LINUX Unplugged - Episode 175: Best of MATEs | LUP 175

Episode Date: December 14, 2016

We get the inside scoop about some fantastic collaboration happening between three Linux distributions that are supposedly big competitors.Plus Google’s response to Ubuntu Core & the big NextCloud news!

Now, before we start, I was wondering if I could get Wimpy to tell us about, I think Wimpy spent the morning, or actually for him, he spent the afternoon evening at a makerspace. Is that true, Wimpy? I have, yes. And tell me about this. This sounds pretty cool. So this is a new makerspace. We don't actually have a space yet. We're just a bunch of makers right now. We don't actually have a space yet. We're just a bunch of makers right now. So in my local area, I live in an area called Basingstoke and Dean. So this is Bad Hack is our maker space.

Nice. That's awesome. This is like the third meeting. I've been to the last two. And we're just sort of drumming up members and figuring out who's got what skills. And we're doing little presentations and show and tells and things, figuring out some projects that we're going to work on and also looking for somewhere that could be a permanent facility

that's open 24-7 so that people can just go there and work on projects whenever it suits them. And this evening was good because we've sort of reached out to the craft makers in the community so we had a load of people that um do um crochet and knitting and all sort of textile crafts there this evening so we've got quite a broad range of skills now to go with all the software and hardware engineers and roboticists and electronics and PCB design guys. So we've got a good spectrum of skills. So we're starting to think up some group projects we can work on, which will be fun.

Yeah. I feel you're making me feel like the worst. I was invited to a local makerspace where they're building an open source car, which has got to be cool. And I couldn't make it work. I couldn't get the timing right. I wasn't able to go. But now I kind't able to go. But now I kind of want to email them back because this looks like a lot of fun.

That's a bit more ambitious than our first projects. Well, you've got to start somewhere. Yeah, we're going to have a go at making a 3D scanner because one of the guys in the makerspace runs a 3D printing business. So we thought we'd have a go at making a 3D scanner using a make our own rotating table with Microsoft Connect as the scanning eye. How much of the open source ethos is alive and well in a space like this? Is there a tendency towards open source and Linux, or is it more about building the thing and less about what software you're using?

and Linux or is it more about the building the thing and less about what software you're using? Yeah, I'd say the latter. So people are aware of what open source is. There's what I quite like is there's quite a lot of the older generation there sort of people that have worked in electronics, all their careers and have recently retired. So they still refer to everything as Unix. But the fact that they know unix is encouraging so people know about linux and open source uh some of them dabble with it and use it obviously raspberry pi is you know prevalent so they're all kind of aware but yes they're they're very sort of practical in the way they use things so it's just a tool so if windows does the job better so the prototype 3d scanner is using windows 10 at the moment.

I intend to change that, but it runs Windows 10 at the moment. That's because it was easy for them to do it. So there is some awareness, but it isn't the core running through the group at all, I would say. I'm the only person that is really sort of invested in open source. The rest of them. Really what you're saying is, is that your work is definitely cut out for you. Yes. This is Linux Unplugged episode 175 for December 13th, 2016.

Welcome to Linux Unplugged, your weekly Linux talk show that just can't avoid feeling that holiday spirit. My name is Chris. My name is Wes. Wes, you got a mug in your hands, too. It looks like you're sipping on, like, some cocoa. One can pretend. You could bring up the eight-hour YouTube fireplace on the screen there. It'd really feel festive in here. I'm sure the mumble room would just love that.

That's not what we're here to talk about today. We're not here to talk about that. Although, maybe in the post-show. You never know. It can go crazy. Stick around. But coming up on this week's episode, there is huge, huge news from the NextCloud project. We'll highlight some of that. And then the Plasma desktop, it's going to be getting all up in that next cloud.

You're not going to believe the kind of stuff they're thinking about integrating in the Plasma desktop. McAfee's making your Linux box way more vulnerable. Linus himself is feeling the holiday spirit. We'll tell you about the present. He's dropping before Christmas. My buddies at Crossover have big news. There's a big security thing to talk about.

And then I want to discuss the completely ridiculous, irresponsible, and maybe even dangerous collaboration happening between distros. If this kind of stuff happens, we're going to start getting results and stop wasting time. And I can't stand for it because then what am I going to come on here and rant about? So we're going to take it to Ikey and Wimpy later on in the show and ask them why they're doing so much damn good work and what the hell's going on with all this collaboration. And then after that, Google announced their plans today to kill Linux on Internet of Things. They're looking at Ubuntu Snappy.

They're looking at all of the efforts by different projects out there to create a free open platform for Internet of Things. And Google says, no, no. for Internet of Things, and Google says, no, no. Let us introduce, let us suggest, let us perhaps recommend that instead you deploy Android of Things. Android of Things is Google's new Internet of Things platform, and the name is horrible, and we'll tell you about their plans to try to crush the competition on Internet of Things, which I think they could have something there. I think they looked over at what Canonical is doing right now with IoT devices, and they went, shit, we better get rid of this.

What did they – Brizle? They had another – Oh, yeah, they did. What was that called? It was like a kitchen – like a brassel. I can't remember what it was called. But it was essentially an Internet of Things platform that also used Weave, their mesh networking. And it didn't go anywhere. Brillo, thank you.

Brillo, yes. Look at Micah68 in the chat room, nailing it. Yeah, so they had Brillo, and Brillo hasn't worked out. Now we're going to discuss the pivot to Android, or Android of Things, or whatever it's called. But let's start with some actual really awesome Linux news, some project updates, huge news from NextCloud and all of that. And it just would be inappropriate to go any further without saying, time-appropriate greetings, VirtualLug. Hello.

Hello. Greetings, program. Hello. Hi. Hello, everyone. So very, very excited. And I'm hoping Joss will come on the Linux Action Show this Sunday to give us the real inside scoop. But I'm very excited to say that NextCloud 11 is out, and it looks like a really kick-butt release.

First of all, off the top of my head, the thing that I'm most excited about is two-factor authentication support. That's really big for me because the whole reason I want to use NextCloud is because I want something that I don't trust putting in Google or other cloud services, which means it's something I really care about. Security is important. Yeah. So they also have support for new browsers, Kerberos authentication, two-factor authentication, and time-based one-time passwords. They have also expanded the brute force protection on all API access points, which is brilliant. And they've made Federation more secure by better use of SSL and TLS. Excellent.

And they're making sure their app store automatically checks apps and enforces their signatures. So you're getting legitimate apps. These do sound like quite a number, you know, a nice layer of security polish on the top, especially when this is the kind of service that you very well

might want to expose publicly so that you can access from anywhere. Also, look at this stat. This release, NextCloud 11, decreases database load

by up to 80% and improves response time by 60% for common server operations. They've also, they have also combined multi-bucket object store support. Now, don't worry what that means. Wait to the next story. Improved handling of previews in Collabra Online,

so they've sped up the online document stuff. This is really interesting, including a lot of nice improvements to back-end searching. So, yeah, I'm hoping to dig in. Yeah, full-text search now. That's great. Yeah, I'm hoping to dig in with JAWS and try it out myself. Wimpy, is this the version? Yeah, it was Wimpy before the show.

You said you tried out NextCloud just recently, right? I've just installed it this afternoon. Was it 11 by chance? Well, when I installed it this afternoon, so I've installed it in a slightly unusual fashion, which is why I've written up a blog post about it, which I've sent you a link to. I see that. I'll installed it in a slightly unusual fashion, which is why I've written up a blog post about it, which I've sent you a link to. I see that. So I recently ordered the Nextcloud box, and that arrived.

And I only had Raspberry Pi 3s, a spare, whereas the image that comes with the kit only supports the Raspberry Pi 2. the raspberry pi 2 so i have installed ubuntu core on a raspberry pi 3 in the next cloud box and documented how to install the next cloud snap on that and access the one terabyte hard disk using ubuntu core so i can confirm that it works because i followed his guide about an hour ago and it works perfectly. What is the process of installing NextCloud from a Snap? What does that involve? Snap install NextCloud. That's nice.

