LINUX Unplugged - Episode 30: Talkin' Tox | LUP 30

Episode Date: March 5, 2014

Two developers from the TOX project, an open source secure Skype killer join us to discuss their new project, the future, and how they hope to become your new messaging system. Plus getting more battery life out of a Linux laptop, the Steam problem, and your feedback.

This is Linux Unplugged, your weekly Linux talk show that knows this call is being monitored. My name is Chris. My name is Matt. Hey there, Matt. We've got a great show coming up this week. So last week in episode 29, we talked about ourselves a lot. We talked about the new How To Linux show. And which, oh my gosh. We got like 40 to 50 emails just on that topic alone.

It's blowing my mind. And everybody has a great idea. So many really good ones. But I just thought like we couldn't go through that again because this week I want to help somebody troubleshoot some battery life problems with their laptop. They've been running Arch for a long time. They switched to OpenSUSE and all of a sudden,

way better battery life. And they're trying to figure out why that might be, so we'll kick that topic around with our mumble room. And I know you've got some battery tips that you might be able to pass along. And later in the show, we're going to chat with some of the developers from the Tox project. Now, Tox is a new up-and-coming open source project to essentially replace Skype, to have a secure peer-to-peer Skype replacement.

And there's a lot of secure messaging programs that are coming online in 2014, obviously as a reaction to the leaks from Edward Snowden. And I think it's going to be really fascinating to watch these as they develop. And particularly because of my focus on the desktop, I've been drawn to ones that are starting with a desktop application. A lot of the ones you're seeing kind of come out of the response to Edward Snowden are mobile applications, which is fine. I want to see a mobile app too, but I'd also really like to have a desktop component.

And Talks is starting there. So we're going to talk with them in a little bit and ask some questions that our audience has submitted in. And then we've got some mail sack we're going to get to towards the end of the show, too. So big show today, Matt. Big show. It might be a really kind of tight show depending on how long. It could be like a big show in terms of what we talk about, but a shorter show in terms of length. You never know.

I shouldn't say that, though. It probably would curse us. So let's start with the topic that I wanted to bounce around. This is a problem that Guggero wrote in about. He says, OpenSUSE has better battery life than Arch on the same rig. Can you guys help me figure this out? And he starts out with, you know, he says, hey, guys, I'm in love with my Arch Linux install.

Well, actually, it was Integros. Have you made the switch with you guys during the Arch challenge? But I recently gave OpenSUSE a spin, and while it has some things I really like, I'm ready to switch back to my Arch install. However, I'm finding it very difficult for one reason. Battery life. Even with similar power management rules running KDE on both distros, the battery life on OpenSUSE has been insanely better.

The trade seems to be that OpenSUSE has been insanely better. The trade seems to be that OpenSUSE is a lot slower. I can't figure out what the actual difference really is, though. I've been digging in the Arch wiki for suggestions, but with school in full swing, I've been unable to find time to figure out a solution, and I've grown very comfortable with not having to hunt for outlets everywhere I go. As I've said, aside from auto-dimming the screen and auto-suspend settings, I can't really figure out where OpenSUSE is getting the dramatically better battery life, but Arch has proven to make my lap be very power-thirsty over multiple installations, even with an Arch proper installation.

Your loyal JB viewer, Don General. Well, so my experience between the two distributions is it seems like OpenSUSE definitely has some moderate configuration stuff done for you. I wouldn't go so far as to say they did laptop mode tools or something like that, but they definitely have something going on. On the Arch side of the fence, laptop mode tools will do you no good

because it's a completely different kind of setup, and out of the box, it's not designed to give you any power savings. It's designed to just run. If you go to the Arch wiki and go to the power savings portion of it, all the tips you need for a laptop are in there. And if you enable every single thing, including down from UDEV all the way up to the kernel modules and whatnot, and actually look at all those different things, enable them, write back mode, the whole package, you will indeed see significant improvement. back mode the whole package you will indeed see significant improvement so i i've this this concept of battery life being different between distros is interesting to me because i don't aside from driver versions and kernel versions it seems like things should be pretty normalized and i've always found the newer kernels recently have provided better battery life early three

kernels had uh power regressions but i've seen like i've always noticed that when i've upgraded my kernel to the latest kernel i've usually seen an improvement in in some component that results in better results in some better battery life like uh so maybe it's better support for the intel graphics or maybe it's a better wi-fi driver and that results in less broadcast usage or whatever it is it always seems like the newer kernel has always been what solves it i've never really like specifically tested this thing, like put Ubuntu on a machine, put SUSE on a machine, and then put Arch on a machine and timed the difference. So basically my experiences are based on if I install PowerTop and I run it on a default Arch installation, you're going to be – and you go over to tunables. It's just bad up and down.

That's just what you're going to see, bad, bad, bad, bad, bad. What you want to see is lots of good. And if you want to do that, that's why I would actually recommend the power saving section for that. Because now on OpenSUSE, I recall there was some bad, some good. Now I noticed there was more good enabled in PowerTop than there was. Oh, yeah. PowerTop is a great tool. Maybe you should run PowerTop on both installations. So, you guys have any ideas on what he could check to see why OpenSUSE is getting such better battery life than Arch? Yes, try CPU power

because I know CPU power behaves pretty good with systemd. Yeah. You think maybe there's like a moderator set? I'm pretty sure because I know in the kernel, Arch has performance mode enabled by default too. That's right.

Interesting. Yeah, so if you modipro the on-demand driver for CPU power, whatever it's called, it'll be pretty good after that. Yeah, CPU freq's another thing to check to see if maybe... Actually, CPU freq is not so good because I think that's been

depreciated. Oh, is CPU freq not just a tool to read what your current frequency is? Because I'm wondering if maybe there is more aggressive frequency scaling maybe under OpenSUSE. Probably. Yeah, there's a number. I've never really driven too deep into OpenSUSE specifically, but I do know that they have some stuff enabled by default. I'd like to play with this some more, I think. I think I would like to try this out.

