LINUX Unplugged - Episode 39: Fragmentation Timebomb | LUP 39

Episode Date: May 7, 2014

In the not too distant future the Linux desktop will face a landscape comprised of users running Wayland, Mir, and X11. Ubuntu will be rolling out their first generation Qt based desktop environment, and developers are crying fragmentation.But how would we shape the future if we could wave a magic wand? And is fragmentation a real problem in practice?Plus: Our thoughts on Magea, producing video content on Linux, and much more!

This is Linux Unplugged, your weekly Linux talk show that's still not exactly sure what the Internet of Things even is. My name is Chris. My name is Matt. Hey there, Matt. Episode 39, buddy. 40 is just around the corner. Are you feeling it? I am feeling it. Yeah, now, it feels like we're, I don't want to break too much because this week Coda Radio just said episode 100, which is pretty exciting. So congrats to my co-host over there, Mr. Dominic, for 100 weeks in a row, too, which is pretty awesome.

Now, Matt, you know, coming up on this week's show, there's been a couple of things that have been sort of gnawing away at the back of me. We look at the Linux landscape in the context of where things are going to be in a few years, and there's a lot of change. You've got the number one Linux desktop. Ubuntu is going to make huge changes over the next few years, roll out new applications based on Qt. You've got Wayland

that's going to be in the market at the same time that Mir and X11 are going to be in the market. That's going to be a sort of fragmentation. But then also you've got this other end of it, this Internet of Things. It's going to be a bunch of little devices from light bulbs that run Linux, don't laugh, that's actually a thing, to devices in your kitchen that control your toaster, Don't laugh. That's actually a thing. To, you know, like devices in your kitchen that control your toaster, all running Linux, all different iterations of Linux, but none of them quite compatible with each other. So there's a lot of types of fragmentation that are coming up. So I want to talk about some of those this week, identify some of the ones that are concerning us.

And then what I'd love to kick around with the mumble room is like, hey, what would we like to see the outcome be? What would we like to see the outcome be? What would we like to do? Like if we could just wave a magic wand and the future turned out in a certain way, so that way applications could be written for Linux desktop and it wouldn't matter what distribution you're on. Things like client-side decoration could just be magically cleared up. The QTGTK war automatically sorts itself out. Like if we could just pick from these things, what would we pick and where would that leave us?

I want to talk about that in today's show. We'll do that in just a little bit. But Matt, there's something I like to follow up on from time to time. Oh, boy. Yeah, that's a Valve update, Matt. We got to do a little Valve update because I think this is, the work that Valve is doing is one of those interesting stories in the Linux landscape where you have a vendor who's working on a lot of open source infrastructure. The end result is it makes their commercial platform more successful. It makes they sell more closed source software, some of which or most of which is even wrapped in DRM. But to accomplish that ends, the means is improving open source. And I think this is a really interesting dynamic for us to watch. So our Valve update this week is a story that ran on Pharonix yesterday.

Improvements to Mesa, or Mesa, done by Lenard G and sponsored by Valve are now an open source patch set. Get ready for this. It's a total of 21 patches that these changes reduce the startup time of applications that can handle deferred compilation of GSL shaders.

Like, okay, that's interesting. That's good. In practice, after these patches are applied, a game like Dota 2 is running with a 20-second reduction in loading times across the board for Intel Power Gigabyte Bricks Pro systems. Valve's already merging those patches into

the SteamOS Mesa branch, and will be shipping as part of their next driver release cycle for SteamOS, which will be open source. This is the third or fourth story now we've done where there's some boring piece of underlying code that was apparently just ripe for the optimization, and now we all benefit from that. And I think this, what we're going to be talking about today is going to be a lot of commercial companies

that are building on the backs of open source, but when it's done right, and I don't want to sit here and just be the backs of open source. But when it's done right, and I don't want to sit here and just be Mr. Valve fanboy, but when it's done right, like Valve is doing right here, everybody benefits. Well, and it seems like they really nailed it because, I mean, like, you know, at first pass, someone hears this and they think,

eh, whoop-de-doo, they're doing something, whoop-de-doo, you know, Valve does stuff kind of thing, right? Yeah. And then you actually see the benefit of it and you're like, you kind of do a double-take and think, oh, hey, whoa, now I care. 20-second reduction in load time is like something that I can actually grok. I understand what that means, and I want that on my games.

Definitely. Pretty cool. Definitely. It's all early days, but it's just, I don't know. It's the right direction. I really like what we're seeing. I really like the way Valve is going about this, and I just kind of want to highlight it because I want to encourage it. Before we get to our first email, we got a Magia question that came in.

And he's going to put us to task, Matt. He's going to stick it to us. And rightfully so, I will add. Magia has been missing from the discussion a lot. So Aaron's going to talk to us about that. But first, I want to thank our first sponsor this week, and that is Ting. What is Ting? Ting is mobile that makes sense.

Ting is my mobile service provider and Matt's mobile service provider. And this is why I'm going to tell you about him right here, right now, because you've got to know about Ting. Here's the best part. No contracts, no early termination fees. You only pay for what you use. You only pay for what you use. It's a flat $6 a month, and then it's your individual usage on top of that. Ting takes your messages, your megabytes, your minutes.

They add them all up at the end of the month. Whatever bucket you fall into, that's what you pay. It included hotspot tethering. Also included no-hold customer support. And they have some serious devices. But what is also great about Ting is they do some fun initiatives. They've just launched their Drop Your Data Challenge. This is one of those things that I think is brilliant because not only is it as a nerd, I'm ready to take on the challenge of figuring out how to game

the system and get my data usage down. I'm all about that anyway, save me some money. This is one of those challenges that you save money while you're doing it, but they're also opening this up to non-ting customers. You don't have to be a ting customer to participate. In fact, non-ting customers can take the Data Savings Challenge and be entered a random drawing to win $1,000 cash, or three people will get $150 of Ting credits. Now, of course, Ting customers, you can save even more while you're doing this by bringing your data usage down. And if you're not yet a Ting customer, in addition to being entered to win a random prize, they'll give you a Ting credit for the percentage of your data savings from week to week to help you make the move even better.

