, okay, so is there a solution? In fact, we've had people write in, Matt, you might recall to the Linux Action Show have been like, guys, this is so frustrating. Is there ever going to be a solution? And I kind of joked. I said, ha, I wouldn't worry about it. I'm sure the systemd guys will eventually come up with a universal package format for all. Well, guess what? Oh, boy. Oh, boy. What? Guess what? You might have heard of him. His name is Lenart.

He's been around. He's done a few things. Avahi. You might be familiar with the Avahi project. It's the auto DNS discovery. You could be familiar with Pulse Audio, perhaps. You may have heard of the SystemD project. You might have been familiar with a company called Red Hat. These are all things you might attribute to Lenart Pottering. Well, another thing you can contribute to Lenart

is fundamentally attacking how we put together a Linux system, changing what defines a distribution forever. And the best part is it's all powered by ButterFS. Of course it is. It's all powered by ButterFS. I'm sure a little systemd is in there too. It feels like a big Reddit meme, doesn't it? All right. Before we get into this, I can see the Mumba Room has a few things they want to talk about. Let's start with Fred and then we'll go to Wimpy.

Fred, what are your thoughts thus far with that Linus talk and all of that? What are you thinking right now? Yeah, so somebody asked him later in the talk if you did watch till the end. You know, he built this parse c front end and then he to solve a problem and then he built git to solve a problem and then now is he going to build some solution to solve that packaging problem yeah yeah and then he says no this is a hard problem and i actually really do think it's a hard oh yeah oh yeah big time i absolutely agree but it's not hard because of packaging issues it's hard because um i just uh so i thought up an example while you were playing the video okay even containers i don't think containers can even solve that problem let's talk about gtk gt and gtk themes so now you have this settings application, which you can go in and choose your GTK theme, right?

And this changes for all applications, right? So the theme, all applications get the same theme. Yeah. So now everybody's statically linking, which doesn't cause problems. But then now everybody is bundling up applications dynamically linked with GTK, but they're just in their own separate container or bundle or whatever you want to call them. Which configuration file do you use to set the theme? Are they separate configuration files? separate configuration files? Do you have a separate configuration file for GTK 2 and 3,

or for GTK 2.12 and 2.14, or for GTK 2.12.1? All of the things, my friend. Yeah, but then you cannot add new settings or add more features to the configuration files without breaking everything else. So how do you do that? And this is a simple problem. That's also not so important. Theming is not that important, functionally speaking. But it's a very good example. It's a very good use case. Something that would drive me nuts.

Yeah. All right. A few of us in the Mumble room were talking about this yesterday, I think. Oh, yeah. Okay, great. And in general, I am very pro-SystemD. I like SystemD, and I don't mind that they've taken a lot of the low-level stuff. That being said, anything to do with package management and changing that scares me.

Because one of the things, my favorite things about Linux, the thing that I think sits above every other operating system is its package management. And I'm not saying that this will screw it up. I'm just afraid this will screw it up. Or at least screw it up for a little while like Pulse Audio did. Well, I'm going to get into how they're going to pull it off in a second from a package management standpoint and things like that. But before we do, I wanted to give Wimpy a chance to jump in, then Daredevil, and then we'll get into the meat

of what the system do you guys want to do. So Wimpy, what are your thoughts at this point so far? Okay, well, I read Leonard's post this morning. That was my morning reading. So I had some time to think about that on the commute into work and i read it again and had to think about it on the way home yeah i had i read it twice too and i'm still thinking about it yeah i was gonna say i think i think it needs a third read to really absorb it but um i think he he and his team are very good at tackling the tough problems in the world yeah yeah and what they've

fleshed out i really like so i noticed somebody referred to it in one of the comments as oh it's it's complicated and i don't think it's complicated i think it's complex but not complicated and if if you're a python programmer one of the edicts in the zen of python is uh choose complex over complicated and it is a difficult problem to solve so there needs to be some complexity to it but i really i'm really quite enamored with the approach and i say i say bring it on i'm really looking forward to a world in which we can isolate the user environment, the framework environment for the desktop and build that with the appropriate configuration and then keep the staple OS separate from that.

I'm really excited by that. So for those of you who haven't read this yet, I bet a lot of you are probably thinking, you know, they're going to use Docker. This is not what... Containers can play a role in this. But in a way, it almost demotes everything to the same playing level. The base OS containers and VMs, in some ways, from a file consistency standpoint, the playing field is a little more leveled. So essentially, you could see why they want to do this.

And Lenart summarizes it very well. And Wimpy, please feel free to jump in as I go over this. He says, what we want to do is an efficient way for vendors to package their software, regardless if it's just an app or the entire operating system, directly to the end user, and know the precise combination of libraries and packages it will operate within. I, you know, if we were to ship a Jupyter Broadcasting media player that just played Jupyter Broadcasting content, was all themed up with our logo and stuff, I'd have no idea what

video libraries you might have on your distro. So I would want a way to know what libraries would be used when I gave you the Jupiter Broadcasting app. He says, we want to allow end users and administrators to install these packages on their system regardless of which distribution they have it installed on. We want a unified solution that ultimately can cover updates for full systems, OS containers like Docker, end-user apps, programming ABIs, and more. These updates shall be double buffered. And what he means by that is, like, if you guys have ever had a motherboard that has two BIOSes or like the way the Mars rover does updates, there's two computers and you do one update.

And then once you have a successful boot, then you update the second install. Or actually in the case of their approach, you have infinite buffering depending on your free storage. But they say we want at least in the minimal double buffer. He says there's an absolute necessity if we want to prepare the ground for an operating system that can manage themselves, that can update safely without administrator involvement, a la CoreOS or maybe some of the directions Fedora Cloud wants to go in, where, yes, the server OS is rolling, but the applications are protected. He says we want our images to be trustable.