Interesting. And I've just been talking to the maintainer of the NextCloud Snap, and NextCloud 11 is in the candidate channel in the snap store and we'll move to stable very soon now and i've just refreshed my install and i'm now running uh i'm at next cloud 11 interesting and what are your thoughts so far i don't know enough about next cloud to comment on that well just a general setup and use what are are your thoughts? Well, yeah, it's great. Logged into the website and the web UI thing and turned on the bits and pieces I needed to access the one terabyte drive and turned on some extra apps like contacts and calendars. And I've imported my contacts and my calendars and I've got those synced to a tablet.

Nice. And all of that was really easy. to a tablet nice and all of that was really easy and i've turned on my wife's phone when you take a photo or a video automatic up automatically upload it and that works as well so so far so good i'm very pleased but this is like i literally started this last night and finished it at lunch time today so it's all very new to me but i'm very pleased so far it's looking good it's certainly the best experience i've had last time i tried own cloud was like own cloud eight and it still wasn't working right for me at that point but this looks much much closer to what i need i'll follow with some interest your thoughts and um just sort of overall feedback on how the

syncing works too as you get a chance to play with that. And those features that you mentioned from NextCloud 11 in terms of reducing the database overhead and everything, that would be great for running it on something small like the Pi 3. That could make a significant difference. So yes, that would be terrific. Pobi, what's your experience been? Just a little bit of time you've had with it so far,

the UI, overall feel of it, and all that. So the only negative thing I have is the fact it's a bit slow on the Raspberry Pi, but I'm hoping that some of these performance improvements you've said are in ownCow11 will help. But I'm not expecting magic instant 10 times performance improvement. But I'm planning initially just to use it for the photo sync sync i want to be able to have the photos from my phone and my kids to be able to take photos and it just automatically sync to the next cloud box so i i followed martin's guide i only used a pi 2 because it was the only thing i had kicking around like you know first world problems

and and i'll install the next cloud client on each of their machines and have them sink in all their stuff and do that over christmas should we get fun but yeah i found it works really well and i and i love the fact that i can just go snap install next cloud there's a little bit of faff to do um ssl but it's only like a couple of lines there's not a lot to it and there was only a couple of lines to paste in to do to mount the one terabyte hard drive so it it wasn't really an awful lot of work and thanks a lot martin for doing that work because i i basically just copy and pasted stuff into a terminal and it worked that's great yeah you shouldn't do that even the author says make sure you try his next article no malicious and i'm just i'm

trustworthy apparently well none of them had sudo in of them, so I feel okay. There you go. Yeah, I think the two things that I also think about, like playing around with the federation, so I could have a local Nextcloud instance and maybe like a droplet ones where I would hit from my phone. It's a little faster. All that kind of stuff. There's a lot of possibilities there. This is next story is what could push me over the edge. Big time. This is so awesome. It what could push me over the edge. Big time.

This is so awesome. It fits with things you've been flirting with lately. So the Plasma folks have been discussing further integration with NextCloud. And disclaimer, I think they're more discussing having a common API that would work with own cloud and NextCloud. And they're talking to both projects about that. But the people they met with was Frank and others from NextCloud. So that's obviously sort of how this is framed. It's written by Joss from NextCloud.

Excuse me. It says, a feature that has been discussed is the synchronization of settings like passwords and file metadata over to the NextCloud server. From an implementation point of view, the latest NextCloud features a key value

store, which could be used to store client settings. This is the same way that iCloud works. When you're an iCloud developer, you can store application settings in iCloud using an iCloud key value store system. So this would be a key value store system that you could use to store

the client settings of the Plasma desktop. Another area, they say, is they could also do the syncing of tags, ratings, and comments from applications in the file manager, which kind of blows my mind. And then, of course, being able to just offer general file sync integrated into the file manager. They say the most important part would be to write the online accounts module, which Nolm actually has this part done, where you could set up NextCloud to begin with. And then they want to evaluate how setting up NextCloud could then be potentially automatically set up

with the KDE PIM software. So your NextCloud calendar works in your local Plasma desktop calendar. And also you could potentially, if you wanted to and they implemented it, use the Plasma Desktop search to not only search your Plasma Desktop but also the next cloud server for files. Wow. Oh, that is cool. Yeah.

That is pretty slick. So, you know, not only is this a way to synchronize your settings in a way that feels like it should be pretty good since it would be supported and built into the desktop environment. But you could keep it completely private. Run on all your own servers. This is what it should be. This kind of collaboration and this kind of communication and this kind of, hey, you're doing this really well. And we're doing this pretty good. Why don't you just do what you do best and we'll do what we do best? Because these guys could have gone with some sort of crazy plasma online that's $9.99 a year that does all of this for you.

They could have done something like that. This is genius, and I would love to see this happen, and I would love to see other desktops go – Wait, what are they cooking over there? Yeah. What are they cooking? Maybe I should do some of that. Maybe I should do that. I don't know.

I think that's pretty cool. I would love to see really solid – I'm going to try to put – I think I have recorded Wimpy's new uh settings sync solution i'll try to get that put in the post show so that way if you guys are curious stick around for the post show where wimpy discusses how he's doing this essentially today but this would be baked into the desktop level would you would you be tempted to give it a shot yeah i mean i think i don't know if i would use it long term or not but with sorry with this number of like that kind of integration if i've already got you know all of my environment settings and then you can you know you already

have it there syncing so why not put your dot files on it why not put in all everything that you need that you want to carry with you you pull up a new laptop you install plasma get the settings configured and boom you're rolling that's awesome and if you do one login and all your settings syncs and your file sync setup, this for me makes like investing – even if NextCloud does have a few issues, like I could build backup systems around that. Like I can figure out ways, especially if I'm taking regular snapshots before updates and whatnot. Like there are ways to mitigate any potential downfall I might find if they get to this level of integration. And if the other desktops ever consider this, or even if just the file manager projects implemented this, or a few other projects did this, this would be huge. I love it. So you can find out more.

There's not a lot of meat to it yet because none of the work has been done. This is all just theoretical. They don't even have the common API yet. So it's not really all that close. One other story, just because I don't really think it affects a lot of us, but I think we'll all collectively like to roll our eyes. Get ready. Get ready. Do one preparatory blink and now roll your eyes.

The Caffe virus scanner is allowing remote code execution as root on Linux. So when you get that virus scanner that you don't need for your Linux box, taking up resources that you don't really have available, scanning files for the Windows users, it turns out that if you have the Intel McAfee virus scan enterprise for Linux, it can be compromised by remote attackers due to a number of security vulnerabilities.

Not one, not two. Some of these vulnerabilities can be chained together to allow remote code execution as root. Yeah. So basically everything they've released from 2015 until now has the vulnerability. The only difference from the older release appears to be updating a newer version of LibC, which makes exploiting these vulnerabilities even easier. Oh, boy. Yikes.

Yeah, so that's what's up. And now it's out in the wild it's known about and uh there are cves and all that to track so you've probably already been informed about this if you run it but they also have a rundown it is 10 vulnerabilities 10 dang yeah each one got its own cve so that's brilliant and uh the other thing that's just wonderful, and it's so great because before I say this, you're not going to be too shocked when I say it, but you have to remember in the context in which this product is sold, this is sold as a security product to run on your enterprise servers that are connected to the internet to make them more secure. It is implicitly sold to make your server more secure. And so when you are designing enterprise software, it's not shareware, it's not $9,

it's hundreds of thousands even with yearly annual subscriptions and support. And you are designing an application to be sold in this environment, you build it for security. And the way, just basic way you could start that is run the individual services as unprivileged users. But not McAfee.