So, in other words, if you cut your data usage by 74%, then Ting will give you a $74 Ting credit. So they're already starting up. You can go get it right now. They've got an app you can use to track this for you, or they'll take the native stat tracking in the Android or iOS system settings. Pretty nice. They've got a couple apps they'll recommend. And they've even got a Google Calendar where you can add to your calendar for the length of the Drop Your Data Challenge. If you want to do this anyways, because why not? You drop your data usage, you're going to save money, you're saving battery, you're lowering your bill, and you have an opportunity to win. So go over to linux.ting.com.

That's where you can find out more, and that's where you can get started. linux.ting.com. Also, if you want to make the switch, linux.ting.com will take $25 off your first device. If you've already got a device, like the HTC One, or maybe you've got an iPhone 4, 4S, or 5, maybe you've got an S3, an S4 you want to bring over, well then that linux.ting.com URL will just take $25 off your first month of Ting. Actually, however long the $25 off your first month of Ting. Actually, however long the $25 lasts. Since it's only $6 a month, it might last you a little while. Linux.ting.com.

And a really big thank you to Ting. And go try out their Drop Your Data Challenge. You can find out more at ting.com slash blog. Go to linux.ting.com. First, that lets them know you appreciate them supporting the show. Okay, Matt. That's tough. Let's get honest.

Let's fess up. Let's read Aaron's email about Magee. Okay. He says, hi. I'm a follower of Last and Unplugged. the show. Thanks for and keep up the great work. So I I'm going to be honest right now. For some reason, I have not been particularly interested in checking out the more recent Mijia builds. Have you have you downloaded an ISO and thrown on a rig at all, Matt? Honestly, never had a real probable, you know, real reason to. I mean, I'll open up to you. Anybody in here want to take the Magia defense position or maybe why they're not interested in Magia? Because I'll tell you where I'm sitting at, Mumble Room, is I'm kind of like, okay, I love that they're still going. That's great. More power to them. But I feel like my needs between

all the other distros are pretty much already solved. What is Magia offering me that I don't already have anywhere else? I think it's a good somewhere else to go if you don't like what options you have on offer. There's a guy I met at a recent conference who said he left and he contributes to Mageia because he wasn't accepted in whatever other community

that he was trying to contribute to. Are you saying it's like a Linux refugee camp? I think so, yeah. But in a good way. Yeah, pretty much. Wow. Magia is like, if you don't want to use OpenSUSE because it's like all the Microsoft backing and all that stuff,

you just don't want to use Fedora and then there's Magia. Magia has an awesome software manager. It's just like by default they have some weird things that's going wrong with it. But the software manager is cool because you can get uh all all different repos and like there's options of like 40 to 50 repos that you can uh activate but by default they're all disabled which i find weird but um are they still using uh ur pmi do you know i'm pretty sure yeah yes okay so it's pretty cool that you can do that. You can easily go from beta packages to stable packages and stuff like that just by getting a checkbox.

But it's annoying to set up in the first place, but once you get it set up, it's very nice. Yeah, I do like that. Also, Magia has the closest thing to Yast that you can get outside of OpenSUSE. Right. Right. Is it still called drac config? Oh, the old mandrake. Yeah, the old mandrake.

Here's a good question. Does Magea have something similar to the OBS or the AUR? Not really. No. No. Their repos are kind of limited, too. It's the only thing. They have, technically, in a way,

they have an AUR with all the beta repositories that they support, but it's not like an official one place you can go to to get everything and have any kind of writing system. And to be honest, I kind of am really annoyed by Magia's packaging structure

because if you don't have Magia installed, you can't really tell what versions they have of anything because their packages is all listed out on mirrors, so you have to find a mirror that has the files. Yeah, forget that. Yeah, from a developer standpoint, that's a nightmare.

From a user standpoint, that's a nightmare. I want to know what I'm using. Or what I'm going to be using, rather. Well, once you have it installed, it's fine. You can easily just open the software manager, or whatever they call it. I forgot. But it's nice, but you have to install it first. So I'm looking to see what dependency they have by default.

I have to load up Mageia before I can find out. Interesting. Okay. Well, good info. And I think it's something that Matt and I will just – it's our nature to keep an eye on these kinds of things. So we'll watch it. And when there's something there that sort of tickles our fancy, we might jump in and do a review at some point. I don't think we have any immediate plans, but it's not something we would never do.

It's something we would consider when there's a moment and it seems like the right time. Corky writes in. He had an idea for a recap episode. I'm not so sure about that. But he said, I came across your episode from April 14, 2013, about Linux's use as a media creator. You criticized Linux's lack of tools for use for you, and it meant you needed to edit on a Hackintosh. Well, a year has gone by, and a lot has changed since then. Perhaps a recap episode would be a good idea.

You mentioned tools such as Telestream. I know quite a lot's changed, so an update. I can give them one right now. It still sucks. You know, honestly. And I do edit stuff on Linux now. Yeah, there's tools, right? There's tools.

And here's my new philosophy that I'm coming at this. I got a new angle for solving this problem, Matt. Okay. One is Rikai works with us now and he's now taken over the editing. So that's taken that off my plate. So that's really nice. And what I've been thinking about doing as time moves on is sort of building up a kind of a mobile setup and that would entail in itself, if I'm out

on the road and things like that, its own separate workflow anyways. And it seems like the time to switch over, you know, I talked a lot with Noah, who joined us for LinuxFest about this, maybe the time to switch over to Lightworks, for example, for editing, is the time when I'm already setting up a new workflow anyways. And now that I'm not handling the day-to-day show post-production, I can kind of zoom out a little bit and look at all of my options. And I can now afford to sort of kill time trying out different stuff and don't have to worry about interrupting the flow of every single show. And so I'm excited about the opportunity to kind of build a new workflow, and and tagging process that I have

done that at different points, but had sort of just abandoned my automation and was doing it all manually by hand. And the great thing about that system is it's taking that encoding process, and that's completely portable. That can run on a Mac if you install Brew, or that could run really straightforwardly on a Linux box. And so we can move it around depending on where horsepower is at. So some of the tools we're using now are open source just based on the new workflow that we have. The editing is still Final Cut.