Now, this is one that might have some controversy because he means signed. And he doesn't just mean like the application signs. They want a fully trustable OS, starting to sound a little bit like Microsoft here, with images that can be verified by a full trust chain from the firmware, meaning secure boot, through the bootloader, through the kernel, all the way up to the init. Cryptographically secure verifications of the code we want to execute

is relevant on the desktop, just like Chrome OS does now, but also for apps, for embedded devices, and on servers, particularly in a post-Snowden world. So what they are proposing on how to accomplish this, it's amazing. It's essentially a scheme built around a variety of concepts that are available either today or hopefully soon in ButterFS. I want to caveat something here. Everything they're talking about in ButterFS is in a read-only capacity. So you're not really talking about ButterFS manipulating live data to accomplish this so much as basing it off of snapshotted data and working from that snapshotted data. So before you get too ButterFS freaky, do keep that in mind, although you probably are

legitimately concerned. So they want to use the ButterFS file system namespacing. There's a large number of features in ButterFS that neatly fit this concept today, he says, and maintainers are busy working on a couple of others they want to eventually make use of. As part of the proposal, they will make heavy usage of ButterFS's sub-volume capabilities and introduce a clear naming scheme for them. So you would have a user file system with a vendor ID,

the architecture like ARM or x86, and then the version. And these would be volumes that could be mounted and mapped to your file system like a drive. So when an application launches and it needs a particular version for a particular architecture, it thinks it's talking to slash opt slash proprietary garbage slash bin, when in reality it's actually in this deeper ButterFS subvolume that's been mapped to that location. Kind of a little bit like that distribution we talked about a little while ago that would let you run applications from any distro into a whole bunch of cherooting on the background. Oh, yes. Wimpy, am I getting everything right so far?

Yeah, you've totally summarized that correctly. And they talk about how you wouldn't even necessarily have to have a ButterFS file system to do this. A lot of this functionality could be contained within a loopback device that is formatted with ButterFS. So if you still wanted to use XFS or ZFS or Extended 4 for your primary file system, some of this functionality may be capable of – I bet that's limited, but I bet some of the functionality could work out of a loopback device. may be capable of, I bet that's limited, but I bet some of the functionality could work out of a loopback device. Here's a couple of things that are really, really, really, really interesting and would fundamentally change the way software is distributed on Linux. They are taking advantage of a feature built into ButterFS called send and receive.

ZFS has something similar where you can take the blocks from your drive, from your storage device, and send them over the network, over an SSH connection, or over the SATA cable, and write those blocks to another storage device in another computer on another drive. It's kind of cool. It's kind of crazy because you can just take the whole thing, suck it up, and spit it on another machine and write it all down there. Well, they want to use that same technology to distribute software. So it would be bit by bit exactly what you got. And using some of things like ButterFS's deduplication features and things like that, even though you'd have a lot of the same libraries on the backend, ButterFS would be doing dedupe to clean it all up. And they would

do delta differentials too. So if you're doing a send and receive of a ButterFS file system, it would just be the bits you don't have. So you would essentially have some Delta capability there too. Now, they're also saying all of this doesn't even eliminate RPM or Deb because the distro maintainers would still be building the distros with those tools. This would be more for the users. The users would stop using the package managers as much and would use systems more like this. And they paint a scenario where, you know, you would have GNOME and LibreOffice and Firefox all doing this system, and it sort of escalates from there.

Any gaps you need to fill in there, Wimpy, you think? Yeah, I think one thing you've missed there is you mentioned LibreOffice and Firefox, but in this brave new world, you can have the last six versions of Firefox and the last three versions of LibreOffice and the beta release, and you can transparently switch between them. Yeah, and that's, I think, where people think it seems a little complex. And it is complex, but SystemD is already capable of doing some of those switching activities now.

So this is building on what SystemD is already doing with its stateless booting and all that sort of stuff. And in this system, everything, as they're calling it, is double buffered or really enfold buffered, depending on your storage, because user and the runtimes and the frameworks and the app subvolumes can exist in multiple versions. You could roll back after a bogus update. You could see how a rolling OS becomes a lot more reasonable because you could individually or completely roll back

different components or all of the components. You're doing bit-by-bit transfers. Wimpy, what do you think of the method they're proposing for software installation? Does it blow your mind a little bit? It's a little bit different, but there are some parallels to the way that we do some things at work currently.

So we work almost exclusively with Python at work. And if anyone out there is a Python developer, they've almost certainly come across Virtualenv, which is a way of isolating a Python runtime and then installing just the packages and tools that you need for your application into this sandboxed, isolated environment. And it means to say that you can keep that separate from the system and stop the system packaging from stepping on your applications and new libraries breaking your application. And it's more complex, but it's like that. It's like virtual environments, but for all of your partitions. Right.

And in an example, Lenart says, I've got three vendor operating systems installed. He's got, like, I think it was two Fedoras and Arch Linux. He's got KDE and GNOME. And the home directories are shared. And using these namespacing concepts in ButterFS, he's free to mix and match apps and OSs or develop against specific frameworks and specific versions on any operating system. It doesn't matter if he booted the Arch instance or the Fedora instance. He can execute both LibreOffice and Firefox just fine because the execution time gets matched up with the right runtime at launch time. And all of them are available for all the operating systems you have installed.

Kind of, I mean, part of this... Yeah, go ahead. As a developer productivity tool, that's really powerful because right now, if I want to develop against the Mart A1.9 branch, for example, I need to boot into a virtual environment with all of that, actually a virtual machine with all of that stuff installed and developed there sort of separately from my main machine. But being able to step in and out of environments like that sort of at will is extremely powerful and could in some way sort of usher out the hypervisor world altogether. Yeah, and he also points out that one fat binary, if you wanted to,

could include all of the architecture. So you could have sort of like when the Mac was transitioning from PowerPC to Intel and you could run either the PowerPC code or the Intel code with one icon. He says the same thing could be possible with this. An installer image could include the Fedora version compiled for x86-64, one for i386, and one from ARM, all contained in the same ButterFS subvolume. And due to ButterFS's deduplication,

they'll share as much of the code as possible, and when the image is booted up, the right subvolume is automatically picked based on what architecture the system's detected it's had. Yeah, brilliant, isn't it? And he says, here in this new system, it would also make installations of operating systems much easier. You know, you could do it manually, but you could also automate this.