I think this would be an interesting – this almost could be an interesting roundup. Is there the ultimate battery-saving distro? Because, you know, on Sunday I'm going to review this C720 Chromebook here, and I can't even – Oh, you will not want to do it without using the power saving portion. No, I can't. You know what's crazy? It would be painful to use otherwise. No, it's not – I mean, so I think I could get maybe another hour.

I don't know. I should try it, to be honest. I will try it before the review. I mean, on my netbook, I can squeeze like an hour and a half to two hours difference. Really? Oh, yeah. I mean, it's a big deal. But you've got to go through every single thing on that power saver there.

Okay. So CPU Freq was renamed to CPU Power. And that's just one piece. So I'm getting like – I don't know. I'm going to time it for the actual review right now. But I think I'm getting like seven hours of battery life with the C720 running Arch. So I'm going to install, I'll go through that Arch wiki that you mentioned, and I'll dig through that and see if I had to do some of those tweaks, if I can even bring that up a little more. And then I'll time it.

I'll try to get like some use case timing done in time for last on sunday but i i find this whole concept maybe once i get this review done maybe i should load open susa on there and see if there's a big difference and if there is then i might try it with other distros too uh tlp is also cool as uh someone in the chat room actually just mentioned just now um that that's definitely good i i'm really really careful of it because it's one of those things to where you can just end up breaking something and then you have to like go back and figure out what you broke i also like i also like the idea of checking uh like uh iotop and using iotop to see what's hitting your disc because it could be possible like by default maybe the kde maybe some search indexer like the the kde search indexer is i'm sorry the plasma desktop Search Indexer is set to a more moderate aggressive level.

Who knows, right? I mean, I'm just throwing things out there, but maybe. So IOTOP would reveal that if something was hitting your disk a lot. True. That's very true. If you have battery experiences and you want to send them in, please do. But include, if you're using an SSD, your generation of processor and your GPU and relevant like screen size info and things like that.

So that way we can take that into consideration when we're talking about this stuff. So we talked something, all, really, I mean, 99.9% of our feedback was about the how-to show. It was so many ideas that it's kind of blowing my mind and it's freaking me out a little bit because there's a lot of ideas that completely are in contradiction of what the next email says. And so it's gotten me a little worried that so many people are kind of disagreeing with each other without realizing it.

Like, well, I get an idea where somebody will say, take a topic and really spend some time with it and, and make three or four episodes at that single topic. And then the next email will say only one episode, make it as long as you need to get to one topic. Don't stretch it out.

I won't watch it if you do like all this different or one that my favorite which would be awesome if we had a staff of 40 my favorite suggestion was um do an intro video that sets up the problem like here on how to linux today we're going to show you how to do x and then kind of talk about a little bit then that stops and that's like a five minute video. And then have like 20 or 30 sub videos that then the person would click on. That would be how to do it in Ubuntu, how to do it in Arch,

how to do it in OpenSUSE, how to do it in CentOS, how to do it in Debian, how to do it, et cetera, et cetera, et cetera. And I could see how that could be really handy for the Linux community.

But yeah, it has a lot of, can you imagine from the production standpoint? From a production, it'd be a nightmare. But I mean, it's one of those things where it's like from an archival point of view, it's awesome. From a show point of view, it'd be a nightmare. Yeah, yeah.

So I think what I can tell you is it would be really interesting to see if maybe the community could come in. And a few people suggested this. Maybe the community could come in and provide the documentation people suggested this. Maybe the community could come in and provide the documentation for the distro they love that we didn't cover. And so that's got my wheels turning

on how we could build an interactive community website that would provide a written documentation version of what we talked about. We would have one official probably that would match the show. And then maybe we could have multiple versions for different distributions that the community could tribute contribute and maybe we'd have to have like three or four or five or six or seven originally you know hand selected contributors at first to

make sure it was good and then kind of grow it from there but it's it's something you know it's a lot of feedback we're taking in um i'm going to try to um send out the uh the official like hey we'd like you to be our producer email this week. So we're going to have that finally done, I hope. And between that and the new studio, we got a lot on our plate right now. But I feel like there's a lot of people out there that want this. But the other topic we talked about last week besides that that did get one response is the one here from Alex. And that was about Outlook's inertia. And we talked a little bit about the exchange problem that Linux switchers face. He says, thank you for the wonderful shows.

I got it started with Jupyter Broadcasting back in the stoked days. Recently came back after finding BSD Now and then TechSnap and Last and Unplugged and CodeRadio, etc., etc. He said, listening to Linux Unplugged 29, there was a comment made about needing exchange or MAPI compatibility, or essentially Outlook. As someone who has been trying to migrate the small business he supports over to Linux, the big issue I see is that open source groupware packages, both server and client side, really do not measure up to the standard that is Exchange and Outlook. For all of its warts, Microsoft has gotten groupware mostly right. When you have requirements that mail be hosted internally and you need groupware and mobile sync, the choice

currently is Exchange. Zimbra keeps getting passed around like a hot potato. I have used most of my two free months of DigitalOcean trying out Zimbra with OpenLDAP. I've got it working and others don't seem to be, but others don't seem to enjoy all the functionality yet. Any suggestions would be appreciated. Alex. Oh, man. So, you know, what do you like? Do you have like a calendar contact sync system you use? Are you using like Google, Matt? What are you using? Right now, I use a little bit of Google. That's pretty much where I've fallen into at this point. For one of my gigs, I have to live in a Microsoft space. And so we use Exchange. And anyone that tells me it's flawless has not spent much time with Exchange, in my opinion. It's painful. And that being said,

it is less painful than a lot of the open source groupware solutions, as was pointed out. And so I agree with that. I mean, that's actually accurate. But none of them are really that good. Google, honestly, while it's missing some functionality that I would like to see, I'd say it's probably the least annoying. They're all annoying, but I'd say it's the least annoying experience that I've ever dealt with. The nice thing about Google is you don't have to deal with the back end if you don't want to.