The problem is it's just a good product. Despite what the internet will tell you, it actually is a good product, especially for online media production. So what I'm curious to see is, can Lightworks be comparable? And I won't know that until I've used it for a while and I've set up a new workflow. But that's my new goal is to kind of go down that path on a new setup.

The Bonobo is certainly powerful enough to do mobile editing. You know, I have two drives in that so I can read from one drive and write to another drive. I've been experimenting with it on the Ultra Pro also. It's completely usable. And you can even get two drives in the Ultra Pro. So there's a, you know, Noah was, when Noah was up here, he had it loaded on his Pixel. He had a Chromebook Pixel running Ubuntu that he had Lightworks on there, and he says it's usable. So I'm excited to give that a go.

We're already, you know, there's other things we use too, like the YouTube download script for certain clips. OSHA Audio is an open source audio editor. Audacity, those are already in our tool chain as well. So some components, more components since that episode in April 14th of 2013, definitely more components are open source. Everywhere it makes sense. But the editing, at least probably for the rest of this year and into next year, I don't think will change unless I really come across something pretty fantastic in the new setup.

And that would be nice to see. I mean, for myself personally, I've all but given up on OpenShot. Kdenlive is what's kept me sane just because it won't crash on me. You know, I love the fact that regardless of the environment I'm in, I can depend on it. It gives me the functionality I want and it's not going to take it up on me. So that's awesome. Audio, Audacity, youacity. So I mean,

I do use the tools and I actually was fighting with Audacity earlier this morning. So it's certainly not without its issues, but I do live in that world. I actually eat my own dog food on that. And I do think it has a ways to go. It really does. Yeah. And it's not just one particular program. I think that's what I want to underscore, right? It's not, you don't, it doesn't just take having an editor, right? Because then once you have a good editor, well, what is your codec situation? And once you have that solved, okay, now what's your motion graphics suite? Because Blender is great, but if you just want to be able to grab a motion template and edit it, well, you're out of luck. If you want to grab an After Effects template and edit it, you're out of luck.

So we don't have to have a – Because you're married to Blender templates at that point. Right. They are limited in comparison because, believe me, I've looked. They really are. Yeah, that that point. Right. They are limited in comparison because, believe me, I've looked. They really are. Yeah, that's it. That's the problem. It's not insurmountable, but it also means it's not just about perfecting and replacing one single application, but an entire suite of applications.

And that is more challenging, but you know what? Things change. Things change in big ways, and I would not say it will never be possible. I think it is possible, and I think we're working towards it right now. And we're just, you know, sometimes you just come up with new ways of doing your work, too, and then the different tools make more sense. All right, Matt, our last email today comes in from Captain Kirk, and he says, Chris and Matt, last week you touched on Ubuntu 1 closing and made some recommendations regarding replacing Ubuntu 1. When I learned Ubuntu 1 was to close, I got a DigitalOcean droplet and I put OwnCloud on it.

It is working fine for me, but since OwnCloud was not among your recommendations to replace Ubuntu 1, I've become concerned. Is there a reason I should not use OwnCloud? Is it flawed in some way that I don't know about? Best regards, Captain Kirk. Well, you know what? I will answer Kirk's question right then and there. should not use own cloud is it flawed in some way that i don't know about best regards captain kirk well you know what i will answer kirk's question right then and there but first this is a great opportunity to thank our sponsor this week and that is digital ocean here's what i want you to do right now go over to digitalocean.com use our brand new promo code unplugged may unplugged may will get you a ten dollar credit. Now, what is DigitalOcean?

DigitalOcean is simple cloud hosting dedicated to offering the most intuitive and easy way to spin up a cloud server. And that is a huge deal. That really changes the game. Users can create a cloud server in 55 seconds and pricing plans start about $5 per month, which will get you, not about, it is $5 per month, but that will get you 512 megabytes of RAM, a 20 gigabyte SSD, one CPU, and a terabyte of transfer. A terabyte of transfer, you guys. We had somebody write in and he said,

hey, you know, I don't really use like half my bandwidth on DigitalOcean, so when Ubuntu 14.04 came out, I just spun up a seed and I used a web transmission to help seed for a little while because I had about 500 gigs left on my cap. And that's a really cool way to just like, hey, I got a machine I can help out, you know, the Linux community a little bit. And the web transmission client makes that super easy. And DigitalOcean has data center locations in New York, San Francisco, Singapore, and Amsterdam. Their interface is so simple. The control panel is very intuitive.

And power users can replicate that control panel is very intuitive and power users can replicate that control panel on a larger scale with their straightforward API. It's really about the dashboard too. The dashboard takes all of these different technologies, things that you really used to have to be an expert to take advantage of, things that I used to be able to charge lots of money for my clients to implement for them are now so straightforward that it's kind of amazing. They have this droplet system where you can create image backups. You can move them around. You can snapshot before you make any big changes.

You can move them between data centers. They also now support private networking. There's lots of really great features all built around this really simple and intuitive dashboard. It gives you an HTML5 console so you can look at the machine as it boots up right at the BIOS level all the way up through post. Two-factor authentication if you want to be secure. It's so great. And there's more and more cool community applications that take advantage of their awesome API. So use our promo code UNPLUGGEDMAY when you check out. That's going to get you the $10 credit so you can try the $5 rig. I've been using the $5 rig for months and months now. When you

combine the power of KVM sitting on top of these SSD drives with tier one bandwidth and data center locations all over the world, the combination is absolutely insane, and you can try it for free for two months with that $5 rig, Unplugged May. And a big thank you to DigitalOcean for sponsoring Linux Unplugged.