Here is the process to install a Linux distribution with this new ButterFS-inspired subvolume system. You would create a GPT partition table. You would add an EFI system partition to that partition table. You would create a ButterFS volume, obviously. You would deserialize. That's what they call when you write one of these volumes right to that partition table. You would create a Butterfest volume, obviously. You would deserialize. That's what they call when you write one of these volumes right to your hard drive.

You would deserialize a single user subvolume into that Butterfest volume. You would install the bootloader into the EFI system partition, and then you boot the system, and it's done. There's got to be a downside to this other than the complexity. And the other thing is, he says right in here, he's like,

A, we recognize this would take a massive level of adoption. That's why we're putting it out there right now. And he says, we think this could also help us accomplish things like factory resets and power wash features where users could press a button and all of their data would be protected, all of their apps, everything they depend on, but everything would go right back to it. Defaults, that could be incredibly powerful for users. But they've got to get adoption, number one.

They've got to make sure people are comfortable with ButterFS, number two. And the real question is how irrelevant does this make a distribution's built-in package management tools? Because for some distributions, their package manager sort of defines them. You look at Arch or really even, you know, dev distributions, there's folks that will only use dev-based distributions. If you're an RPM-based distro, you're dequalified. This would sort of nullify that differentiator, right? It could do. you could build your framework and application sub-volumes just by compiling and installing the code into them.

And maybe that's the simpler way to do it, thinking about it. Maybe that reduces some steps and some overhead. So yeah, thinking about it, maybe it does kill off the package manager. But it doesn't necessarily mean you have to do that. There's still some value in being able to package things and install things in a reproducible way. Right. Yeah. And in fact, I want to talk about that in a second. Before we move to that aspect of it, does anybody else have any thoughts on, it's early days on this proposal, so I know there's a lot of you haven't chewed on it yet. We'll have a link to in the show notes too if you guys want to read it all, but anybody else in the mumble room want to chime up before we jump to?

Yes. Definitely. So this is one of those things that I actually think, so far from the proposal, I can see the feasibility of implementing it. I can see the benefits of implementing it. But I think we're misleading a little bit on a couple of things that we gained from not doing that before. So one of the things is people are now statically linking and the reason we weren't doing that is because there was this major benefit of updating the library and getting it all updated. Yes, it breaks packages and makes the distro and the package manager much more relevant. However, it does make our applications to be following the path of later security, the path of the latest upgrades that were made in technology,

which is no longer true when you have this kind of hybrid system. The other thing that I also think that this can actually provoke is a situation where you start just bringing everything to a monolithic state and you lose the distro variety. It's not really the software, it's not really the... I think it's the experience that developers from different distros are trying to convey to their users and i think that might be harder to tackle that philosophical issue of getting people on adopting it than the technical sides of it developer i agree that could be a big issue i understand why linus so understand why linus is looking at it i don't package for linux i mean i developed

the kernel i want to use wherever my kernel runs and that's why he looks at it that way and he's not packaging it but to be honest and fair i think this will eventually become something of the past and be simple if you want to develop an application want to target your end users actually currently if you're developing a server application chances are you're developing for a head red hat open SUSE or something like it and if you're doing for the trying to get to the end user you're doing it for a winter and that's simple you are focusing on OS for now the docker idea OS. For now. And having said that, the Docker idea is pretty cool and dandy,

but you're just bringing all your baggage to the desktop of the user. And that's just assuming that our machines are ready to just take it all. And yes, we have the hardware right now. We didn't have it in the past, but I think that's completely bloated. It can be fixed in

tooling for developers, not necessarily on this. And to that point, you know, the storage point, the chat room is mentioning what about ButterFS storage limitations. I assume as this gets worked out, storage will go up and those things have to be worked out too. I know Heavens wants to jump in and kind of bring this

one home, and then we'll move on to hey, while package managers are still relevant, there's some big changes coming that I want to talk about. But first, let's thank Ting. Go over to linux.ting.com right now. Everybody go there. linux.ting.com. First of all, that'll put Linux in your history. Let's be honest. You need more Linux in your browser history because I know what else is in there. And it's not what you want in your history. linux.ting.com. Go over there. That'll also get you a $25 discount off your first Ting device. If you've already got a Ting-compatible device and you want to bring it with you,

well, then good news. They'll give you a $25 credit for a lot of you out there. That gets you more than your first month. So why Ting? It's mobile that is zero BS. You only pay for what you use. Ting takes your messages, your megabytes, your minutes, adds them all up at the end of the month, whatever bucket you fall into.

That's what you pay. It's just a flat $6 for the line and your usage includes hotspot, tethering, picture message, caller ID, all the stuff, everything, all of the things, plus more, all controlled by an amazing dashboard. Ting gives you the controls you need to see exactly where your account's at, maybe turn something off, turn something on, disable voicemail, enable voicemail setup, caller ID, disable caller ID, all those things exposed for you right there in a crazy easy to use interface. Go over to linux.ting.com right now and you can check out their Ting savings calculator. You put your current usage

in that bad mama jam and not like what you pay, but your actual usage and just get an idea of what you'd save by switching to Ting. If you're listening to this right now between 8 a.m. or 8 p.m. Eastern, why not try giving Ting a call? 1-855-TING-FTW. Guess what? No hold customer service. They'll answer your questions right then and there on the spot. It's kind of incredible.