Nope, nope. Not one, folks, but two separate services running as root and only one running as an unprivileged user called Nails. Nails. And the great thing is the main scanner service runs as root and listens on a local Unix socket. So that's beautiful. So if you get on the box, if you use any other vulnerability to get local access to any of the other nine, if you use any of the other nine vulnerabilities to get local access to the box, you then can just access the local web server, the web server listening on local

host running as root, which is brilliant. Then the main scanner service also runs as root. That's just brilliant. And then the web servers, there's another web server that runs as the user nails, also listening on 0.0.0 on port 55443. Interprocess communication is basically a web server that sits on top of other pieces underneath it. And when the user makes a request to the web server, the request is reformatted and sent to the root service. And then the user is shown the response rendered in an HTML template. The web interface doesn't do much

to limit what data a malicious user could send to the root service, so you can probably figure out a way to just use the web service, and it'll just pass it right along to the thing running as root. Just passes it right. Here you go. Here you go, right? It's just raw input from the web. It's just brilliant. Built

as a product to make your Linux enterprise servers more secure with the intention these boxes are likely connected to the web because they're mail servers and things like that. And this is what they do. Crazy. Well, expected, though. Don't.

If you're going to run antivirus on Linux, just use Clam Antivirus. All right? Do me a favor. Do you a favor. Don't make Chris mad. Don't spend thousands and thousands of dollars to make Haffey. Go do a donation to the Clam AV project and call it good. Right?

Clam AV has been around for years. It gets updated like nobody's business. It's easy to integrate. It doesn't have a ton of overhead compared to these things. And it's open source. You just, I mean, there is not a competitive market. I don't know why – I don't – like I've seen this before.

I've seen people asking about antivirus before. If you must run antivirus, use Clam AV and if you need to spend money on antivirus, give a donation to the Clam AV project. Problem solved. Problem solved. Go ahead. Arcanos? Problem solved. Problem solved. Go ahead. Yes. I have found several third-party Clam AV virus repositories. There's one that I'm using on my personal squid proxy at home to scan all the data.

How do you know it's a legit third-party? How do you know? How do you find these and how do you know they're okay to use? This one, they have paid repository parts and they mentioned something about, I think, US government using them or something like that. This is like a business, this is like a way they generate

revenue. Yes. Oh, okay. Interesting. So they're like a curator for ClamAV databases and can you use multiple Clam AV database sources? All it is is adding into the Clam, the Fresh Clam config. Yeah. Fresh Clam. That's right. I forgot about Fresh Clam. Wow, it's been a long time.

So that is really interesting. I don't think I ever got down that far in the rabbit hole when I messed around with Clam AV. Thankfully. I just set up a cron job. Let me see if I can find the website, and I'll link it to you guys in the IRC. Yeah, of course. Like Ike says, you can still charge for open source. Of course, Clam AV could choose to sell it if they wanted to.

They just don't. It's just right there in your repo. I love it. There's a lot of value adds like we're seeing here. That is really a thing, though. Just those root services, Wes. Anything that runs as root and then claims to make your computer more secure. Just I know.

Right. And walk away. Walk away. I mean, I mean, I only run as root, but that's because I know what I'm doing. You know what you're doing? You meant to delete those files. Didn't want them anymore. Oh, man.

You know what? I get grumpy at actually. This is such a first world problem. But like when I had this is such a first-world problem. But, like, when I have – this is such a first-world problem. But when I have, like, a lot of packages to update and I've pseudoed, and then sometime during the build of an AUR package or something, the pseudo time window is expired. And then the package just sits there waiting while I reenter the password. Usually, because I've walked away, I come back. Yeah, you're gone.

And then it's timed out again. And now it's asking for the root password. And I'm like, I don't know. I come back and then it's timed out again. And now it's asking for the root password. And I'm like, I don't know. I never log in as root. I set up some crazy password I generated. Like, oh, come on. Can't you remember? So I'm always tempted to just pseudo the whole session.

But then, of course, now the package managers yell at me. You really shouldn't run this as root. To run this as root, you must do tack, tack. Come on. It's like the way my truck beeps at me when I'm driving down my driveway to empty my garbage can. It's like, you don't have your seatbelt on. Yeah, you better put that seatbelt on. You don't have a seatbelt on.

I know. It knows best. Just trust the system. Oh, man. Speaking of the system, help break the system, the duopoly that is the mobile industry. Go over to linux.ting.com and sign up for Mobile That Makes Sense. This is a gift you could give because the SIM cards are like $9. The phones, they have a whole range of phones starting at like $20 after you go to linux.ting.com.

All the way up to the Cadillacs. Linux.ting.com will take $25 off a device or give you service credit in $25 if you bring a phone or a tablet. I was just – this is going to be a funny story. So this is – I have a – just last week I have a new use for Ting, which I've – when I got – ages ago, I got an old iPad like Air or something. It's like the first iPad Air. And this is what I'm using when I fly my DJI drone.

So I have a nice big screen to view the DJI drone on. Here's the thing, Wes. There are regions like where the studio is at that are restricted air. And I have to sign in and authorize myself before I can fly in that area. And you got to have data. So here I am like a jerk out in the freezing temperatures with my drone and I can't take off. Endangering everyone. I can't fly because I don't have data on my iPad.

So that took about 30 hot seconds to fix. So I get a SIM card from Ting. I put it in the iPad. It's $6 a month for the line. I never pay for phone minutes. I never pay for text messages. And for the once a week this thing has to hit the cellular tower to authorize a login token, I will pay for the, you know, I mean, I'm nothing, basically.

And if I was a new customer, if I went to linux.ting.com and I signed up for a device like this, just like a tablet that has like an LTE connection as a backup, you probably wouldn't even use that first $25 for a couple of months. It's just $6 for the line and then your usage on top of that. And there's no contract, nor the termination fee. They got great customer service, CDMA and GSM networks to choose from. I know, it's so cool.

And they have great tech tips and tricks on their blog and often things about giveaways and whatnot. So check it out. Go to linux.ting.com. You can find a device for a great value if you just want to get the SIM like I did. When I got the iPad ages ago, I was like, well, I want GPS. I don't really care about the cellular,

but I do want GPS. And so I thought, well, maybe, all right, fine, I'll do it. And then for years, since I've put that, it just sat there, never with a SIM card in it. And then just a week ago, I'm like, you know what? I got a reason. It's six bucks a month. Why not?

Now you're glad you got it. I really am. Yeah, it's so awesome. You can just anything like that, from your main phone to just a little device you want to add data to. Linux.ting.com and a big thanks to Ting for sponsoring the Unplugged program. The Unplugged program. The Unplugged program.

Linux.ting.com. I don't really have much to say about this next story. Other than what does kind of jump out at me is the register is like bipolar about its coverage when it comes to Linus. It's either Linus is the devil bringing rage to the internet and scaring people away from open source development because of his toad or he's Santa Claus and he's saving Christmas.

That's literally the range for the register. So this headline reads, Linus Torvalds releases the biggest ever Linux 4.9 and then saves Christmas. Now wasn't it just a few headlines ago where Linus was the worst human being on the internet? That's just what Christmas needed. Yeah. So Linus writes, I'm pretty sure this is the biggest release we've ever had.

It's definitely at least the biggest in the number of commits. This was written on the kernel mailing list. Other notable additions this time around include support for the Raspberry Pi Zero. Hey. Hey-o, plus another few 28-arm devices. There's also things in there that you might be interested in, like the AMD GPU support to enable the use of virtual displays from GPUs. Hey-o.

William, we might be hearing your activation key from time to time. I think that might have been you. No. Anyways, here's the thing, Wes. You got to hustle in because Torvalds has also confirmed a small change to the development process starting with Linux 4.10 in the form of a shorter than usual merge window. Get ready, Wes. His kernel patch has got to get in. The usual two-week merge window will close on December 25th, which might look familiar, as Linus says, because it is Christmas Day.