It's a great service. Okay, so to answer his question about OwnCloud, I haven't brought up OwnCloud as a sync solution simply because every time I've implemented OwnCloud, I have ran into some serious performance issues. And I don't know if it's my setup. I don't want to presume that Own cloud inherently has some sort of problem. But I have had significant performance issues.

And I think it's because of the amount of files that I have deployed. And I have tried it on everything from a Raspberry Pi to a fairly high-end i5 home server. And I still have not been happy with the performance. That's not the case for everybody. And that's why it doesn't come front to my mind. But if it works for you, then have at it, man. That's awesome. OwnCloud's a great solution for you.

Well, and to elaborate, so when you're talking about performance, if I'm not mistaken, you're actually talking about the actual performance of like accessing a file or, you know, actually getting something to sync. Not merely bandwidth, but going deeper than that. Yeah, yeah. Okay. Like the UI and retrieving the directory listing and all those things for me has been a little slow. Now, a fader in the chat was pointing out you can change the database backend, and perhaps it was just my database backend selection. Now, I've done it a couple of times. I do not remember which database backend I used each time. I believe I varied it up with still limited success.

I don't know, Mumble Room, anybody in here who had similar experiences with own cloud or an opposite experience with own cloud that you want to chime in with? I just tried setting it up one time just on a server, just on a little server rig, and I don't know

if I did something wrong or what it was, but it was just a pain in my butt the whole time. It never worked right. I got it set up easy enough, but I didn't have a use for it because I have a ridiculously fast network at my house, and so I don't want to sync to the internet. Yeah.

Well, Derek Devon, what was your experience? Well, my experience was really weird. I think at least I got the impression that they're trying to make it like hard because they have like this duality of the community edition and their own edition for enterprise so i think they're kind of trying to get people to approach their services i i don't know it just like even you do the installation everything is correct and you just trickle everything and images start missing, things

not loading. That was my experience with it and I've retried and it appears that each version installation gets difficult. I hate image duplication. Doesn't Sysus Studio have a version of Unclouded that you can use?

Oh yeah, and it's not difficult to install in pretty much any distro. And, you know, I'm seeing comments in the chat room right now saying, hey, we're using it in production at my work. It's just fine. So I think it depends on maybe the database backend setting, how many files you have in there, you know, those kinds of things. Because to be honest, I always load mine up

with a ton of pictures and music. And, you know, if maybe you're more document-focused and more calendar-focused and contact-focused, it might just work just fine. And, you know, maybe if I didn't have to put 10,000 pictures in there and 20,000 songs, that could work a little better. Okay. So I want to talk about a topic with you guys. Now, Poby, don't take offense. This is just food for thought. But, and he actually doesn't, I'm going to say, let's look at this in the context of a post Ubuntu dominant desktop world, but it actually is just as applicable if Ubuntu is still the number

one Linux distro. So it actually works either way, but something that's been sort of gnawing away at the back of my head, uh, for the whole year, at least is a very awkward future we're going into in the next couple of years. It's come up a couple of times on the show, but this week I want to kind of take a different take on it. We have a few things that are coming up soon, sooner than I think we'll all realize. They'll just be here. Number one, let's stick on the Ubuntu thread.

Now, caveat, 14.04 will be out for five years, LTS. Okay, we're all establishing that. Now, the newer releases of Ubuntu based on Unity 8 are going to have, on the desktop side, brand new applications written in Qt that, just like always happens in the Linux community, some people will think are great and praise them for being innovative and the right direction. And some people will hate them. And they will declare that Ubuntu has ruined the desktop and they'll leave. And it'll probably be pretty, there'll probably be a lot of people

on both sides of that aisle. And I think the likely outcome of that would be people who are frustrated, they're not just all gonna jump to one distribution. They're not all gonna go in unison and say, all right, it's time to make OpenSUSE our new distro and all of a sudden OpenSUSE becomes the big fat lizard of the Linux community.

Or they're not all going to go to Fedora or Mangeo or Arch, right? It's not going to be any of those. Instead, you know, of anything, it's probably going to be a mass spreading out across all of the distributions. And it won't be like devastating for Ubuntu. It'll just be – I would foresee more folks moving out that are unhappy with some of the changes. But that brings with it that elephant in the room that a lot of developers who are targeting Linux desktop like to mention, and that is fragmentation.

And not only will we have this sort of user-based fragmentation over the next couple of years, but we will also have display server fragmentation. We'll have Wayland. We'll have Mir. And we'll also have distributions that just elect to stay on X11 for very legitimate reasons for a lot longer than we all expect. And so we will have fragmentation at that layer. there is a really interesting development battle brewing where GTK is starting to maybe not become the preferred targeted framework

for new desktop Linux applications. A lot of people are talking about Qt. It offers a lot of advantages. It's great with mobile platforms. It's a good technology in general. Obviously, the canonical folks are switching over to it. So there's a fragmentation there. And then on top of that,

you've got just a general desktop environment fragmentation where some desktop applications handle notifications in one way. Another desktop environment handles it another way. Some draw borders. Some don't draw borders, right? It's all over the map. And this is just going to be in the next few years. And I'm trying to wrap my brain around this. And I'm thinking at the same time, at the hardware side of things, it sounds like calling it the cloud, but you've got this Internet of Things, right? All of these devices, from Philips light bulbs to the toaster to the TV to your garage door opener to your front door locks, there's going to be devices out there running Linux, talking to each other, but they're not going to be compatible Linux boxes. You know, they're not going to be able to run the same applications. You're not going to be able to write once and run it anywhere. They're not going to have the same processors. It's another type of