Right. Yeah. As long as you understand that your stuff's open to whatever. But it is anyway. Yeah. Gosh, that's the part. I mean, I'm not a huge fan of that part, to be honest. But I think, you know, I like Zimbra. I just think what the problem is, is exchange, you sort of create a workflow around Outlook and Exchange.

And when you move to one of these other options like Zimbra or any of the other ones that are out there, it's a different workflow. And if you started on that, it would be fine. But because you've trained, you know, those mental pathways in your brain are laid down for the way that it's sort of like why I sometimes have a little bit easier time using gnome is because honestly, I've, I, I learned a lot of things the gnome way. And so this just way in my mind, like that's just

the way the neural pathways are is like the, the, the electrons fire faster down those paths. And it's the same thing for outlook. If you've learned how the whole Outlook calendar system works, and I know people out there like depend on the rich text formatting of Outlook and all this kind of junk that Outlook does. But to them, it's an essential part of the job. And sometimes these replacements do it differently. Sometimes they don't offer something at all. So a lot of times people say they don't size up.

In reality, it's just different workflow, I think. Yeah, I'd say workflow. And muscle memory in general, I think, is a real tough one to get past, regardless of what your skill set may happen to be. I mean, muscle memory kind of rules the day. Colab is a great choice. Chatroom is right now talking about Colab.

And you can find them at colab.org. It's a free software initiative building a unified communications and collaboration system that you can find them at colab.org. It's a free software initiative building a unified communications and collaboration system that you can install on your own server. And it offers a lot of the functionality that we talk about a lot that Zimbra also offers. And Colab's on my short list of like the three that I'm going to be considering installing for Jupyter Broadcasting this year. This is one of the goals. I don't know if I'm going to get to it because right now, this is one of those things

where Google's doing a good enough job. Right. And I don't feel super pressured to replace them. Even though I know there's a privacy problem there and a lot of other issues there as well, tracking and all that kind of stuff. But because it's my work email, it's not like I have some sort of dirty exchange with some bankster in there or something like that. I don't

want to get caught. Right. I mean, it's not like it's just it's my work email. And most of the email that's in there gets read on air anyways. Eventually, this is true. This is very true. I'm not super compelled to do it, but I am looking into my options and Zimbra and Colab are on my shortlist. So I, you know, I, I, I actually really, uh, thank Alex for sending that into us just so I can kind of get his thoughts on it. Cause you know, to tell you the truth, I'd probably roll it out on a DigitalOcean VPS too, which means I should probably stop right here and thank DigitalOcean this week for sponsoring Linux Unplugged. DigitalOcean is a sponsor of the Linux Unplugged program. And as you've heard,

a lot of our audience is finding really awesome ways to try out digital ocean now you're not familiar with digital ocean well i'll tell you right now digital ocean is simple cloud hosting dedicated offering the most intuitive and easy way to spin up your own cloud server you get root access to this bad mama jam and users can create a cloud server in 55 seconds although i bet alex i bet alex was in the 40s somewhere although he didn't say in his email pricing plans start around five dollars per month that gets you 512 megabytes of RAM, a 20 gigabyte SSD. Yes, an SSD makes a difference on a VPS, trust me. One CPU and a terabyte of transfer. A terabyte of transfer. If you're using this for email or Yassi search or a Pastebin repository or a BitTorrent sync server,

a terabyte of transfer is going to get you a really far, a long ways. Also, DigitalOcean has data center locations in New York, San Francisco, Amsterdam, and now Singapore. They have a simple intuitive interface. But if you want more power, you want to automate some stuff, you can also implement your own set of scripts and things like that using their straightforward API. For example, you might want to utilize their API to automate image backups of your VPS systems. And they have a system to let you do that. But here's the best part.

You can get a $10 credit to try out. They have a $5 rig. That's the one I told you about. That's a $5 rig. You try out the $10 credit, you're going to be able to use it for two months for free if you use the promo code Linux Unplugged March. Brand new month. DigitalOcean is back because you guys have been loving them.

We've gotten a lot of people to try it out. And you can go over there right now. Use the promo code Linux Unplugged March. You get one of these bad mamma jammas for two months. You get tier one bandwidth connected to those SSD hard drives and backed by KVM virtualization. It's a really good system. They also have hourly pricing if you just need to test out your app. They also have images that are set up and ready to go

with LampStacks or Docker. They support Arch, Fedora, Ubuntu, all of the stuff, CentOS, all of it. They have two-factor authentication, backups and snapshots, all that good stuff. So go over to digitalocean.com, use the promo code LINUpluggedMarch to get $10 credits so you can try it out for two

months. And I think you'll see why I've been talking about them for a while now. Also, a big thank you to DigitalOcean for sponsoring Linux Unplugged for the month of March. You guys are awesome. Love it. Alright, Matt, so I want to welcome two developers from the Tox Project onto the show. Joining

us on Mumble right now, we have STQ and iRun Gen2. So I know TOX started up as sort of a discussion thread, and I'm told I think it started on 4chan. And now here we are. We're, what, seven, eight months down the road. It's become a very active project. So where is that from where it started versus where it's at today? Where is that from where it started versus where it's at today?

Well, talk started as, well, I actually, I had this idea, this very basic idea. And then, well, I was on a thread on IG. Well, that was speaking about how Skype is, well, privacy issues with Skype. Skype is a privacy issue. So I just decided, okay, hey, let's... I was actually a bit maybe drunk at the moment or something. And I said, okay, let's make a Skype replacement. So that's how it started. And then I was very impressed by the positive response from everyone. So I said, okay, let's start writing code. Let's start planning stuff. And that's how talk started. So it started kind of as a crazy idea that maybe after you thought about it for a little bit, realized this could be a good idea.