And so if you're originally from Finland like he is, it's the day you relax. Well, so might as well just finish up that. Yeah, I guess, right? So, because the real party, he says, happens on Christmas Eve. I will certainly stop pulling on the 23rd at the latest. So you better hurry up if you want to get your patches in. Hurry up, Wes, because

Christmas is going to affect the Emerge window. Now, this is the story I'm actually, I mean, yay, new kernel. This is the story I'm more excited about because it really actually represents a four-year completion of a goal for Crossover. And it's a major milestone for wine. Crossover has launched version 16. Crossover is the CodeWeaver's product that gives you a graphical front end to set up Windows applications with guides and scripts and downloaders. It's a great product.

And it uses bottles to isolate them so you don't break your different wine environments for different apps. And they have finally, I know you might think this is funny, but they've finally, after four years, nailed Office 2013 support, or at least damn close. And I suppose if I was going to want to run Office on Linux,

2013 probably. That'd be fine. That was the last good one. So that's fine. And the other big note here is it's shipping Wine 2.0, which is I don't think technically out, right? I thought Wine 2.0 was still in development, but they're shipping Wine – they're using a Wine 2.0 code base. Now, this makes sense because a lot of what CodeWeavers does goes upstream to the Wine project. So they would be very likely maybe running on stuff that hasn't been accepted upstream yet. They say they're also going to be focusing on Android support and DirectX

11. Yeah, Android support and DirectX 11 in 2017. It's pretty cool to see this development all going upstream and getting better. Thankfully, I don't have occasion to use Wine that much in my day-to-day life.

But if you have office needs, support, that kind of stuff, interop, this goes a long way. Where does Crossover go once most of what we need to do is online? Like if I can now run QuickBooks online, I don't really need Crossover as much to use QuickBooks on the desktop. And if I can now run Office in the web and Office on Google stuff do I really need to go bend over backwards to run Office 2013? I mean absolutely in the business that's still got to be a use case but it feels like that's going to be yeah so that must be why they're

switching to Android right I remember when they announced a Mac version I was like well shit there goes the Linux version it's going to go down but really they've done a great job. They've kept it really feature complete on both sides. If anybody could do it, it's probably them. But it does seem interesting that they're moving to Android.

And does that mean Windows applications on Android? I can't imagine that would be pleasant. What Android device are they thinking of where that's a thing you need? A tablet? A laptop? Is it just for – you know, I actually – I'll speculate a little bit here.

I bet you the way they do this is with a strict focus on certain games. So they're going to bring a couple of Windows favorites to Android and you'll download – it will be powered by crossover I bet. It might even just be a completely wrapped solution. So they'll have the individual app in the Play Store, and it's using a Crossover backend to run. Huh? How about that? That would be actually kind of interesting. I think that was Chris's first 2017 prediction.

Yeah. I think I just did that. You heard it here first, Linux Unplugged. Yeah, just wasted a good one for the predictions episode. And then this one would be interesting for those of you running CoreOS up on DigitalOcean or on your own. We don't know anyone who does that. No, no, no.

This is huge though. CoreOS developer Alex has announced that they're switching the name to Container Linux. No more CoreOS. Container Linux by CoreOS. Yeah, they're doing a little Plasma desktop by KDE on us here. The decision

to rename the OS was made earlier during the Tectonic Summit 2016 event presented by Core OS. And they also unveiled the next-gen Core OS Tectonic Enterprise Kubernetes solution for deploying and managing containers at scale. Does that mean anything to you? I mean, if you have a lot of containers,

you gotta have something that tells them what to do. Yeah. And they should be running on Container Linux. Container Linux. It's hard to say still. I'm not used to it. All right. I'm not huge on the name.

But it definitely, it's like calling it what it does. And so, hey, what's the best Linux to run containers on? I hear containers are the big thing. Well, there's Container Linux. Container Linux, of course. Just use it. Sounds like they're trying to copy Noah's Nick. It's like Kernel Linux, Container Linux.

Yeah, maybe. Maybe. I wonder how this is going to affect people that are running CoreOS now because CoreOS kind of gets these rolling updates. Will you just be one day running Contain container Linux and no longer CoreOS? Like all the branding will just get refreshed? Let's see. OS release just changes. I mean, it's just branding.

Yeah. Yeah. Yeah. I know. I agree. It's an interesting change. It's an obvious try to play for... I feel like there's been a lot of renames, rebrandings recently.

TrueOS from PCBSD. Yeah. Yeah. 2016 has been rough for a lot of people, man. It's been a rough year for a lot of people. There's nobody in the lug there that's running CoreOS, right? Nobody, I would imagine. I've played with it.

I've checked it out. They've got some of the actually most interesting things about container Linux. I see. Is the technology they've open sourced to make Container Linux possible. etcd and some of their other... Yes, definitely. Yeah, that's where some of the cool stuff is. They've also done some interesting work in trying to use like Secure Boot or TPMs to help.

So if you are deploying your own data center that you can verify that only your software is running on your hardware, which I think that's interesting. Evolve OS to Solace. But we don't mention that one no we don't we don't mention evolve os it's a pretty good name though although you know what i think solace is growing on me more yeah me too i actually think i like solace better so i think long term i think it was a better play ikey uh you know speaking of solace we should talk about that we actually that's that's on the docket that is really what i'm looking forward to the most

today because there is some ridiculous shenanigans going down and I'm not going to have it. I'm not going to have this turn into some sort of community that's productive and with users benefiting. That's not good for the show. That's what we have to think about here. So let's take a moment before we go into that next subject. I was just mentioning DigitalOcean.

DigitalOcean.com. Go there and use our promo code D1Plugged after you create an account and then apply this to your balance and you get a $10 credit. CoreOS, which is now Container Linux, up on DigitalOcean. This was one of the first times I really got insight to after DigitalOcean became a sponsor and how they work upstream with a project and I

was like really impressed by that. Obviously, the other things that really get me about DigitalOcean is how fast you can spin up a Linux server on demand when you want one, how great their interface is to manage all of this, and the solidness in which the platform has performed for me now that I've been using it for years. That's another thing about it that I don't really mention in these reads, but if you're new to it, you want to know how reliable is it.

But they're not a newcomer. They've been doing this a while now. They've gotten damn good at it. They were one of the first, and it really has paid off for them. They were one of the first early-ons to invest all SSDs back when – I mean this is years ago now when SSDs were extremely expensive, especially in the enterprise. And they still are in the enterprise. But DigitalOcean realized that this is a key part to the performance value. And so when I say you can get a droplet for $5 a month, that's all SSD-backed.

All of them have 40-gigabit e-connections to the hypervisor. They've got fast processors. And you can also price a rig for $0.03 an hour, which is 2 gigs of RAM and multiple cores. And it's just phenomenal how fast it all is when you combine that disk I.O. with that network connection, that great interface, and the API, which allows you to do so many things from your phone or the command line that just make you more productive. I always joke about this, but,, but there's things I can do. I just go to a tab and drop down a Guake terminal, and I can issue commands that start and stop servers that make streams connect all over the place. And I can do that from any of my machines that I have Guake on and an IRC client.

And it's so awesome. Just right at your fingertips. It's so much faster than going to a web page and logging in. It's just I hit the tilde button, and I issueC client. And it's so awesome. Just right at your fingertips. It's so much faster than going to a web page and logging in. It's just, I hit the tilde button and I issue a command. And if you're in the chat room before I show, you see me issue those commands. Those are spinning up DigitalOcean droplets on the back end. And then they're kicking off scripts connecting to the endpoints that we're streaming to using their API.