And if you want to sort of charge up your Ting device, if you're a recent Ting customer, they just posted a great top five app pick from one of the Ting associates, Ben, and he talks about some of his favorite apps for Android. He's got Waze on there and a couple of others for those of you in relationships or not in relationships too, or that one, which might break up a relationship. He's got a whole list over there on the Ting blog. You can go over to ting.com slash blog, but do me a favor. Start by going to linux.ting.com, linux.ting.com. linux.ting.com.

Go grab a device. I've got the Nexus 5 on Ting right now, and I got multi-ROM on there. Well, Rika did. And I've got Sailfish OS on that bad man, a Firefox OS on that guy, Ubuntu Touch, and Android L all on the Ting system. It's cool because there's also other Ting customers who are doing the same thing. And every now and then I chat with them in the Ting forums. Yeah, Ting has forums too. They have a great community where people are trying out stuff all the time, always pushing the envelope. Linux.ting.com.

Go find out why so many of you out there have switched over to Ting and been super happy. You might just save enough money to buy a new laptop every couple of years. Linux.ting.com. And a big thank you to Ting for sponsoring Linux Unplugged. All right, Heavens, did you want to jump in before we moved off of the uh new universal package format that's going to save all of us proposed by lenart and the system decabal oh may i yes sir you may all right sounds wonderful well anyway this is okay what happens if this wasn't from Lenar?

Would people be more accepting of it or not? I would actually be more skeptical. Yes, well, that too. But there is another situation here, which... Okay, most of the time... Or way back when SystemD was just coming out, we were saying, there is a here of linux that we are going to come to that we have never seen before and guess what this is one of those side effects

that the creation of systemd has spawned it's actually one of those future things that we pretty much need because what libraries that's a figment of our gcc linking past just because we need to statically or dynamically link to a certain library in the system that's pretty much gcc's legacy just as rolling distributions and statically released distributions this lack where model versus the rolling model I think those are a figment of the past. This is a very good headway into our future when it comes to partitioning the system, even though I'm not very much

of a big fan of this ButterFS. It just so happens to provide the technology required in order to make this feasible, or at least not insane. Yeah, I agree there. It does seem like it is a glimpse of the future that kind of has to come.

And Crossroads... And we know. Well, we have to really... Well, yes, because Lenart, he has the overview of the system which none of us have ever seen before because this is the future which we are coming to. No one has seen the abstraction level that he has.

He is at that point. He's the epitome of where he sees all these distributions and all their problems. He's the person at the forefront. If he can see all these little weird discrepancies and can have a reasonable way, or at least an idea on how to bring us forward, he is in a pretty good position when it comes to his position, where he is and what he sees to propose something like this. All right. So I want to give Crossroads a chance to chime in because he's been waiting. Crossroads

too, though. Is this kind of a big win for Red Hat if all distributions are more like Red Hat, then Red Hat wins? I don't know about that. However, I think we can all acknowledge that this is an issue. That developing software for Linux is more difficult because of Linux's variety. And I think we can all recognize

that that's an issue. And once we realize it's an issue, we have to think who can solve this? Who in the Linux community, or what in the Linux community can solve this, right? Systemd, I think, is the only thing that can, because with the exception of Gentoo and Crux and a few of those other things, Systemd is on every Linux distribution, is or will be, on every Linux distribution. So if there's going to be one thing that can unite the package management of every Linux distribution, it has to be system

B. Yeah, I agree. I know that's why I made that joke a few weeks ago in Linux Action Show, is it's going to almost have to be. And you almost have to have a company that has some real reason to put that much effort into the game. And I think Red Hat is that company. Uh, Wimpy, I know you wanted to kind of, uh, did you want to touch on anything else before we wrap up on the topic? Yeah, I think Crossroads and Heaven's Revenge have, have totally hit on the right point

here. Um, for, for decades, Linux has often been described as a Unix-like operating system. It was neither one thing nor the other. It wasn't. I mean, when I worked at Sun Microsystems, we used to joke that it was Solaris for a pro and Linux for show. Oh. Yeah, oh, indeed.

And I think that what SystemD is doing is it's very clearly setting out the future of Linux, and it's giving Linux its own identity for the first time a distro agnostic identity and Linux and system D are going to be the future of

our platform and I think people are either going to have to choose to adopt it and learn it and like it or they're going to have to choose to go to yeah definitely but I mean I've said that despite it and learn it and like it. And maybe now ButterFS is in that. Yeah, definitely. I've said that despite ButterFS' foibles, I do

use it on my root file system. Yeah, that's why I've been using it too. I felt like this was coming too. Linux has to have a file system with these capabilities, and this entire initiative gives you some insight as to why it's necessary to have functionality like this to have an operating system to overcome these limitations despite

its problems. Yeah, and I think the only question that leaves to be debated is do we continue calling it GNU slash Linux or systemd slash Linux in the future? Wow. We'll wrap the discussion

there on that. We'll talk some more. I know Fred's got some stuff he wants to we'll wrap the discussion there on that. We'll talk some more. I know Fred's got some stuff he wants to chime in on. We'll do that in the post show. But the show's gone long, even though we didn't get a chance to get Aaron in or the DNF discussion. So and Eric's been looking at DNF, too. So maybe we'll just bust that into its own segment next week and talk about Fedora's new package manager, because we got to call it right there because I got to get home. But a fascinating conversation. If you're curious, go read Lenard's post. He wants as many people talking about it as possible right now because I think he acknowledges to make this work, to make this successful,

it's got to have huge community adoption. Buy-off. You thought SystemD had to have big buy-off? No. No. I mean, this is right. I mean, think about this, Matt. Universal package management using the file system across every Linux distribution.