Have you worked on this kind of stuff before or something like this before? Well, I played around with some BitTorrent DHT. I did a couple of scripts. I once made a script to find out which to try to track everything on the BitTorrent DHT. But, well, there were some little bandwidth issues. I would have needed a big server and something, so the idea kind of died. But that gave me a bit of experience on how peer-to-peer software

works and everything. So that's, that's how. Yeah. Well, can you tell me a little bit about, and I guess it's actually pronounced salt, but the, the networking and cryptography library that talks is, is sitting on primarily is, is this the big piece of functionality here how does what is the what is this component play what role does that play well that's the it's uh that i seriously i i like that crypto i love that crypto library it's it's very simple to use and everything and it's very secure i don't think it talks would have worked without it it's uh since uh well well talks when talk started we didn't know uh what uh what crypto library would use right i mean it seems like a huge choice to make right

yeah but then someone suggested it in a thread and then i looked at it and i just saw this is perfect it's simple to use it's uh it's fast it's uh it's very secure it protects against all types of uh timing attacks etc so that's uh so we picked that, and I started reading a lot, okay? So how do I implement this correctly in talks without screwing up things? That was my question because I've heard that I think also the Telegram messenger program uses the Salt library, but I guess they maybe, I don't know the details, but I guess they've implemented it incorrectly. I don't know the details, but I guess they've implemented it incorrectly. Have you looked at that situation and tried to kind of balance what they did wrong and adjust accordingly for talks? Telegram, they don't use the salt library. They use their custom crypto implementation.

If they would have used the salt library, their crypto would have been fine. Nobody would be complaining about it. Right. Gotcha. So something that seemed kind of like a big deal, you guys got accepted, the TOX project got accepted into the Google Summer of Code. But you know, a lot of people have asked, is that going to influence the project in any way?

What about their independence? What are your thoughts on the Summer of Code and what it means for that aspect of the project? I'll go ahead and answer this. Sure. Well, I think with Google Summer of Code, we're really going to be able to finish a lot of things that we wouldn't have the time to do that due to our prior commitments we would never be able to do. We would be able to have people who have different skill sets in areas we might not have. It would be able to help out that way.

Okay. And what about this criticism that people have said, oh, well, this is a bad sign, this is about influence. What do you think of that? Has there been communications from Google about suggestions or anything like that? No, they don't do anything like that. They exist purely to foster the development of open source projects. No secret influences, no link to this secret library, nothing like that.

No insert go-to-fail line here. That's good. That's good considering that there's a lot of hopes riding on a secure messenger, and I've got to tell you guys, this is a pretty contested space right now. So I think people are looking for a lot of things to pull out to kind of criticize on. And to that end, I think, and I don't really know if there is something to be worried about here, but a lot of people have criticized that they started as a threat on 4chan. And have you guys dealt with growing pains as a result of that?

4chan? And have you guys dealt with growing pains as a result of that? Well, we didn't start like that, but as we developed, we got people with skill sets from Reddit and all these areas, and we sort of nurtured. That's exactly, I mean, not to interrupt, I'm sorry, but when I, so I've seen this question, it was sent into our show a few times, and to me it just seems like these are locations that technical people hang out. True, so 4chan specifically, that little area, a lot of trolling. I think it was pure luck that the people who we have now, that all the skills matched up just perfectly.

Yeah, yeah. yeah yeah well i and i and i know i know it's something that's on people's radar and it maybe seems like one of those things down the road it probably it it probably won't be as big of a factor but what any other growing pains you guys are seeing as the project is as becoming more aware people are becoming more aware of it not really well there's uh always a bit of issues, but we've dealt with them. Yeah, good. Well, okay. So that's the big picture stuff. I want to talk a little bit about maybe features and stuff that Talks is hoping to support down the road. What are your plans for three – there's three major things I've seen requests from the audience.

Conference call setups. Do you guys have any plans to do things like group calls and things like that down the road? Yes. Okay, good. What about multiple presence? As in I can be logged in on multiple computers and the message goes to all locations.

Or maybe I'm on my Android phone and I'm also on my laptop and I want my messages and calls to go to both locations. Is that possible with the security model you're using now? You know, we haven't worked out something like that just yet. We're still working on a lot of our group chats, how we can do audio and video, how we can do things like that. But when the time comes, I know we'll be able to reach that area. Yeah, it's possible, but it's later after we've done the TCP stuff, after we've done the group chats and the work. So is that how you're focusing right now?

It seems like the talks clients that I've tried are all text chat based. You get that working, and then is the plan to then to get the next set of functionality working and then the next yeah it's it's how well I think it's the best way to develop software to do make a feature test it correctly add another add another and that's so what's on your immediate list and then what's your sort of, after I get this next immediate problem solved, after I get this thing that's bugging me fixed, I really want to start working on. So what are those two things? Well, we have first, the thing that I want to finish first, well, that I'm working on right now is the TCP stuff. is the TCP stuff so that I want to add TCP relay functionality to Tox nodes so that people that are behind bad NATs can also use Tox.

Currently, Tox only works people that can hole punch through their NATs, but that's not everyone. Yeah. Some people are behind enterprise NATs, some symmetric NATs that are weird and block certain... They only have a maximum number of UDP connections. That can be a problem because, well, Tox needs to connect to a bunch of people. For discovery and whatnot, yeah.