fragmentation. Not, but not one I'm particularly worried about. You know, the Linux core itself, Linux core itself is, it's pretty solid. You know, you know that you're going to have glibc. You know, you're probably going to have gstreamer or qt. You know, you're probably going to have cups for printing. You know, you're probably to have Glibc. You know you're probably going to have GStreamer or Qt. You know you're probably going to have Cups for printing. You know you're probably going to have WebKit if you need to draw an HTML box. So those things are kind of standardized. The core, the kernel, systemd, all that kind of stuff is pretty standardized. So I'm not so worried there, but I am pretty worried about the desktop because I can't see where this is going where end users don't lose out and where developers don't get so frustrated they just put

their hands up and say, screw it, I'm going to write an application for the Mac because I know they'll spend money and I only have to write it for one operating system. And let's be honest, all this is happening when the dominant desktop Linux or the dominant commercial desktop platforms are getting pretty boring. You know, they're not changing very much. And a lot of developers like that. They like that particular brand of boring because it's easy for them to make money on. That's definitely not a brand Linux has to offer. So, I thought maybe we could discuss

with the Mumble Room, with our virtual lug, how we'd like to see this play out. What if we could wave a magic wand and push the future of desktop Linux in a certain direction to prevent some of these problems, to save ourselves this obvious hassle that we are all about to encounter? I mean, it's like we're all driving down this freeway going 80 miles per hour, and we know that there's a bunch of trouble up ahead.

The bridge is out, but we're still going. And we just hope that when we get to the bridge, we'll just jump that gap and nobody will notice. Ah, the bridge is out. Nobody notices. But if we could somehow build that bridge before we got there, what would we make that look like? And that's one of the things I thought maybe we could toss around today.

I have a certain thing. Go ahead, Ick. Yeah, I think we've got a temptation to say a teapot. So many people I talked to at LinuxFest Northwest about the direction Unity is going and how it's going to be mostly Qt-based seemed really happy about that. Yeah.

They seemed like it'd be more solid. That was my takeaway, too. Yeah, easier to program on. That guy keeps seeing Qt programs on Windows, too. So, you know, it's cross-platform. You can go anywhere with Qt. So it's easier for purposes of making cross-platform applications. So you're saying in your Linux desktop of the future,

Qt has sort of emerged as the dominant framework that applications are developed with. I think so because it's more cross cross platform therefore more people can program with it and it can be port easily easily portable and i want to see more of that all right i'll go to heavens next and then popey's up so go ahead heavens what were you gonna say well i would like the next iteration of what I wanted Chrome OS to be. Right now, Google kind of screwed up what I imagined Chrome OS to actually be. Qt and GTK both have embedded JavaScript interpreters, or ECMAScript interpreters,

which actually make them do their beautiful magic. QML and GJC, the JavaScript inside of GTK, which makes it absolutely Mac-ish pretty. So I think because JavaScript was invented to control UI elements as its fundamental, that's what its design goal was. It's cross-platform. It can go to Mac, Linux, Windows, or run on your car. It doesn't have to be Windows or C++ or compiled in order to run on that platform. No, it's perfect for this Internet of Things future. Now, before we go to Popey, Tyler, you wanted to kind of dovetail on something that Ick was just saying. Oh, yeah. I wanted to also say part of our discussion when we were at LinuxFest Northwest

also involved talking about Mirror versus Wayland and the direction I'd like to see the video card companies go is develop the kernel driver as open source like AMD wants to move towards or Intel. So that kind of reduces some of the compatibility issues

that we would have with proprietary drivers. Now, Poby, I want to throw something in your face. So let's say here, when I'm saying the future desktop, what I'm saying is there's some sort of overall consensus in the development community that when we're writing a desktop application for Linux, these are the things we target, and it's sort of the whole hive mind in some magical future

has sort of bought off on this. Pobi, what would you like to see in that scenario? So I'm going to take a step back and not answer that because I think your argument is flawed for two reasons. First is even if Unity is pure Qt, that doesn't stop applications being GTK or some other toolkit. So whilst it's great that Unity is moving towards Qt, that's just one of the platforms Unity has been implemented on.

It's been re-implemented multiple times, and we've still been able to run Qt, GTK, and other toolkit applications on top of that. We don't stop that happening. No, no, I wouldn't say so. So implying that in six months or a year or wherever the full Unity mirror lands on the desktop, implying that that means the world is cute and forget GTK

or forget anything else. You can't run those things on Ubuntu. Therefore, it's gone. What's the post-apocalyptic world going to look like is a somewhat forward argument. I think you're a little sensitive because what I'm saying is what the frustration is in the developers that I've heard from is that there's too many choices, right?

And so what I'm trying to come to is something that doesn't look fragmented, something that doesn't look out of place. I'm trying to come to some sort of consensus where we can have a desktop operating system that's filled with choice. Because one of the good things, honestly, is like each distro solves a particular niche, and they're good for those areas. like each distro solves a particular niche and they're good for those areas. So how do we come to something that is sort of each in each area looks like it belongs,

look like it's in place. And what I'm suggesting is maybe if we could sort of come to a consensus to perhaps, let's just say, you know, some sort of utopian future where the cute transitions go so well that more and more people jump on board wouldn't that actually be a good thing i don't think that's dystopia i think that's good right so i'm not saying there's a dystopia but my my reading was that the implication was you know that there's going to be this mass exodus from ubuntu when when this happens maybe i've read that wrong oh what i'm too oversensitive well yeah no what i'm saying there is that there's going to be whenever you change applications like, there's just going to be people who will make a big fuss, right?