You just have to have ButterFS support and this version of Linux kernel. I mean, that's the biggest change ever. It sounds awesome, but there's so many. It's huge changes. I mean, it's absolutely mountain moving. And it's one of those things to where you can look at it seven days from Sunday and you think you got it all figured out. And then you're thinking, oh, wait, what if you do that, though? What if you do that other thing? There's so many

things. How will it work on mobile devices? How will it work on embedded systems? All these kinds of questions. How will we distribute software to switches running Linux now? It's fascinating and it's going to be one I think we will probably track for a while as it sort of develops and grows and we'll be

curious to see A, if it makes it and B, how much it changes from this point. But I think tonight I'm going to give it another read. I've got it saved up to Instapaper. And I'm going to give it a go-through. But Matt, guess what? That brings us to the end of this week's Linux Unplugged. Now, I don't know exactly what we've got scheduled for Sunday.

I'm waiting to hear back from a guest. But I already have kind of like a let's get serious about, A, productivity under Linux, but, B, how to take the Chromebook and actually make all these Chromebooks that are coming out right now a drop-dead awesome Linux experience, not a second-class experience, but an awesome experience, how to do that. I might cover some of that on Linux Action Show. We've got a whole range of topics.

So, Matt, I'll see you on Sunday, okay? All right. See you then. All right, everyone. Well, thanks so much for tuning in to this week's episode of Linux Unplugged. Don't forget to join us live. We'd love to have you here on a Tuesday over at jblive.tv, jup All right, everyone. Well, thanks so much for tuning in to this week's episode of Linux Unplugged. Don't forget to join us live. We'd love to have you here on a Tuesday over at jblive.tv, jupyterbroadcasting.com slash calendar to get in your local time zone and jupyterbroadcasting.com slash contact to send in your contact info

and Jupyter Broadcasting or also our subreddit, linuxactionshow.reddit.com is a great place to go. All right, everyone. Well, if we don't see you on Sunday for Linux Action Show, I hope we'll see you right back here next Tuesday. And Fred, why don't you in the post show now get a chance to jump in with what you were going to say? Because I know you were mentioning you're seeing some problems if the distributions kind of normalize out. Yeah.

First of all, you shouldn't feel that bad. I mean, you have to run the show. And for me, it's really no problem if I don't get to speak. But so first of all, I'd like to say something about Linux or at least how I see the whole distribution thing, you know, spanning out. So essentially, it's free software and open source is an evolutionary type of thing, yeah? So make the best win, yeah? And distribution is essentially like branched off

to solve different kinds of problems, to sort of even solve the same problems in different kinds of ways. And it's time to converge back by taking the best of each, right? So, you know, it's really not a problem if they all just converge back, right, into a single system.

It's actually a good thing because we get the best of each of them. That's a good perspective. I like that. Yeah, I actually think it's like the natural course of things, right? It is the sort of organic evolution that open source seems to take's it's it's like the natural course of of things right it is the sort of organic evolution that open source seems to take where it sort of forks out the strongest survive and then it gets reincorporated and we kind of focus on that thing exactly exactly and also on that point um you have so much of the differentiation between the different distributions

is the low level plumbing that defines them so So ultimately it's the package manager and the release cycle. If you take package management and release cycles out of the equation and Upstream is now responsible for distributing their binary sub-volume, the distributions are going to have to focus on the end user experience to differentiate from one another rather than technical masturbation. Yeah, yeah. And philosophy and things like that. Oh, I wanted to give, before we go too far there, I meant to give a quick plug to Aaron

Saigo. He has a Patreon page up. He is bringing back Illumination of Free Software, kind of like a talk show, discussion show. He's going to have it structured. He's looking for backers. You can go to patreon.com slash L-O-F-S and I'm a backer.

And he's got a video up there talking about what he wants to do. And, you know, we talk a lot about getting good open source coverage and I think the thing about a Patreon campaign, just like a Kickstarter, you can back a Kickstarter and get absolutely nothing, or you can back a Kickstarter and actually get the product,

and I think it depends on who's running the project. I think if anybody's going to be able to pull it off, it's Aaron. I think if he sees success with it, he'll keep at it. So go over to patreon.com slash L-O-F-S and kick him a few over on Patreon if you want to. I want to give him a plug for that. We're going to have him on the show, but we had some mumbled connection problems. Okay.

So what about the fact that this is powered by ButterFS? That does seem to be the elephant in the room that we're just kind of glossing over. Yes. Sorry, Fred. Go on. No. I have a much bigger problem here.

So Leonard is the guy who should be doing this. I absolutely agree. He has the vision, he has the view and everything. But the problems he tackles usually, they're quite big, they're complex, they're made of many small parts. And he has always a grand vision, but never a stepwise vision for solving something. He never gave out like, okay, now we're solving this little piece, and then that little piece on top of that, and then and so on. He just goes out and says i'm groundbreaking i'm changing

everything and you know it's good for everybody might be good for some people might not be good for some people yeah until this day i've been running arch with systemd on my laptop and i have some serious problems because i encrypt so i raid my encrypted partitions i don't encrypt, so I raid my encrypted partitions, I don't encrypt my raided partitions, yeah, and systemd does not handle that properly. So there are special cases where... Sorry? It actually does if you handle the dependencies correctly for your file system startup if systemd is in your initrid, although I don't think Arch has systemd in the initrid. Hmm. I don't think Arch has systemd in the initrid.

What a funky way to say initrid. So I've read quite a bit. I have went on the systemd RSC channel asking. I've read a lot of the documentation. My system starts up fine. It does not shut down properly. And then actually for debugging shutdown, by the way, I have to log to a text file. So I'm back to, you know, sysv init style of logging.

I cannot use journal CTL for some reason, this journal D stuff. So there's a huge fuss. And then there is this one corner case that maybe somebody did or didn't think about and that he didn't manage to cover or the or system didn't manage to cover or think about it somehow yeah and this is the kind of problem we have in today's technology and linux is gonna turn into windows if we keep doing this stuff windows works fine no i'm serious yeah i'm serious because windows works fine for the general case not for the corner cases right and well for our general this is not for the corner cases, right? Well, for our general case. This is why Linux was so flexible and susceptible on a use case basis.