Yeah. So, okay, I look at this and I think if Tox was really successful, if you could truly create a Skype replacement that was secure, you would have a system that would be used by journalists, by dissidents, by the paranoid, and by cheaters. And it would be, these groups of people would fundamentally rely on this technology to be absolutely ironclad solid. Have you considered code audits and maybe at a certain, like, quote-unquote official

release point, saying we're going to audit the code, have a third party look at the code? Well, yes, of course. But we want to wait till a lot of our major features are done. And talking about what you mentioned earlier, we're designing talks to be something that everyone can use. Someone talking someone somewhere, I have no idea what I'm saying right now, but long story short, we want to design something that's so fast, so easy, so smooth that people would

use it instead of Skype. I mean, we're not going to truly accomplish our goal unless everyone's using this, unless there's nothing that can be spied on, nothing that can be snooped. Security isn't a selling point for end users, but we want to give it to everyone. That's an interesting way to put it.

I mean, it's definitely for some, but yeah, you're right. For the general user, it's not really a big deal. The only really other question I had sent in from the audience that was kind of kicking around is, how do you guys plan to make money, and are you going to take donations? No, we might take donations. And so, Gentoo's been talking about it a bit. So, right now, all of our expenses, our build systems, everything comes out of my own pocket, and I haven't really been taking anything in.

Well, you're a good man, SDQ. I haven't really been taking anything in. Well, you're a good man, SDQ. So have you considered a donation system or a Kickstarter system or something like that? Can you repeat the question? I couldn't hear you. Are you thinking about... Because the reason I'm asking is we had a few people send in... Longplay wrote in and asked if you guys would... He wants to send you money, but he couldn't find a way to donate.

Are you considering doing a donation model? We might put a donation button on the website. Yeah. Because the reason I didn't want donations to, well, us accepting donations at the beginning is, well, what if we just started and then everything fails? And, well, people are going to be angry and everything. But, well, we've gone a long way from there.

So, well, we might. That's good. It's good. Yeah, you want to make sure you got something that you can show for before you ask people to start taking money. Well, so I'll open it up to the mumble room and see if there's any questions that the group has. And you guys, if you do, just ping me in the chat room with any questions you have for the folks, and we'll get through those.

And the chat room, I'm also checking your questions as we go. But guys, I want to thank you for coming on Linux Unplugged. It's this project I'm going to be watching. You know, I'm still really interested in picking the mobile messenger of the future. And right now, the Jupyter Broadcasting Network really kind of uses Skype a lot for our shows, because at the end of the day, it seems to be the sort of the best combination of deployment and ease of use and actual video and audio quality simultaneously. I'm really excited about the Talks project. We talked about a little bit this last Sunday on the Linux Action Show.

And I demoed a couple of different apps you can use for it. It's early days right now. But I want to encourage you guys to keep going at it, keep working at it, because this is a space that people are going to be more and more interested over the next couple of years. Crossroads asks, how does this compare to other chat programs, e.g. BT Chat or Tor Chat? And actually, that's a great question. Have you guys seen the announcement of Tor Chat? And what are your thoughts on that versus Tox?

Well, yes. Do you want to? Yeah, I can answer. Sure. Okay. Well, Tor Chat is they're going more on the anonymous. Well, they want to make an anonymous chat, but talks isn't anonymous.

You connect directly to the people you're speaking to because we want performance because, well, streaming high-resolution video through Tor is, well, doesn't really work well. That's true, right? Yeah, that's a great point. Yeah, I hadn't thought about the video aspect of it. And to reiterate further.

Yeah. Go ahead. I think you... I mean, we're trying to design something that everyone can use, not just the paranoid. Well, and to that end, is that why, I mean, I'm seeing a lot of different talks clients. There's not like one official talks client. I guess it would be conceivable that you could have chat programs like Pidgin out there and others that support talks and TorChat at the same time, right? Yeah, Pidgin, there's a talks plugin in development already, so it already supports talks.

Very good. That's very interesting, you guys, and I'm glad you're working on this, to tell you the truth, because one of my predictions for the Linux Action Show for 2014 was that there would be an explosion in secure messaging. So you guys are, right now,

confirming my prediction. So that's awesome for me. Most definitely. Yeah, I know, right? So you guys, I know you guys probably are not big fans of giving out timelines, but do you have a general big picture timeline or roadmap of when features will roll out or be available? I can tell you the order at which we're going to roll them out, but timelines are a bit... I'll take that.

I'll take order that you want to roll things out for sure. Well, first of all, we're going to roll out the TCP relays. Then after that, we're going to work on fixing our group chats and making audio-video work in group chats. And, well, that's the two main that should take a while because they're they're two very big features that's they're gonna need lots of testing and everything and after that we're gonna concentrate on well maybe whatever well maybe optimizing code or trying to implement other features like offline messaging and how someone who wants to be logged in to computers, that we might work on that. It depends on what we feel people want the most.

What they're asking for. Yeah. Well, let me ask you this. What if somebody, Daredevil is making a great question, making a great point in our chat room. What if you wake up one day, you go to Slashdot or Techmeme or Reddit or whatever, and you see a top headline, new secure messenger launched available on Android, iOS, and the desktop. It's awesome.

It's rock solid. And you dig through it. Oh, and you see they're, and the desktop. It's awesome. It's rock solid. And you dig through it. Oh, and you see they're charging for the app. And then you'd read through the details and you realize, holy crap, they're using Talks. How do you feel about people creating clients or potentially business models on top of a system that you're building today? Well, the core is licensed under the GPL v3, so they can't take the code and not give the source back. So, well, if someone does take it and improves their source, and if they make money, I don't care because they're improving the software.

Dare Devlin, I didn't realize you were in here. Go ahead. Did you have more you wanted to add to that? Dare Devlin? No? Sorry. I'm here. Can you hear me? Yeah. Yeah. So what happened is the reason I got this question is because I was wondering more. I know it's GPL and people are contributing back, but there's kind of business models that could involve, for example, a company making their own nodes in a way that would allow them to mimic users. I mean, it depends, of course, on the intrinsics of the project, but then goes to something that is a little bit deviated from original tax plans. And it goes like it uses stock source code.