They make a big fuss. Change equals bad. So that'll add fragmentation. What I'm saying there is that's going to add fragmentation in the user base, what they're using. So I would say the best thing about this is you can take the comments that Ike made and the comments that Heaven's made, which is basically QML and Q are great. JavaScript is great as well. We've implemented our toolkit in QML, but we're also implementing it as HTML5. So if you want to bring your JavaScript app across,

it will look the same on Ubuntu, whether it's a QML app or an HTML5 app. So you can get that same look and feel across different toolkits. That is possible. You know, that was sort of the pitch. Well, in a sense, that was the pitch too that the Firefox OS guy made at Linux Fest is he said, you know,

the developer can make a phone app and then, you know, they just redesign the layout or maybe it automatically relays out and it's a web app too. And you've written the application once and it runs in both spots and it saves the developer having to duplicate work and if you could you know even with minor minor tweaks to make it integrate properly take one application and move it across ubuntu touch ubuntu desktop firefox os and anybody who has a browser capable

of rendering html5 that seems like a pretty compelling incentive for developers to target that, even if one of the individual platforms doesn't have that much traction. Because in whole, there's a considerable amount of traction and a reduced amount of development time, which that's why I really do think Qt has a strong future, not just on the Linux desktop, but everywhere. And maybe it's Qt and a combination of HTML5 apps. Are we worried about that, though? Is that a good thing? I mean, qt and a combination of html5 apps are we worried about that though is that a good thing i mean do we want a bunch of web apps um now my response as far as

cute goes i'm really hoping that if ubuntu is heading towards cute popular applications like let's say firefox or chrome will start pulling their theming from cute instead of gtk chrome has stopped pulling from gtk. They're doing their own thing. Aurora, right? Do you guys have a lot of problem with the fragmentation? I'm honestly sitting on a Windows computer running a GTK app and a Qt app side by side with my

Windows app. I don't care. This is exactly where I wanted to go because I think if you look at actual in-production I don't know of a Linux desktop that at least is on x86 that doesn't have Chrome or Firefox like Thunderbird, right? I'm just naming applications that are already cross-platform VLC. We all have these even though there's supposedly so much fragmentation already that developers could never target Linux. But yet I've been using applications across multiple different Linux desktops for years and years and years, and they're the same applications.

I get them at the same time for the most part. I haven't had any problems. So let me ask you this. Is fragmentation just a bunch of crap? Is it hype? Is it just like something that developers use as an excuse to be lazy? I don't understand. Alan, I'd say as far as Moshpots goes for wanting a beautiful desktop, GTK apps, and I'm a KDE user, do not fit into kde that well at all unless you are

limiting yourself to like a certain set of themes but it works it works but it doesn't look good right yeah that and that bugs me well well that doesn't bother me and it doesn't bother most normals i don't think it does either yeah i think most users mugg... The guy who writes programs, it would probably be an easier idea if you have a layer to connect all the toolkits and stuff so that they only use one set of interfaces. Well, but you know who does care? Actually, our developers themselves. They don't want to make...

PC Wiz, close your mic. They don't want to make something that they think looks ugly, right? They're not going to put their time and effort into something that... Well, that depends on which kind of developer you're talking about, because, you know, some types are like, I don't care what it looks like. It has this really odd design, because that's how I like it. Right, and then you look at...

I'd like to add on to it. Normals might care about it, because, you know, you have this beautiful desktop, and then you have this GTK app that might look like it came from Windows 98. I mean, I think Normals might care when, like when like for example the client-side decoration thing where you've got two clothes bars two title bars i mean like that that's just confusing yeah wow you have a real b in your body about client-side decorations that's like the fourth episode of the show second episode but fourth mention uh yes second episode fourth, fourth mention. I agree. No, it's because it is the beginning of this particular type of this brand of fragmentation

that I'm talking about and where it's happening right now. And as a gnome user, you know what, I'm of two minds of it. So I'm trying to wrap my brain around it because I picture myself showing this to my mom, who has been a graphic designer for almost all of her professional life. And that means she's used Mac since the 80s to right now. And I try to put her down in front of an Ubuntu box because she wanted a new computer and I was showing her GIMP. And I could not even begin to describe to you how badly that went. And it's things like that that she just would look at and she would think, think is this why it's free that's literally what she'd say ouch right now for the client side decorations

i think we're just in a phase it is something that the desktop environments have to solve it's not yeah on the it's not the responsibility of other things right now the client side decorations is desktop specific. Yeah. I guess I want to just kind of close out my thoughts on the perfect desktop. I would love to see more targeting of things that work across all desktops, not just Linux desktops but all of them. And I'm willing to

give this transition to more web apps a proper go because I feel like in the last couple of years we've actually started to see web applications that are impressive. And I don't know if you guys have played around with it, but if you install Chrome and then there's the Chrome apps, I think all of the Lynx desktops, it will add those Chrome apps as individual launcher icons like in your menu

or whatever activity screen, whatever you're looking at. And I launched the Plex one, and it looks like I'm running a standalone Plex application, and it's just a Chrome application. And it actually works. Go ahead. The only thing I don't like about the Chrome applications is that they use one

session for all of them. So if you have multiples of the same web app you want, you can't log in separately. You have to log in just once, and then it makes a duplication. But if you look at Midori, they have an applications thing very similar to the way Chrome does it, but every single application being made has its own individual session. What if I have a Firefox version of that?

You can do that with profiles. Prism. Prism was the application. Yeah, bad name now. Dave, you were going to say maybe it's more about how the applications communicate with each other. Yeah, I mean, think about how these different devices that we use actually connect to each other. Right. I mean, I have two phones, two laptops, and a desktop in my home, and they're all on the same Wi-Fi network.

So I don't think at the end of the day it matters what graphical toolkit is actually running or what the UI is. If they're all accessing the same file and writing to and from the same file, then they'll work together. For example, the Plex app. The Plex app is still reading from the same server, whether it's on your phone, whether it's on your laptop, or on the cloud.