That's why it worked fine for so many technical people. But if we're going to keep skipping that, we're really going to end up quite badly. I also think I want to, you know, Blackout24 in the chat room brings a good point that it's not just a Lennart effort. I think he's the name that has clicked the most with everybody because it does add some credibility to the effort, in a sense. He got his way.

Yeah, but I don't want to, but I guess I don't want to over-cult a personality this thing, too, because it is not just his thing. However, let's not kid ourselves. He will become, if it's successful, he'll have to become the figurehead of this. And it will become Lenard's baby. Well, he's done it before. He might be able to do it again.

Well, he's just, simply put, he's one of the most followed online. He's one of the most well-spoken from like a, he'll come at you with a technical rebuttal that is both scathing, but yet also technically accurate, so therefore hard to quibble with. He's really good at that. And he's got the name brand. So, I mean, it's not, this is not a Lenart quote-unquote initiative, but I think by the end of it, it'll look like one. A lot of the time. I just think this trend continues forward is

that we keep seeing innovation starting around the red hat fedora type people like lenaert so it's i just wanted to point that out that's pretty interesting i wonder if our pace of development is getting a little bit too quick in a way that scares most of our technical users now. Oh, one Linux that fits all is not bad. I'd say... Well, you know, that's the thing that we have. Our problem that we have so far is our variety. A lot of people bitch about there's too much choice in Linux.

This helps, you know, unify. Just choose a different... A lot of people just like to do this. I mean, I think part of the problem mean i think i think part of the problem i think we are part of the problem each time that we go and we say we're using linux i mean seriously i'm not coming for for the new plus linux thing i'm just saying you use a distribution and the distribution represents a whole set of things not only the kernel they could have changed they could actually change that should have mattered to. When you're using your computer, like for us technical people, yes, we care about it.

But let's face it, a regular user comes to the desktop and he doesn't know or care about what kernel is in there, what is going on on the background. If it's called Unity, if it's called GNOME, they don't care. They care it works for them. And that's the experience they want. And that experience is under a name of the distribution that's the name you you you go for and if you do that a lot of the issues of oh it's so fragmented just don't exist that exists for us we like to bitch about it but let's face it the user doesn care. The user just wants it to work.

Also, if we don't allow for change, if we don't allow for innovation, we can't move forward. Of course, but just a minute. He has a ton of distros, so it can always happen. He can do it in the distro he's on. can do it in the distro he's on and if people in there like it some people like everything else when it proves useful and it proves to be worthy it appears on the other's distro that's how we work yes thank you thank you i'm trying to avoid the whole celebrity hacking topic because you know it's funny that not to even get into like the body parts aspect of the celebrity hacking but i feel like so much of what we've been talking about, building your own cloud team for,

you know, fortress Chris or going with own cloud or, or sync thing or bit torrent sync and, and all of the, you know, encryption and all of this stuff. And then you just have this huge story that drops this sort of the personification of all of the stuff that we've been a little worried about.

People have been a little nervous about, and you know, it, nervous about. And it has flavors. If you zoom out and look at the meta story, it has flavors of NSA spying. And now it's a new angle. Look at that. It has flavors of personal privacy and protection. It has flavors of controlling your own data. It has flavors of what the responsibility of these new corporate overlords that are sort of forcing you into these new systems and these new ecosystems that just do this stuff for you. And at the same time, when one of their marketing messages is, don't worry about how it works. That's the selling point.

You never have to think about it, which then encourages exactly these kinds of things to happen. So I think it's a huge meta story. So yes, there's girly parts that are involved, and of course there has to be that adolescent aspect to it that sort of gives it the traction. But I think when you take that away from it and look at more of the violation and privacy aspect of it and the implications that this, I mean, this is happening to very public figures. And so the reaction is very public and a lot of people are talking about and a lot aware of it. and a lot of people are talking about and a lot aware of it. But I was just saying a little while ago that if you actually follow the news on this kind of stuff,

this kind of thing happens almost on a weekly basis on a much smaller scale, and usually it involves kids and it involves a high school and the shaming of someone and then that someone either sometimes gets suicidal or they have to leave the school. There is a very real-life aspect to this story. And I think sometimes when we start talking about it from a technical aspect, we'll talk about it from a, oh, well, they need local encryption. Or we need transparency. Or the solution is open source code that can be publicly audited. much smaller scale and probably realistic on a daily basis, really, if you think about the amount of devices around the world that are out there that this kind of stuff happens on, there's probably these kinds of violations going on almost daily.

But this story has given us the opportunity to actually stop and think about it because it involves people with names we know and faces we recognize. Although these women have had an egregious privacy breach, for which there is no apology by anybody, I'm not so sure they were the target of this. Do you not think that Apple were actually the target of this, in that the timing of this seems to be very precise? Oh, snaps. Look at you, Wimpy, busting out the be very precise. Oh, snaps. Look at you, Wimpy, busting out the conspiracy bacon. Very nice, sir.

Very nice. Where's the bacon from? Well, okay. So, you know, they're about to have their big unveil. This is Apple's time of the year. They're going to introduce a new payment system supposedly partnered with Visa, MasterCard, and American Express, right? Built on NFC, probably connected to your iTunes account, I'm sure.

Yeah, and you look at the scale of what was disclosed improperly. This wasn't something that just happened overnight. This must have taken weeks of planning. Yeah, I agree. So I think that the timing of this is very deliberate, and I think it's designed to embarrass Apple. So it definitely accomplishes those things. It publicly shames famous people for using iProducts, which is something that, you know, it's one of these things where you'll see, like, Samsung will pay a celebrity to take a selfie. And, like, Ellen, for example, she'll be out on stage.