It uses all the costs of the tax. Like Bitcoin's Litecoin, that kind of thing yeah kind of sort of so it will be with some differences that will make more uh a company driven a perspective and not actually providing the users direct access so what are the measures in place for you to incentivize people to just come to your project instead of just working doing something crazy that goes completely out of town well if if a company does something like that well and users actually use the software from that company that means they're doing something good so that's a win for you. Well, we can always take the code and put it in our project. Right.

And, well. So, Josh. It's like SSH. Go ahead. That's a good point. It's like, okay. Oh, okay. I mean, that's an interesting way to think about it.

Yeah, like OpenBSD makes SSH and they give it away for people to do whatever they want with. Right. Great point. I mean, that's an interesting way to kind of frame that. Josh, you had a kind of a question more about funding, right? Yes, I did. So you guys talked about maybe monetizing or accepting donations in the future. So what is your opinion on allowing people to back certain features with something like Bounty Source and allow you guys to get paid to implement certain features like group

chats and maybe help speed the process along or provide some sort of monetary incentive? You know, I think something like that really hurts people's trust if something doesn't need a timeline. If something looks good, looks like it's easy to do, but it doesn't come up and it ends up taking a lot longer than it should or something. People get disappointed.

Well, the main issue is that implementing something is relatively easy. Implementing something that's bug-free and works perfectly, that's hard. That's an interesting point. fascinating to really, it truly tells you what people's priorities are when they assign a dollar value to it in a way that is so much more tangible than emails and bug requests and all these things. When people are actually willing to put money down for something, then you really know it's literally of a value to them. But I do understand you don't want to overset expectations. And you guys seem to be walking that line pretty closely. I'm going to take a quick break right here. And when we get back, I want to talk a little bit about maybe your thoughts on the future of mobile for talks. But first, speaking of mobile, I want

to thank Ting for sponsoring Linux Unplugged. Go to linux.ting.com. Yes, linux.ting.com. That's how awesome Ting is. Now, what is Ting? It's mobile that makes sense, my friend, my mobile service provider, and of course, Matt's mobile service provider. You bet. And Ting has no contracts, no early terminations. And by the way, you only pay for what you use. It's $6 a flat rate every single month. And then just your usage on top of that. Every Ting plan includes hotspot and tethering. So you just go on Android, you just check that box. Every device you buy, you own it. So you're not annotizing that cost over like two years where you're still paying into something that's not nearly worth that. Plus, they have a super awesome dashboard that supports web standards that is, while simple to use, extremely powerful unless you set up a whole bunch of stuff you want for your phone and they have no hold customer support. But what I

really like about Ting is they're actually good people. They're a good company, they're good people, and they care about their customers. I'll give you an example over on their blog. They just posted yesterday six simple steps to save money on your smartphone mobile data. And they go through some of the big apps that a lot of smartphone users use, and a lot of the settings that already exist in those most popular apps like YouTube and Google Maps and Chrome and Facebook and Spotify and Gmail. They have all these different settings you can go through to actually conserve bandwidth, to save a little bit of money. And they've done posts like this in the past, too. They talk about Opera Max, for example. They've done posts like this on the minute spectrum.

And this is, Ting is, because the way Ting is structured with their flat rates and the pay for your usage, they really have a system that allows them to super serve the customer and not try to nickel and dime you like every other carrier does. And even the carriers that pretend like they're trying to make a big difference, they're actually still nickel and diming you. So go over to linux.ting.com. That'll save $25 off your first Ting device. If you already have a device that's compatible with the Sprint network, well then bring that with you.

And then Ting will take $25 off your bill. They'll give you a $25 credit. By the way, they've got an early termination relief program too. And that's awesome because if you have a contract and you know that they're screwing you, you know you want to get out of that contract, you're sick of the BS, all you have to do is grab your Ting device, port your number, and then you submit an ETF claim over to Ting and they'll pay you up to $75 per line that you had to get canceled.

And then once you get in there, the savings is crazy. In fact, check out their new savings calculator. They've updated it recently. If you have Verizon or T-Mobile or AT&T, you can plug that in there. It'll go get your bill information for you. If you don't, you can just plug in your data like you normally would. Tell them how many cell phones you got, and they'll tell you how much you would save. It's crazy when you thank you to Ting for sponsoring Linux Unplugged. Alright guys, so how about it? Any plans for mobile support? A Tox mobile client down the road? Yes. Well, of course. You can't get adopted without

support. You need to run everywhere. It's like a must-have. So are you going to run everywhere yeah that's like a must have so are you going to build that app yourself no we already have people building the apps we have tcp relays need still be added but that'll help with a lot of mobile connections we have apps already done in there's a city app and there's an android app already. Android doesn't work yet. And the Cydia app needs a special repo. But this is all being done already.

We'd love to see that on Ubuntu Touch. I bet. That would be a really good app, though, right, to have at launch? Yeah. And they'd probably get a lot of exposure, too. And they'd probably get a lot of exposure, too. Yeah, I know some people within Canonical are very keen on selling Ubuntu Touch on the fact that it's possible for people to very easily create these kind of secure apps. And we're not somewhat tainted with the three-letter acronym companies uh like right poking our users yes uh i know dare devlin you had a comment about freelance developers i wanted to let you get that in go ahead yeah um

the reason i was talking about freelance developers is even though you guys don't want to compromise or commit yourselves to a timeline i guess uh if like it will be a way that you allow people to just come around and say I want to do this for just as my freelance gig if someone wants to I think the risk of the time and things do you want those people to be directly associated with the project do they need to contribute the patches on their own, in their own branch, and then involve later? The reason I'm asking this is because actually I'm pretty fascinated by your project, and I've actually gone through two universities

and there's quite a lot of papers, and I will be interested, but there's business models on top of this. So that's why i kind of have these questions to you well if you want to contribute just come to our our irc channel and tell us the the only problem with uh contributing a lot of code for example the core is that uh well core stuff is very touchy. It has to be no security bugs or else some blogger is going to destroy us. Yeah. That's how that happens. No, no. But I feel you didn't answer my question.