And I use it multiple times a day. It's so brilliant. And then you combine that with the speed, the reliability, the great price, and all of the fantastic documentation, and the fact that they stay up to date with the latest and greatest Linux distributions and FreeBSD. And then they have block storage on top of it all, up to 16 terabytes of SSD back storage, which is nuts. And if you just want to even just mess around at all with LVM or ZFS

or any kind of disk orchestration system. Go make yourself a nice big array, play with it for a couple hours. Yeah, man. Yeah. I was watching a demo of a system that had essentially, to the system, it appeared to be 30 different block devices. And that's for pennies on the hour,

that's a system that you couldn't replicate even at some of the most high-end IT shops. I can't afford one of those right now. No. No, it's been so awesome to be able to go there and then play with the latest and greatest software too. And if you're a total beginner, you can deploy an entire application stack with Linux installed and all the stuff you want. And if you're messing around with containers, there's really no better option out there. DigitalOcean.com.

Use our promo code D-O-N-P-L-U-G-D. It's one word. You apply it to your account. You get a $10 credit. DigitalOcean.com. Use our promo code DEOUNPLUGGED. It's one word. You apply it to your account. You get a $10 credit. DigitalOcean.com. And thanks to DigitalOcean for sponsoring the unplugged shenanigans. So I saw this Google Plus post from this wimpy character. Oh, I'm sorry.

I think he goes by Martin on Google Plus. He says, I'm extremely pleased that Solas and Ubuntu Mate are deepening the collaboration. So much so that I said these words in a recent This Week in Solace post. Ubuntu Mate have been working to – oh, I should have Wimpy read this in his voice. Why am I reading this? Wimpy, do you have this post handy enough that you could read it for us in your own Wimpy brand? I could do it in a few seconds. Yeah, that would be good because I could – why would I?

Why should I if I could have it in your own voice? This is a great – and really it's a great thing that's happening here. I've been teasing the whole episode. But that's because we're excited. I am very excited about this. And I think that the choice is right on both accounts. I will take this moment to pull up this week in Solus that Wimpy is referencing so we can show the pretty images on the stream. This is great.

Okay. All right. images on the stream this is okay all right so the words that i said in this week in solace to josh strobel were ubuntu mate have been looking to replace the mate menu which was forked from mint menu for some time it's clear the brisk menu despite being in the very early stages of development is the project that can achieve that given the quality of work coming out of the solace project yeah wow very complimentary towards Solus. And interesting that this is a departure from a menu that has its heritage in Linux Mint, something that's, when I saw it in Ubuntu Mate, I actually was glad to see it.

Because I like that menu a lot. It's just, it does start, it is starting to feel old and slow. But I don't understand what's happening it's been slow i don't i i thought uh this this seems like cats and dogs to me to a to a to a certain degree because solace is is uh i don't know i this feels like an odd couple pairing to me for some reason i think it's fantastic you know a bunch of monte had come from kind of a what was before what we thought we could continue with. Solus kind of broke out on their own and does things their own way, their own style.

That's kind of the shtick. Shall I give you the back story here? And also, let's just touch on this. Although I posted these things about Solus, this isn't the only collaboration that's happening with Ubuntu Mate at the moment. The only collaboration that's happening with Ubuntu Mate at the moment, we've got Daniel Foray, who's just completed doing some icon work for us. Yeah, Daniel, not to interrupt, but Daniel is from Elementary OS, and Daniel did work on, it's like the humanity icon theme originally. Right, that's his history, and so now he's working with the Mate project, just for people that don't know who Daniel is. Go ahead, Mimpy. So I'll explain a bit of the backstory with Solus first. So behind the scenes, what people don't see is that myself and IK are chatting fairly frequently about stuff.

And we have collaborated on things in the past. What was started here by IK, we could have silently just let happen in the usual way that we get on with stuff. But recently, there's been an uptick of reporting in the online press about the divisions within the Linux community and projects working in silos on separate stuff and not collaborating. And that isn't actually how things work at all. stuff and not collaborating and that isn't actually how things work at all so i decided to use some of these collaborations that are happening recently to sort of counter uh what we've been seeing in the press recently nice so to point to that when ike announced that he was going to make a mate spin of solus i contacted ike and said here's all of the things I did in Ubuntu Mate to change the branding. And this is probably stuff that you'll want to do.

And I also watched his hackathon six-hour video fest thing to see him actually do that initial enablement of Mate and sent him some feedback based on what I'd seen. And I've also had some conversations with Josh as well around that. And Aiki and I have also had some conversations about, so Aiki is also submitting some pull requests to the Marte desktop to make notifications interactive. That was a recent thing that Aiki did. And nobody knows about that because we didn't publicize it. That's huge. And nobody knows about that because we didn't publicize it.

That's huge. And, you know, Ike and I have talked about some other areas where Ike could contribute to the Marte desktop where he has specific skills and know-how that the Ubuntu Mate project is generating to fund these people in the community that are capable of doing good work that benefits Mate or Ubuntu Mate. So if you want good quality code and a modern contemporary menu that's well implemented, you go to the guy that writes highly optimized code that understands the traditional desktop metaphor so that's ikey at solus and if you want artwork the standout project in the open source community for aesthetics and design is elementary so that's who you go and talk to so ubuntu mate is funding both those initiatives through our crowdfunding to improve Marte and Ubuntu Marte, but also ensure that some money goes to those projects as well. So they've got some money coming in to improve their projects

and distributions as well. So that's the whole thing. So I think, Ikey, that's me doing the complicated talky talky bit. Yeah. Now, Ikey, which I love everything that Wimpy just said, Ikey, don't you know better? To compete, you're supposed to keep all of your best stuff for yourself and fork everything and relame it and keep it and just bloat the project with all of these own little independent islands of applications that you maintain. Don't you understand how this works? No.

But I mean, even me, I don't fork things. Solus itself is independent and there is a there's an immense value in competition so when we brought out the solos mate edition there was a few things that we had to look at there is the what i'd like to think of the reference implementation of mate which is oddly ironic uh wimpy myself were talking about it earlier a bunch of mate does diverge from stock Mate quite a bit. But it's the Mate when you think about it. So I decided to

compete with them because then you set the bar at a certain level. Even if you're not there at the time, that's what you're aiming for. And in doing so, you start to expose areas of weakness. And for me, immediately, that was the menu. So we were going to be putting out a new

release of solos a new snapshot but the the menu i just couldn't get over my issues with it so i then investigated the uh the mint menu fork which is i believe wimpy did that himself it basically took the existing mint menu and stripped out all the silliness, basically the hooks into package management and things like that. And even with all that done, it runs dog slow. It's terrible. I cannot bring myself to do it because I know what that code looks like. I've seen that over a few years. So it's like I just got to a point, it's like, well, you know,

I've written menus before this can be too difficult uh so i started writing it up and then yeah we agreed that the the best thing to do is you know have that as something that's agnostic because that's something solace is really trying to focus on lately make sure our software is agnostic that everyone benefits from it and in the way that we're doing this so uh a bunch of mate through mb is funding the development of brisk and in turn that money would then be able to fund projects that solace does or solace gets invested in so you kind of have this to use a real corporate term here you have like a virtuous cycle of growth

some synergy is happening that Nice one, Ike. Virtuous synergy. Ike, the other question I have to ask you is, in 2016, we do docs now. Everything's docs. This looks like a menu. Can you explain yourself? Yeah, that's definitely a menu.

So for the Mate desktop, you want something that's basically a little bit traditional. You want something that's nice, but by the same token, you don't want a full-screen launcher. You don't want docs. You don't want dashes. You want to hit your Windows key, sorry, super key. Thank you. And have a menu to come up, and then you can just type.

Or if you want to filter by category, you can do that. But you also want it to be fast. Because when I was testing the Mint menu, when you're searching, there's a visible lag. While it filters the results, it's not acceptable. Not in 2016. Oh, I am excited about this. This is going to be my favorite launcher, I can tell. I've got an i7 Nokia, and it was lagging.