So what I'm saying is I think we need a standardized backend network interface for phone applications and desktop applications and HTML cloud applications to talk to each other. We need a backend connection, like say something like Unix sockets, or a standardized network interface where these things can talk to each other and they can work together with each other and just exchange data regardless of what the individual UI is. Right, as long as it can be implemented safely. Now, Daredevil, you were going to make a point about how you just want to, like,

get into a nice pair of comfy toolkit jeans and just walk around and feel good, right? Did I make your point for you or did you want to add to that? No, yeah. I mean, if I'm familiar with the toolkit, I probably know how to leverage that toolkit to look just exactly how I want. And I'll do my application to look the best I can, and that's going to be my effort. Now, if a distribution later on that uses a different toolkit wants to pick it up, well, they better support my toolkit. And in the end, it will benefit every single user because there's more developers that just do like I do. They focus on what they know because it's easier. I mean, if you're going to jump on toolkit to toolkit,

sometimes you will gain a benefit. Like, for example, you go to elementary OS, you will have a very valid GTK-focused environment. You will get runtime performance consistency with their teaming because they have these toolkits, these widgets that you can use. No, I mean, it works, but you have to have the elementary OS photo viewer and the elementary OS music application and the elementary OS this and the elementary OS that. And to be honest with you, they're not a multimillion-dollar company that can hire dozens and dozens of developers to work on each individual application. At best, they get a mild feature improvement after each release.

And this is my issue here, is when you reinvent these applications, when you recreate them, there is a massive technical debt that you undertake to make them as featureful as all of the other applications that have been around for potentially 10 years. But that's not what I mean. I'm not saying that the elementary way is to go. I'm just saying that if i selected that toolkit in

those tools when i'm going there it's going to make my life easier there and because they select this path it's going to be very clear that if i do gdkr will get a few things for free such as the widgets when they change it will change in my application as well same way that you do in mechaosx with cocoaa but I really don't think this is a problem just because I mean developers have been using their toolkits and when the toolkit is not there some of them will actually compile the toolkit there yeah I think the best is to the developers to focus on getting a good experience on their application regardless of the toolkit and then worry about about, okay, can I port?

Unless it's like a mobile platform where the resources are constrained and then you have to deal with just one toolkit. I guess I just look at, I'm sitting back and I'm looking at the successful, like really successful Linux-based platforms. And two that come to mind are Android and Chrome OS. And both of those involve a limited set of choices for developers. Both of those involve a very narrow scope of what you can and cannot do and how you do those things and then where you publish those things. They all have a very set of strict lanes that you have to drive down

and rules you have to follow. And the desktop is nothing, nothing like that. It is a wild west and a cornucopia of choice, but that's not necessarily a good thing. Now, that said, I mean, I don't want to be all Mr. Anti-Forking and whatnot because Jake Edge over at LWN.net had a good piece about Linux and the Internet of Things. He was talking about Tim Bird, who was giving a keynote

at the Embedded Linux Conference. And this piece was at the end of April of this year. And he brought up a point that Linus made about forking the kernel. But Linus is actually, this was in an Ask Linus column from years ago, when somebody asked Linus, he said, do you get all upset when people fork the kernel and sort of start reinventing the wheel? And Linus says, actually, he thinks that when folks targeting a new market where Linux does not have a presence, it actually makes a lot of sense to fork. He says, just as a new market, and we need a new base camp to attack from it. So when they fork, they're essentially creating a new base camp in that new market.

And, you know, you look at how Android started. Android was originally a fork, and then later they've been folding much of the Android changes back into the mainline Linux kernel after years. And so he says, and Linus, I have to agree with, he says the key to getting in there is to have some place to start from. And sometimes fragmentation at the beginning

looks like the end of the world or it looks like it's going to be a huge mess. But because the code is open and the changes can be readopted if those upstream want them, it sometimes actually works out for the better. And, you know, in this case, the Linux kernel has gotten a lot better because of these kinds of forkings. And I think that's telling because if you can do it at the kernel level, you can certainly do it at the file manager level. If you can do it at the kernel level, you can certainly do it at the file manager level. Or you can, I mean, nobody, who says somebody doesn't take the new Qt file manager for Ubuntu and it becomes three, four years down the road, it becomes the default KDE file manager. They fork it and they make a new dolphin based around it or something like that.

I mean, you just never know what could happen. So maybe it won't be that bad. Well, you have to bait the hook to suit the fish, right? I mean, that's what he was saying. Yeah, that's very good. Yeah, that's very true uh yeah i think the market will just eventually sort it out just like it's always done like the livings community is pretty resilient when it comes to that stuff so like eventually it might not happen right away but over time things will get sorted out so right well

as far as the client-side decorations those just need to go and burn in a pit somewhere. Wow, that's a little extreme. Let's play it quickly. With my app, actually what we're doing is separating the UI and the core. So the core is being written completely without a UI, and then we can put any UI on top of it. So it could be Qt, it could be GTK or whatever. And the reason we're doing this is because it allows us to make one core that can support mobile and desktop.

So we're working on an Android version, and that core will work also on the desktop with whatever toolkit we pick. Yeah, that's what the KDE guys were doing. I saw this video, OS something, OS Connect, where this is KDE's vision as far as I understand it. The basic toolkit has to be the same. KDE Libs hasn't been upgraded from 4.9 because they're kind of halting it.

The UI, they believe, has to be different for tablets, different for phones, and different for desktops. And this is exactly what is going on. You know, fragmentation is good if it's done properly. We have to, you know, bait the hook to suit the fish. A phone UI is different. It needs a different implementation from the ground up, from the kernel to the GUI toolkit to the actual interface. Right. The same with the desktop.

Right. So, you know, I think that's what we're ignoring here, that we can't have the same UI across the board. And we have to both embrace the fragmentation and find a way to connect the dots, basically. Yeah. Connect the different, yeah. I also think, you know,

earlier the inter-app communication was brought up. I think the difference this time with this type of future we're going into is actual, reliable, and performant back-end processing is literally today's reality. You know, you could throw something up on a DigitalOcean droplet, on a Rackspace machine,

on Windows Azure, and you can have essentially as much processing power as you need. So if you need to have different applications across a phone and a desktop communicate, we have a solution for that now. You know, you look at all these, the augmentation you can add to help smooth out the rough edges between these different form factors and different types of devices that need to communicate with each other is much more of a reality now. It doesn't have to be device to device anymore. There can be a middle processor that does that logic work, that does the storing, does the state information. And that's going to be extremely powerful. And I think, I'm hopeful, that's going to help smooth some of this stuff out too.