Sorry. Is that, so a developer that I want, I have a client. That client wants me to implement something. I'm going to be paid for this. So where do I go? Do I go directly to try submit to these patches and these new feature with you? Or do I do it for the client on my own branch and then invite them to go meet you once the feature is done?

Well, you come and see us first to see if we have plans to implement a feature like that very soon so there's no duplicate work. And then, well, you just work on your feature or you do whatever you want and we'll adjust. Okay. All right. And Crash, you wanted to make a point on the sort of cross-platform adoption we're seeing already, right? Yeah, well, I mean, just talking to the guys in the IRC, in the Jupiter Broadcasting IRC before, there was a guy in there saying there's, said no because they didn't want to sign the FSF CLA purely for the fact that then they could have an iOS client as well,

which I thought was a very smart move because the CLA, like there's, it's a good thing in many ways, but I'm just doing that so that they can even have a client for iOS, which it just makes it more ubiquitous so that everyone can use talks. Yeah, that's interesting. Thanks for sharing that. And I know, go ahead, STQ, you wanted to respond?

Sorry, I was just testing. I couldn't see my microphone icon, so I was just making sure. So, Crash, you've been following the project since the beginning. Did you have any questions you wanted to ask the guys? Well, not particularly. I was actually quite interested in what their opinions are of WebRTC as far as the video part, because I thought that might be, I mean, that's more of an idea thing. But, you know, if they built a little,

in one of the clients, maybe it's a little WebKit engine or something that can connect to WebRTC, but I don't know if that's P2P or not, so they might not be, they might not work with the goals of the project, but that would be an interesting way to do some things. That actually kind of dovetails to a question I was going to ask. What are your thoughts at this point,

maybe this is probably best for iRunGen2, on the actual implementation of video support? Are you looking at things like WebRTC? Are you looking at codecs yet and how it's going to work, resolutions and all these kinds of things? Or are you not at that spot yet? Well, audio-video already works in the core it's it's basic but it works currently it uses a vp8 vp9 isn't quite ready but switching from vp8 to vp9 it's a

two two line change in the code so uh it's it's not really a an issue but uh yeah but first uh first before developing uh audio video more we need to get it into clients because right now no clients really support earthsit we have our test client that supports it but that's not enough you want to get some real testing for it. And once you have some real testing, you can start maybe adding more stuff to the video. Right now it's very basic. You set the resolution, then there's no congestion control. Yeah, I mean, the fact that it's the it's you it's it's going to be using vp9

down the road and i actually get to set the resolution already puts it a notch above skype my friend so you're already it's already a good thing i know i know you had a question sort of uh sort of maybe about the fundamental separation between the core project and the clients and maybe that would that would sort of get around some of these licensing problems? Yeah. So there is this problem with the CLA for a free software foundation, but the way I understand the free software foundation CLA

actually ensures that even if the free software foundation for some reason goes to bankruptcy or fails, a company that takes over cannot transform the free software, the proprietary software, which at this moment is something that, I mean, I'm not saying that you guys are going to get bought out, but it's possible. And the project will just buy from there

or go to a different direction. And I see that if the reason is the iOS client, couldn't just the iOS client be completely separate from the main project and be something that is just kind of, it can be developed by the same people, just not under the same umbrella? No, unfortunately, everything has to go through our core library

to essentially communicate. Well, our core library is GPLv3, meaning the clients would have to be compatible. We've been working on making sure we have the copyright for all of our code so we can selectively assign a dual license for an OSI-compatible license so we can allow support on iOS while still ensuring every client, every library is 100% open source. Aren't you just recreating a CLA, though,

if you want to keep the copyright under your umbrella and everything needing to be through you basically alright and one last obligatory question before we wrap up you guys, how far along is the Firefox OS version of the Talks Messenger that's like HTML and JavaScript based, right?

Yeah, actually, what about a web chat? Is that possible? Is that something you guys are considering? Well, I mean, on Chrome OS, you can use Native Client. On web apps, you can use Native Client. I don't know how you communicate with a C library in Firefox. Guys, I appreciate the fact. This is really the reason why Tox has moved to the top of my list right now is just good desktop support and that promise of replacing Skype.

Because as somebody in the podcasting space, a lot of podcasters are very dependent on this system. And it has a lot of limitations. And we're not their focus. And you joke about it, but really, just the ability to set the video resolution right now is not even an option under Skype. And if you think about some of the major multi-million dollar networks out there, that are relying on Skype to do video production, and you can't even set the video resolution. And that's aside from all of the massive security problems that Skype now has. Thank you, Microsoft. And also just that, you know, it's a compromised piece of software. People DDoS Skype nicks for joy all the time. And it's

completely just messed up. So I'm watching talks with a lot of interest, you guys. So thank you very much for coming on the show today and chatting with us. Before we run, I wanted to just get a question out there. And I got it in from Scott. He said, hey, guys, I want to express a growing concern I have held for the past decade with open source software and the recent activity, things, and going-ons in the Linux community right now. Examples are Firefox putting ads in, Steam bringing DRM games to Linux, NVIDIA proprietary drivers exploding in popularity, etc. Linux is always touted as being a free and open system, but are we moving into an era when less and less is free? When I talk about free, I mean as in freedom. The reason I bring this up is because you are very positive of Steam

coming to Linux. He says, in many Linux groups, however, I think some of the coming to Linux. He says, in many Linux groups, however, I think some of the sentiment is opposite. I'm good with paying money for legitimate software, but will the future hold me paying for software that is not open source? Do you think as Linux gets more popular and more big players come into the arena, that the whole open source notion is kind of just a nice dream?