Not acceptable. How? How? There's 16 gigs of RAM. And when Ike and I were discussing this behind the scenes, you know, I explained I'd put as much effort into the Marte menu as I was prepared to invest because, as Ike says, it's spaghetti code. And frankly, the hotspot in the loop that causes all of the lag

in the mint menu, I can't decipher what the heck is going on there. It's just impenetrable. So I was never going to touch that and i've made a number of improvements to fix the way that it works and some integration um and there's some bugs with it in the way that it hooks into comps and things and i simply can't fix that because of the the wrong way things have been implemented there so when i saw this we've investigated other menus and launches along the way but when i saw this i thought i know this is going to be done right and we really need it it's the thing that people keep it so along ubuntu mate's journey there has

been things you can identify as this is the pain point for people and the pain point right now is people want a better advanced menu and here it is yeah so ike's mentioned that we're offering some funding for this and also my other end of the sort of the social contract on this is that i will package brisk for debbie and ubuntu and see that it gets into both the official archives for those awesome um and uh replace the mate menu with brisk menu as soon as brisk menu covers all of the core functionality of mate menu so i really i really will be using brisk soon then uh i i really think this the dock thing needs to go i i i have a few docks but i do you're out of the dock huh i do it more like the unity launcher a persistent sidebar sure um because i don't

like this am i going to have the screen space Because I don't like this, am I going to have the screen space? And I don't like this constant, like, I need to resize my window to make everything show up on the screen. When I can just have a menu that I click and it drops down and it's just not even an issue for me. So, Ike, if I'm a Solus user, how does this work? Is this something that you'll ship eventually in a certain point release? Is this something that just rolls in? Is it something that I could get my hands on soon or do I have to wait a while?

What do I expect as a Solus user for developments like this? So I've basically put in about just under 24 hours into this. I started it over the weekend and then back to work and whatnot. But most of the core of it is actually done in terms of having the category system and having the filter in and having it filter results by relevancy because that was one of the problems we had in the budget menu where we put a hack in to sort things by most used, but now it'll actually sort them by relevancy to your term.

So in terms of pace, I mean, that's going to be out real soon. We'll go through a number of releases. Our first validation one will go into the next Solus Mate ISO, and then we'll continue working with Martin just to make sure, you know, that these are the things that we want in there, things are working as expected. So there are going to be a number of releases now that come out of it, and that's off on GitHub like you'd expect it to be. And checking with Martin is going to be real simple

because I just trust Ike's judgment on this stuff. So what is it written in? C. Oh, okay. Did you expect anything else? No, I was hoping to have a chance for us all to freak out when you said Python. I thought we were all going to have a group like, Oh!

That's what we're replacing. Yeah, I know. That's what we're replacing. I, I know. That's what we're replacing. I know. That's why I was hoping. I thought if they're on the off chance, that would be a good gasping moment. But no, that makes sense. There are other benefits to this menu as well.

So we've had a couple of people that have freaked out on Google Plus basically saying, when you first announced the Solus Mate edition, I knew this would detract resources. Now you're not able to do it. It's like, whoa, whoa, whoa. Calm down a second. In terms of doing this menu, it's already benefited Budgie, which is going to sound really, really strange because it's not

for Budgie, but it's already benefited and fixed two bugs already because I'm doing a clean room implementation without none of the old Rust code and things are actually being done if you buy the book and I'm not using things that shouldn't be used like Valor Rust code. And, you know, things are actually being done, if you like, by the book. And I'm not using things that shouldn't be used, like Valor and Python.

Anyway, so one of the bugs that's already managed to fix in Budgie was the icon task list. It wasn't tracking the startup notifications of applications properly and cancelling that notification. So in Budgie, you'd have this timeout of 30 seconds with a spinny wheel of doucheness as i've come to

call it and it just wouldn't go away so it's like it's managed to fix that and now the run dialogue can properly track when the application is fully launched and then close itself and that's come from doing brisk menu so you know like there's all these little things it is kind of like throwing a stone in a pond you don't quite know where it's going to land and there's all these little things. It is kind of like throwing a stone in a pond. You don't quite know where it's going to land. And there's all these ripples coming out. You don't really know what they're going to affect. But it's already having positive effects,

and it's only like essentially a day old in terms of code. Sounds like things are proceeding very quickly. I think those kinds of projects, too, are the perfect thing for getting more people talking about the distribution. So even if it didn't actually end up fixing a single bug on the main desktop, it still feels like it's good publicity for the project. That seems like a good thing. There is that, too. It's good to raise the profile of what are the emerging distributions, maybe the new guard of distributions for the yeah the next era of desktop linux exactly well put you know also it's important to point out you've

talked about competition ike and i do compete with one another um but and also uh we also both compete with elementary desktop but all three distributions are talking to one another and collaborating with one another. And the competition is healthy because it drives us forward. And where we can share things and where somebody solved a problem and we can all benefit from that, then we're moving together and the landscape is improving for everybody. and the landscape is improving for everybody.

And frankly, I don't really care if Ubuntu Mate is the dominant operating system, because it simply won't be. But what I do care about is making sure that the Mate desktop remains relevant and has a future, and it's initiatives like this that will help ensure that. And through other funding initiatives will also help ensure that we stay current with other technologies that are coming along as well. That's a great point. Yeah. Just making sure that people want to keep using the desktop and a menu like this.

If you're listening, it's a really just nice straightforward drop-down menu that you're going to just recognize immediately. In response to the sort of commands you'd expect. So that's the Brisk menu. We'll have a link in the show notes if you want to check it out. And to the point of publicity for the project, just before we move on, literally every time we have a conversation like this,

I'm like, I've got to get this on a machine somewhere. It's time to try it out. Yeah, yeah, like every time. And I am just, I am so married to the AUR for production purposes, because I sit down here and I install three apps. I'm like, all right, I try them on. This is the one I'm going to try.

This is the one I'm going to talk about today on Linux Action Show. And just being able to bust it out like that super quick is key because sometimes I'm already sitting down on the microphone just playing around on the live stream. That is a hard thing. Are you suggesting you need a dedicated machine on which you can test the two Solus distributions? Yeah, well, that would be true. I think he is saying that. I actually thought it would be a good job for the Apollo, but I had something happen to me. We'll talk at the end of the show.

We'll make that happen for you. So, Chris, AUR, while it was something that was distinctly not possible in the past, and personally myself, I don't want to say that for mainstream users of Solus because we provide binary repos. You can now provide your own binary repos in Solus as well with the new SolBuild tool. You can run it on any Linux distribution that has overlayFS in the kernel. And you can build Solus packages from there. You can also create your own Solus repos using that tool from any Linux distribution. Nice. Okay, that seems like a big deal. I was looking at this week in Solus number 40,

and I saw you, or Josh, whoever wrote this, Josh, I saw a mention in here about SoBuild with a great gif. That got my attention, but it went over my head as to what it is. So you're telling me, what are you telling me exactly? What knowledge bomb are you dropping on me? Because all of a sudden I just got really excited. Okay, so our old build system was Python. You know, we all make mistakes.

There is an initiative in Solus to move away from Python because we basically want compiler safety in all core tooling, which I think is a fair assertion to have. And Python has no performance at all. So I've rewritten in Go. Go is now the tooling language we use in Solus. It's hip, it's sexy, and it gets the job done. So I've rewritten the whole thing in Go and basically reverted a lot of the design mistakes from the past. So now basically you have this base image. If you think about it kind of like in the terms of Docker,

except that this has done this longer than Docker has, if you think about it, you've got a base image, and each image is either unstable or the main repo. That inside has got a very minimal solar system, which is enough to have a compiler in there, basically. So we store those on our server. You select the profile you want. When that downloads, it puts that as the bottom layer,

and it gives you a temporary working route over the top of it. So the bottom layer is never touched. You have this temporary route, which is using overlayFS. It's basically a container now. It used to be just a troop, but it actually uses namespaces as well now. It'll enter that environment and build the package entirely within the environment. Networking is disabled as well, so you can't exit the sandbox. You run as an unprivileged user inside the sandbox as well.

And once the build is done, it'll emit the packages into your current directory and re-own them to your user. So if you invoked it as sudo, it'll take the original sudo user and make them owned as yours. Oh, that's a nice touch. If you run the index command, all of the packages you have in the current directory, it'll emit an ELPackage index file, which is a repository.