A backend. Yeah, exactly. Which we're seeing different iterations of that now., that's going to help smooth some of this stuff out too. A back end. Yeah, exactly. Which we're seeing different iterations of that now. It's just going to become more prevalent. You know, imagine a Linux distribution that maybe has, instead of, now I don't know if people would like this, but instead of each individual installation going out and doing a package refresh and pulling down all the repo information,

maybe that distribution just has a single server that just, when you do an update, your Linux box just connects to that server and says, hey, what's the latest package? Okay, thanks. And it's like this, they're doing the backend refreshing of all the repos. They're doing the backend caching, even with your own personal repos, and they're just delivering you a single answer across all your devices. It's a weird example, but you could see how back-end processing could be applied at all different levels of these things.

It could really change things up, and that's just something we can't really wrap our brains around right now because we're just starting to see the beginnings of that, and that's going to be a major game changer down the road. So that's my thoughts on it. Cloud computing isn't a buzzword anymore. It is a reality. Yeah, we just don't have to call it that. Let's just not call it that. It's still a buzzword anymore it is a reality we just don't have to call it that

let's just not call it that it's still a buzzword you can't describe to me what it actually means but computing there you go I did it for you let's just bring it down to the earth and make a fog yeah or how I describe it

it's anytime the system is opaque and you can't tell what's going on inside of it. Right, anytime it's magic. Before we go, I wanted to give a special shout-out to IrrationalNumber in the Linux Action Show subreddit. He's been doing an Ubuntu challenge, and he's really going at this. This is a really interesting journey he's taking. It came up, actually, during an IRC conversation from Linux Unplugged a few weeks ago. So with 14.04, he made the switch over as a longtime Arch user to using Ubuntu.

He did a great post that outlined his goal, and then he's done a great write-up on his first week and what he ran into and what kind of struggles he had. And how he just posted two hours ago, week two, another really good post. You know what? This is the kind of thing you start reading it. It's actually very interesting. And then there's some good conversations that are starting up around these posts, too. So I will link to these in the show notes if you guys want to read these. I thought they were pretty interesting to follow. And then one last thing I wanted to cover before we run was a question that came in today to our subreddit. It says, my children's school district, this is a gray wolf night,

my children's school district is looking at testing various options for replacing XP. Well, probably a lot of people are. Right now, they mostly have one main CPU, three to four workstations per room. They're trying Chromebooks for teachers and Windows 8. I would like to provide them with another solution. The school district doesn't have a lot of money, so I'm hoping to get a solution to help teachers and students. I was hoping to reach out to see if anyone

out there has ideas for a desktop window manager and programs. I like Sugar Idea, but I have a hard time with it, so I'm not ready to make that work. I was able to get IT department to try out OpenOffice, but I need more solid programs, especially for teachers. Thanks to anyone who has some suggestions. I will tell you what I've seen work. I don't know if you're going to like my answer.

Linux Terminal Services Project, Google Docs. The thing about the Google Docs is, and this is what you can't get around, and you can make the same argument with OpenOffice, but then it requires parents at home to install software on the computer. The thing about the Google Docs is they just need to have the web browser. And Google just launched

Google Classroom, where schools are going to be able to do assignments and integrate that with Google Docs, and students will be able to check in. And Google just launched Google Classroom, where schools are going to be able to do assignments and integrate that with Google Docs, and students will be able to check in. And again, all you have to do is just have the browser. And Google just announced that they also will not be doing tracking or ad sales against educational accounts. So the creepy factor has been reduced, too. I don't know what that means exactly, but they're supposedly not going to be scanning the content of children's Google Drives and won't be displaying ads against that. So that seems good.

So that's one solution. Linux terminal servers work really well as long as you have even semi-decent connectivity. It works even over Wi-Fi, one box, and then you just have a multitude of front-end clients. You will be amazed at how smooth it works. I don't know, Matt. Do you have any other suggestions for a cheap way to get boxes in a school? Boy, I think you really nailed it. I know here at a local community college, Google Docs and Moodle, those are like the two.

They have great success with that, and it actually works really well. I feel a little dirty recommending Google Docs because it's not A, open source, and B, there is going to always be some kind of tracking. But at the end of the day, it's kind of like it you just can't get past that they don't even have to run linux at home right they could be on windows they could be on mac and they could still work on those same school docs that's a pretty compelling use case for for students in schools so that's well yeah i mean because my nephew uses a windows machine at the school then comes home and we only use linux here so he has a linux uh linux a Linux laptop in his room, and he just logs into that, and it doesn't care.

It just works. Yeah, and Chad was bringing it up too. Edubuntu is a great distro to run. Sure. I've also done it with Fedora. There are specific spins of distros that are targeted for Linux terminal services. So try it out. And Edubuntu is a great start because it's really focused

at the educational market. Okay, Matt, well guess what? That brings us to the end of this week's Linux Unplugged. How about that? Did you know? Were you aware, Matt, this brand new, we just launched this, did you know you could watch this show live? How about that? You can! I know, it took a little

working, but I found this company called Scale Engine and they hooked us up. Yeah. So here, you could, hey, maybe you guys didn't know it. So come over at jblive.tv. We start at 2 p.m. Pacific, or go to jupiterbroadcasting.com slash calendar to get in your local time zone. Oh, and go over to teespring.com slash CR100 for the Coder Radio 100 shirt. All right, Matt, I'll see you on Sunday, okay? See you Sunday.

All right, everyone, have a great week. We'll see you right back here next Tuesday. Thank you. you