Love to hear what you think. Scott, anybody in the mumble room have strong thoughts on this yeah uh i do if um in my opinion anyway uh as long as the core components of linux remain free so as long as like the kernel and all the um intrinsic parts of linux remain free that's all that matters. It's fine if you want to put proprietary software like games on Linux, but as long as the core components stay free, it's fine. What do you think, Riley? Yes, this is why there's still things like Driscoll Linux and official Free Software Foundation districts.

So if you're not concerned about it, just use that and leave everybody else alone. Wow. Get off my lawn is essentially what Riley is saying. Exactly. I think what, you know, Scott is kind of saying here is he's worried about the groupthink syndrome where we all kind of just, you know, as a group decide that sort of allowing this encroaching intrusions on freedom is okay. And so, well, it's just our video driver. Well, I really need good Wi-Fi.

Okay, well, this video game cost them a lot of money to make. And we keep rationalizing it. And every single time we rationalize it to some sort of group consensus and we all just kind of collectively give up on that particular thing instead of holding the hardware vendors to actually making an open source GPU driver and things like that. So I actually think what Scott's bringing up is something that we all have to just stay vigilant on.

And I don't think it means that we stop playing Steam games, but at the same time, I look at it from a business standpoint and I say, okay, well, core infrastructure components for sustainability, for long-term predictability, and for transparency, those just have to be open source. All other things aside, freedom aside, just if my business is going to rely on this technology, I need to be able to know where the hell that technology is going. And if the person making that technology disappears or gets burnt out, at least I can fork it and keep it running myself. That's why people who have their entire, you know, like server infrastructure or desktop system dependent on Apple. Man, that would make me stay up at night because you never know what Apple is going to do next. Are they going to kill the Mac mini?

Are they going to drop the Ethernet port? Are they going to release iOS for the desktop? It would drive me crazy. Are they going to drop all PCI Express expandability and internal storage? You just never know. Whereas with open technology, you at least have the mailing list, you have the roadmap, and you have that almighty fork. And I think for core infrastructure technology, and you could argue with the GPU driver, and that's one of the reasons I've been eyeballing Ultra Pro, because I kind of like the idea of getting an Iris GPU, just to make one more component less dependent on some upstream vendor. Even with all of that considered, to make exceptions and where certain practical matters come in, like in some cases video editing or in some cases video games, I think commercial can be accepted. I mean, Matt, this is probably something you've struggled with for a long time.

Where do you fall down? You know, I think at the end of the day, it comes down to the point you were making about core components and that sort of thing. I think as long as it stays out of the kitchen as far as getting really deep into stuff, I don't think this is really such an issue. Because for years there's always been the concern of this becoming a major, major catastrophe and oh my god, it's the end of Linux as we know it because of this and that and the other thing. As long as we have distributions that are able to maintain

those core components pure, if one distro wants to go south with something, that's their business. I don't see it as being a big problem, I guess that's what I'm trying to say. I really don't. Yeah, and it's come and gone over the years. In some ways, it's actually gotten better. These days, a lot of systems actually have a lot less binary blobs powering that hardware.

I mean, there's still a few. But there was a time when I got a Red Hat. Jeez. I mean, I think it was back before Fedora Core was even a thing. So I want to say like Red Hat 6. And I had a binary driver for the Adaptec or some sort of SCSI controller. I had a binary driver for the network card because it was a special network card to communicate with an IBM System 390 mainframe, right? And then it ran on top of it a proprietary management software to monitor and manage it, which has now been completely replaced with things like Puppet. And so, in some ways, while we have Steam and we have NVIDIA binary drivers, in other ways, we are

actually working with a lot more free software. And as the technology landscape moves forward, and there's less entrenched interest in these certain areas, like there used to be a big market around creating server SCSI adapters. And you wouldn't even use the adapter that came with the box on the motherboard. You'd never use that. You'd always use an external adapter. And you'd have to have have the driver and you'd have to build the module for that version of your kernel. You couldn't update your kernel and you were really stuck. And that's not even a problem anymore. Now those guys work directly with upstream and those drivers ship with the kernel and they're open source and you don't even have to worry about it. The drives just show up now

to your system and you kids don't even appreciate to appreciate what us old farts used to have to deal with. So in a lot of ways, Scott, I actually think it's gotten a lot better too. I think we are always looking at the individual trees, but if you were to fly up like Superman above that forest, you'd see that a lot of it actually isn't that bad. That's my take on it.

Alright, Matt. Well, so we're going to wrap up here on Linux Unplugged. Thank you big time to the Tox developers for joining us today. It was a very fascinating chat, and good to get their perspective on some of the things and kind of get an idea where they're going. Look, on Sunday on the Linux Action Show, I'm going to review that Chromebook, that Acer C720,

and I'm going to try to answer the question, can a Chromebook running Linux, not Chrome OS, but actual full-fledged Linux. Can you get decent battery life? What's the performance like? And is it actually something that's worth spending money on? Or should you just stick with Chrome OS or maybe get a tablet, something like that? We'll try to answer those questions on Sunday. If you've got any questions, feel free to send them in.

We can start a thread over at linuxactionshow.reddit.com. You can post them there. And we'll try to get to those questions in Sunday's Linux Action Show. All right, Matt, well, have a great week, okay? You too. I'll see you on Sunday, and thank you, everyone, for tuning in to this week's episode of Linux Unplugged. If we don't see you on Sunday for the Linux Action Show,

I hope we'll see you right back here next Tuesday. Bye-bye. Bye-bye. Thank you.