The next step of that is creating a tool called BinMan, which is binary manager. I'm not good at naming things, by the way. I'm not sure if you've noticed. It's also a tool in Go to replace our repo management script, which again, is Python. So what kind of workload are they looking at, say,

if I wanted to install, I don't know, is Telegram in the main binary repository? Okay, so an application like Telegram or something like that, what would be the, for the end user, what kind of process would that be for me to do that? Yeah, I mean, to build something like that, what would be the, for the end user, what kind of process would that be for me to do that? Yeah, I mean, to build something like that, like, our package files are called package.yml.

They're YAML syntax. So the very basic things our package have, you have, like, a name, a source, a version, and a release. So you're describing the meta information of the package, and the source is, like, where it comes from. That can be a tarball, or it can be a git source uh we got built in git supports into soul build now and it knows how to do tags branches commits uh sub modules it'll handle all that stuff for you and it'll create a system-wide cache that you can reuse for the builds to speed them up and then you basically have three steps which is set up build and install any one of those can be missing it doesn't matter, but you're

going to have to have some sort of process to say how to make a package. And they kind of map to the configure, make, install of a normal package you would do. But instead of that, they're basically bash scripts inside each one of those steps. So you wouldn't be lost.

What we do is we spice it up by using macros, kind of like RPM, but better. So we'll have these predefined variables, like you'll have the configure, but with a percent sign. That will pass all of our config effects to it instead of just doing a dot slash configure. So you use some of the macros instead, and then, boom, you have a package.

So you can make a package in like 11 or 12 lines. This is pretty neat. What it does as well is it automatically splits that package up for you. If you package a library, and let's say it's got some header files and it's got some package config files, you'll automatically gain a sub-package called devel, and that will automatically depend on the main package as well. If you pass the emul32 option as yes,

it will rebuild the 32-bit version of that library. So you get a 32-bit and a 32-bit developer package on top of that. So, do you mean 32 and a 64? Yeah, so it has built an emule 32 support, so it'll build 64-bit and 32-bit at the same time.

That is pretty nice. And not to diminish, that sounds like a really badass system, which has stuff in it that only stuff that's getting popular now, like overlay OS or overlay FS and YAML files. But honestly, my thought when you described all this is it'd also just be nice if I could just install a Snap package. That would be really nice too. You could, but, I mean, snap packages require containment

and they require containment because of the fact you download them from the internet. If something is in our repositories, it is implicitly trusted, so it does not need containering, right? So we ignore the container part of it because that's not relevant to our discussion. It's about the simplicity and distribution of a package.

In terms of the simplicity and distribution of a package itself, why package and sole build win every single day? Because we've got an advanced macro system. We've got advanced build system built into the different types. We support multi-lib out of the box and naturally. You can pass optimization straight away to the package by setting the optimize flag. We've got advanced macros.

So in terms of the core of a package build r1 is superior to snap the only reason snap quote wins end quote anything is because it's supposed to be distributed in an untrusted fashion and downloaded in an untrusted fashion so it has the container aspect if you take away the container aspect r1 is superior because it's meant for native packaging. The advantage, though, still remaining to Snap might be developer adoption and availability of software for me as an end user.

Sure, but the fact is all of our software for building stuff, for doing what we're calling basically the soulless build machinery, obviously it's all staying open source, but we're making it as agnostic as possible. The reason being is we're sharing that build machinery with anyone who wants to use it.

And at some point, we will also share that with partners. So while they might not want to share their source code with us for us to build their packages, if we share our build machinery with them and set them up, then they can build them themselves without us even having to interfere and provide native packages. And with the repository management tool we're now creating as well, that will automatically maintain the repo, create the Delta updates they need, automatically trim old updates, allow them to have unstable and stable branches.

So with our tooling now going to them, they will be able to. Can I just interject very quickly? Yeah, sure. iKey, some of what you've just described about snaps there, you're a little bit off we'll talk after and i'll fill you in that's fine it's fine okay absolutely fine and also there's some new stuff happening in snaps and i'll give you the inside skinny because you might want to consider supporting some of the new snap features that are coming along and i'll

i'll talk to you about it do you want to talk to talk? You don't want to talk about it on the show? No. Okay. I want to know, though. I'll just say this. Stay tuned, everyone. I'll say this regarding all the stuff that Ike just described is note it down now, folks, because look at it coming to a Linux distribution near you in probably about five years. Well, and we're just talking about this cross-distro work.

Maybe this is one of the next things that should bleed over. It's a VHS versus Bet beat em x thing isn't it like uh mine is the superior technology in terms of building packages right uh in terms of distribution obviously snap wins because we they have a center they have they have everything but they're the common one that everyone wants and everyone is going to you well i'll give you this you got a good track record of being ahead of where the uh the puck is going and so there's probably a lot of stuff. We're not going to win on popularity at all there. Snap has already won.

There is nowhere for us to win. So I'm not, as much as I was originally completely opposed to Snap, there comes a point where you're going to have to say, this is what everyone is going to use. It's not going to be flat back. Snap has already beaten them. Snap is already the thing. I was going to give you a whole flatback pitch, but alright,

okay, I was going to ask if you had a minute to talk about flatback, but that's fine. What about AMP image? Yeah, I was going to say AMP image too. Oh, and there's others that are just not coming to the old noggin right now. Anyway, I'll have a chat with Ike and give him some inside info later.

I like the inside info. That is always good. I'll give you some good source at iKey. No bother. Well, if it's good, then I'll integrate it. Yeah, it sounds like it. It's a position that is not impossible to be moved on, but it has to be worth it.

Well, that's interesting and now makes me pine for Solo S. Maybe next week we should both be running it. Ooh. Alright, we'll talk more about that on the makes me pine for Solo S. Maybe next week we should both be running it. Alright, let's talk more about that on the post show. I want to just briefly, because we went on for a while because I was fascinated by that last topic, and I think rightfully so. So let's take a moment

and let's thank Linux Academy for sponsoring these shenanigans. LinuxAcademy.com slash unplugged is where you go to keep it going. If you enjoy the conversations and the discussions and want to hear more of us, please visit our sponsors LinuxAcademy.com slash unplugged is where you go to keep it going. If you enjoy the conversations and the discussions and want to hear more of us, please visit our sponsors, LinuxAcademy.com slash unplugged, where you can sign up for a seven-day free trial of the Linux Academy platform, where you can learn a heck of a lot more.

In fact, they've just launched two new courses. I just got an email about it today. It's super wicked. Check them out at LinuxAcademy.com slash unplugged. It is the place to learn about Linux and all of the big infrastructure that runs on top of it with hands-on scenario-based labs, instructor mentoring, nice servers that spin up when you need them, that match the courseware you've chosen like distro-wise. You SSH into it to get your work. They got practice exams so you can prepare for certs. And they got tiny little nuggets when you don't have a lot of time.

You just want to go into a single topic as well as schedule availability picking depending on your busy, busy schedule. But probably the biggest feature is the instructor mentoring, full-time human instructors. Also, by the way, I'll mention this. Check out on their blog. Linux Academy is hiring right now for instructor positions and for back-end positions. Linuxacademy.com slash unplugged is where you go to support this show and learn more about the Linux Academy platform. They have new tools to help you promote the hard work you've done at Linux Academy to an employer or a client or a school. And they have team accounts if you're a small business or a big company or an enterprise that wants to have your staff learn up and train up.

Linuxacademy.com slash unplugged. And a big thank you to Linux Academy for sponsoring the Unplugged program. Linuxacademy.com slash unplugged. And go check out the new courseware they launched today. Now, what do you think? I was thinking about Ubuntu Core when I saw this. And I actually, it feels like a competitive response to Ubuntu Core. Google is launching Android of Things.