She'll take a selfie with a Samsung camera, and then she goes backstage and tweets about it on her iPhone. And that drives people crazy. So Wimpy, are you implying then that this was orchestrated by someone? Well, I don't know who it might have been orchestrated by, but I'm wondering if the motivations of the individuals behind it was to sort to slander Apple and use some salacious means in order to accomplish that.

Maybe it was an activist that wanted to derail the centralization of the payment system, right? Maybe it was like... Possibly that or possibly it's just simpler. I don't like Apple and I'm going to show them in a bad light because I want product X to win the mobile wars. I'm watching a video of my daughter right now on the baby monitor at home. I think that bacon was brought to you

from Frank's Meat Market in Black Diamond, Washington. Ha ha! Oh, man. But yeah, I think that reasoning about us activists or just a few people, because the nature of the simplicity of the brute force doesn't look like there's been that huge of a technical effort behind it. Right.

And also, the other reason why i think this is a an attack i'm not sure if that's the right word an attack on apple is looking at some of the files that were taken it's implied that other storage cloud storage platforms were potentially compromised but those cloud storage platforms haven't been mentioned. They're very squarely pinned this on iCloud. I do recognize that as well, but I also, I've noticed just in the tech news industry, quote unquote, if you can even frickin' call it that,

if the story has Apple, you put that in the headline, it does not matter anything else. Because A, headlines are of limited length, and B, it seems Apple headlines drive traffic. You could put Microsoft, Oracle, Amazon, Google, eBay in the headline, and they still drive less traffic than when you put Apple in the headline. So I think that might explain that aspect of it, is that's just the click-baity nature of the tech media. That might explain that aspect of it. That's just the click-baiting nature of the tech media. But I do, yeah.

I mean, I think also if it is a simple, if it is truly as simple as a password breach and perhaps aided by an inadequate lockout timer after a certain amount of invalid attempts, then this is truly something that could affect any cloud service. Because, for example, if I have a script and a computer and time on my hands and say Dropbox allows me to do nine invalid login attempts, on the tenth attempt it locks my account. Well, then I would just make a script that tries to log into Jennifer Lawrence's Dropbox account nine times and then sleeps for the hour or whatever it needs to sleep and then tries another nine. And I would let that sucker run for a year, just like people Bitcoin mine when they get nothing, because if you get

anything at all, it's going to be such a payday. It's totally worth it. So, yes, it absolutely makes a difference if you rate limit the login attempts. Fundamentally, though, if these kinds of things, this something, this private, something this important is being protected by a password and a password alone. I think that's a failure of the system. And I think you can give some of the blame to the user for using a dictionary password. But I'll tell you guys, I've been guessing people's passwords since I was literally in elementary school. I figured out my teacher's password just by looking around at their pictures. I've been doing it for years. Nobody has changed. I mean, I'm getting to be an old person now, and I'm telling you, for 20 years now, human beings have been using the most

ridiculous, 25 years, people have been using the most ridiculous passwords to log into their computers, and that has not fundamentally changed. And, you know, it's funny because people think they're really good at picking passwords, and they think they're really clever, but I've been able to figure out my mother's passwords, my father's passwords. I won't say other people's passwords. They might be listening. I've been able to figure out teachers' passwords. Human beings are not good at picking passwords unless you are educated on how to do it.

The nature of this attack was to use the Find My iPhone, which I believe you need to supply an email account and a password combination. And once you get a valid credential, just think how many systems that can unlock. Bearing in mind, the majority of people use the same password everywhere. Email accounts are often used as the assert your identity field. So you've now got an email account and password and you can just go around trying that at Google, Dropbox and everywhere else. Well, for most of those,

you don't even need to try the password. You just need the email. Yeah. Because you take your email and you go to whatever the service is and go, I forgot my password. I mean, that's one of the brutal things about current online identity is that it's built around your email and maybe your phone, two things which are remarkably easy to get your hands on. Right. And especially when you have physical access, like if you can sit at somebody's desk and then you go to any website that they ever visit, and the chances are, even if that website's logged in, their Gmail or Yahoo or AOL mail is logged in, and you just initiate a password reset, it sends the email to a tab they might already have open on their computer,

you open up the email. I mean, it's ridiculously easy to do this. We talk about this constantly on TechSnap because people really, really, really underestimate the quote-unquote value of their email inbox, and really that's where it's all at. underestimate the quote unquote value of their email inbox. And really, that's where it's all at. And I think, too, like we have to ask ourselves if maybe if you can't consider a service secure unless it requires more than a password. And would it be ridiculous, A, for a company to require two factor and B, if any company could pull it off, wouldn't it kind of be Apple? I mean, they do all kinds of stuff that's just like, you can't do that. Then all of a sudden it becomes

normal. Like what if Apple like required to make iCloud changes like this, you can't do that. Then all of a sudden it becomes normal. What if Apple required to make iCloud changes like this, touch ID and a password? If you have an Apple device that has touch ID, then they just require you use that as a form of two-factor authentication. They could do it. At least what I see among my peers and people that I know is the reason how they justify using the same password

and the same weak password is because they think, oh, no one will attack me. Yeah, I'm not important enough. Yeah. But they don't realize it's not them that someone's going to try to hack. They're going to attack Facebook or some site they have an account on, get all those accounts, and then automate the attacks against everyone else. Right, yeah, yeah. Yeah, then everybody just gets swept up in it.

So back to the app thing that you mentioned about, you know, they could do it. What would be their motive? Like what would motivate them to do this? Marketing? I mean, it's a good marketing point, right? So two reasons why I think Apple could be the company to do it. One, because they'll do things that seem almost like suicidal sometimes. Like they'll launch a product that kills another product of their own. The other reason is because they're not making money off the monetization of

the data. They're making money off the monetization of the hardware. So Google and Microsoft, both of them monetize the data for advertising purposes. But Google's way beyond that. We all understand that. So Google, for example, like even though they're working on encryption for Gmail, I would be pretty impressed if they ever actually encrypted the actual email contents that they couldn't read themselves because that then they lose out on a huge AdWords opportunity that would probably literally cost them millions of dollars. But Apple, if they can't read my data, if it's impossible for them, if it's encrypted at the operating system level on the device using their 64-bit

fancy processor, they lose nothing. It doesn't cost them a dollar. Wouldn't that take longer to send the message though? I mean, that would be one of the... It's not a sexy feature. It's just not a sexy feature. Yeah, that's what I'm saying.

It wouldn't be a time thing. I'll tell you what, after what's happened, it's a sexy feature. Yeah, that's what I was going to say. It wouldn't be a time thing. I'll tell you what, after what's happened, it's a sexy feature now. Well, maybe. Maybe. I was hoping the NSA stuff would make it a sexy feature, but that seems to have sort of petered out. Most iPhone users don't care. I mean, they just don't care.

Most people I know really don't. It doesn't matter to them. They don't care. The thing about my generation is that we've grown up with Facebook and we've grown up with sharing everything about ourselves. So most of us, and I'm not included in that, but most of us

really, it doesn't matter to us if our information is out there. Well, maybe to an extent, but so take one of these friends you're talking about. They care if all of a sudden their dick's on the internet. Yeah, although as long as their dick isn't on the Internet, they'll just say, oh, well, he shouldn't have done that. Right, I agree. But every time a dick or a boobie gets on the Internet, I think it brings it a little more home. Like because now it's somebody you know.

It's a face you recognize. It's a body part. You know, it's not an abstract concept anymore. It's a dick or a boobie. It's not an abstract concept anymore. It's a dick or a boobie. And I think that is more powerful than an abstract PowerPoint presentation published by The Guardian. That is true. So maybe in regards to the nudes they're sending, sure, that might get them to care a little.

But especially about all the stuff that Google collects, it doesn't matter to them that everyone has a huge database on their life. I bet amongst Joe public, this story has resonated more and made them think more about their own personal privacy and how they use their, not just Apple devices, but how they use these devices

than anything that was disclosed by Snowden and The Guardian over the last 18 months. Yeah. Yeah. I pictured Labor Day weekend here in the United States was millions of U.S. consumers going through and purging their photo streams.

And then there's the other half of them that were actually going and looking for those other photos. Right. And I'm thinking, hey, you know, maybe I should be taking, that was my thought process, hey, maybe I should take some nudie photos. And how would I do that exactly? Polaroid. That's what I was thinking. If you want to succeed, you've got take some nudie photos. And how would I do that exactly? Polaroid. That's what I was thinking.

If you want to succeed, you've got to have a sex tape. Jupiter Broadcast would really be catapulted into the spotlight, but not in the right spotlight. I feel like a calendar train wreck. That'd be – All right. At work, when Heartbleed hit, I just quietly went about the business of patching our stuff. And then at the next board meeting i said you know there was this thing and did you hear about it nobody knew what i

was talking about really well we weren't affected in the first place but we've done the necessary things today my boss comes up to me and says can i have a word with you for a minute and i say yeah sure i go into his office and he says so so what do you know about this iCloud hack with all of these celebrities? Really? So it is grabbing. So I think this story resonates. Hardly, no, not so much.

Not on the radar. No. But then this, everyone wants to know. Well, there you go. It's got everyone's attention. It's another reason to brag about the you have an iPhone. Boobies bring us life, and maybe they'll finally bring us

some privacy and concern about security. I feel really bad for those gals that, or the women that had to, you know, have this happen. Oh, I do too. I can't imagine if that were like my wife or whoever. I know. It's really awful. And I think it's important to remember it's not just them it happens to. It's happening on

a smaller scale that is affecting lives all the time. And we can't do a lot in our position, but as advocates of free software and as folks, I think a lot of you in the audience are those of you who work with your friends and family to help them do these things. You have an opportunity to at least help the people around you. Show them alternatives. Don't be heavy handed about it. If they want, you know, like my mom's using iCloud, I let her go ahead, mom, have at it, whatever. But if you want an alternative, I'm there to help you set it up. It's more secure. Here's the benefits. We don't have to be heavy handed about it. But as folks in our position, I think we do have some opportunity

if we want to engage in it to help people with this. Yeah. You know, I was just thinking back. It was a couple of years ago i i saw this on abc there was this girl who got hacked she was like maybe 14 years old at the time and her life was essentially ruined by all these pictures that she was taking for her boyfriend who i believe she's now engaged to or something or at the time of the thing and they they were kind of risque photos in a lot of cases and her life was ruined from that point everybody's once her account got hacked yeah uh her high school friends you know it spread all over the internet and her high school friends uh called her a slut and whatnot and she probably

wasn't that way but you know the fact that they were treating her that way is just sad. And a young person is probably not properly emotionally equipped to deal with it either, which is sort of the unfortunate part. Exactly. She had to withdraw from school and everything. And, you know, this happens every day. And a lot of people don't understand that and don't recognize it. And I hate that it took a celebrity getting hacked or several celebrities, I guess, before this became a public problem.

It's like natural disasters. People don't care until it happens to them. Unless they have sympathy. What is it? One death disaster? We're kind of a martyr for our security systems at the moment. And at least right now, we know there's a problem, or at least normal Jews of America,

and hopefully the world, because it's not just just America celebrities that got used here in this situation. The world's or at least a lot of different countries got involved in this. At least everyone will know that there's a problem that needs to be fixed versus just the ladies being exposed or used this sort of publicity. I just hope we don't have a overreaction from the political side of the spectrum, but we'll see it. Now, I think we'll leave it at that.