And they plan to juice it up by integrating it in with all of their Google Play services, back-end cloud computing capabilities. Oh, who would have guessed? Yeah, Android Studio and the Android SDK for development. And they're going to also partner with Intel for the Edison, the NXP Pico, and the Raspberry guys for the Raspberry Pi 3 to get images, as well as Belkin and LIFX and Honeywell and Wink and TP-Link and First Alert. They've also said they're going to work with Google on Weave and integrations with Google Assistant.

So the whole picture here is Android on the IoT device itself, Weave for the communications protocol, and Google Assistant for the command and control. That's the world of Android things. That's quite the ecosystem. Yeah. And I wonder if this is Google's, Google, I know this sounds like it's maybe a little fanboy or something, but I honestly would not be surprised

if they tried Brillo and went, shit, that didn't work. And these things that canonical, this Ubuntu core stuff over here, this looks like a good idea. We're starting to see people experiment with this. Well, how could we jack that up? How can we juice it up?

How do we make it Google-ified? We've got the Play services. We've got the Google compute infrastructure that we could bring into the mix. We've got the Android development that's already – the Android SDK that's already massively deployed and the Android development studio, which has a huge user base. And they come in with a lot of territory advantages. Yeah, a lot of things can be ported right over, already run there. Yeah. a lot of territory advantages. Yeah, a lot of things can be ported right over, already run there, that kind of thing. Does it feel like a response to any particular product to you,

or does it just feel like them still searching for a market? Yeah, it seems kind of like a natural evolution. They want to be here. You know, there's like the article talks about, like there's a whole Apple HomeKit, you know, kind of stuff. And so I can see where they'd want this. And, I mean, Android's their platform, right? And if they have another chance to push the Play Store, to push the reliance on the Google-managed services, it seems like it's something they'd really be interested in doing. I don't know what they mean when they write this in here.

But they say something to the effect of being able to also Google – what do they say? Here it is. I don't even know what this means. This is from TechCrunch. One interesting twist here is that Google will soon enable all the necessary infrastructure to push Google's operating system updates and security fixes to these devices. What does that mean? Does that mean that they've introduced a product as it stands right now that doesn't support getting updates? Is that what that means?

It kind of sounds like it, yeah. Like the updates is a new thing. And you know what would be really great? I mean, they're following the IoT tradition there, I think. Yeah, you know what would be super awesome? If Google come out and just be really clear and plain spoken about what the update policy is for these types of devices

and what the agreements are in place with the OEMs to make sure that you can update these devices. Is this more like Android Wear where Google has a little more control? Or is it like the Android we know and, well, tolerate? And even in Wear where they promised that they would be able to update it directly, LG held back my updates. So – and I have to wait for my Wahe watch or whatever it is to get updates from them. It's the same exact thing I have. And in fact, when I had the LG Watch R, they rolled out a new version of Android Wear that enabled Wi-Fi support, but not for my model.

It had a Wi-Fi chip in there, but they didn't have the right driver in the image, so my model didn't get Wi-Fi support. So the same fragmentation issues in Android handsets are in Android Wear today. So I have very little faith. issues in Android handsets or in Android Wear today. So I have very little faith. And what strikes me about this is one of the core principles of some of the competitors' offerings, like Ubuntu Core, is transactional secure updates. You know, it's like part baked in. It's like the core.

It's the foundation of the product. Am I wrong, Wimpy? But isn't it sort of a foundational aspect of Canonical's IoT approach? Sorry, you're going to have to repeat the question. I was distracted. Updates for Ubuntu Core, is that a transactional update? And wasn't that one of the core aspects of creating a software platform for devices like this is making sure you can reliably update them? Yes, it's correct. And it appears as of right now, the Android of things IOT platform doesn't actually support updating the software and that they're going to roll that out soon, they say. And it seems to me that that's a fundamental design

aspect of an operating system like this. Not like something you bolt on, but something you start with, especially in this space. So to build on your point, when we started this show, I was running NextCloud 10. I've updated the Snap to NextCloud cloud 11 and i found one of the next cloud apps that isn't compatible with next cloud 10 and i could choose to roll back now if i wanted to because that one app doesn't work and in addition to rolling back the app i could also choose to roll back all the data associated with NextCloud as well. So you've got an and or option there. Yeah. This is going to be an interesting thing. We'll see where, I mean, I'm sure Google will have some amount of success.

And it's kind of interesting to see what their, like, focus is. Like, obviously, at least here on this network, on this show, you know, we talk a lot about IoT in terms of security, in terms of security failings, in terms of how it makes Linux look, how maybe it'll make Android look. So it'll be interesting to see what evolves as Google's focus here, right? Is it just ownership? Is it just getting their services going? Or is it doing things in a secure way that's better than what we have now? Yeah. And there could be a legitimate competitive functionality to having the Google Compute Cloud

behind your IoT device. Obviously, these things are pretty low power, low CPU. So if you could do a huge computation job off on Google services and then send the result to the device, that could be really, really competitive. And something that would be harder for smaller players to match. Yeah, at least depending on the pricing. But Google seems to be willing to throw in some of this compute just as a, here you go, here's some free compute. Go ahead, have at it. Anyways, we'll

just watch it. This is nothing today to really balk at, but I do feel like the Ubuntu core solution is getting more and more traction. You know, you look at the Nextcloud box, but there's also a lot of other boxes out there that are running it behind the scenes, and you look at something like this and you think,

Google must have noticed. They must have noticed. That was my thought when I saw it. And we'll have a link in the show notes if you want to read more about their efforts and maybe even how you could jump in. And I guess we say goodbye to Brillo, which still exists in some form, but it's not going to be the solution for us. This show will go on in another form next week. Join us, won't you?

Tuesdays over at jupiterbroadcasting.com slash live or just go right to jblive.tv and get it converted to your local time at jupiterbroadcasting.com slash calendar. Also, you can submit content or open source projects we should talk about, updates, and feedback at linuxactionshow.reddit.com. Also, if you want to watch my biggest Linux switch disaster of all time. Oh, it's embarrassing. It's bad. YouTube.com slash Chris Fisher.

That's my personal channel. That was the last, the latest video I have posted over there, as well as about a dozen other ones, which you can check out. You can follow me on Twitter at Chris LAS. You're at Westpain, aren't you? That's a,

yes, I am. That's a great Twitter handle. Oh, thank you, sir. That's nice. You really nabbed a good one. At Jupiter Signal for the network. And we'll see you right back here next week. And we're out!

I've never installed GNU slash Linux. Get it out of here. Thank you, MumbleRoop. You guys are wonderful. Yeah, yo. Love your faces. Thank you for being here, and thank you for participating, and a special thank you to Ike and Wimpy for doing a bunch of the heavy lifting this week. That was good discussion.

I really liked that. Like you stood about. Snap into a slim cloud. That's funny. That is funny. What's that? What's this? Oh, what about snap into a slim cloud that's funny that is funny what's that what's this oh uh what about what about snap into a next cloud that's not bad either uh snap install next cloud android of things that's also not bad i would have dropped that last topic if i could have got wimpy to talk about it whatever it was on air i would have dropped the android topic like it was hot the the only reason i'm reluctant

is because i'm not sure if it's public information yet and it will be very soon so uh when when when it when it all comes to pass i'll explain it but i'll explain to wakey because it's of relevance to some of the points he was making earlier and he's's a distro maintainer, and we want distro maintainers to think thoughtfully about supporting snaps. Sure, I bet you do. That would be great. But it's a super cool thing. Yeah. And I'll tell you more at some point.

I think it could be the great equalizer for distributions like Solus if it got enough adoption. I mean, you've got to look at it from my angle here, right? From my perspective, I don't have the package of problems. And people have tried to call me out on this before and have failed. From my perspective, Snap is something that came into existence to address the limitations in the whole stop rating systems that created them. And nobody can deny that because every statement you see about snap it was so hard to maintain the debian package i created white

packages so building all the tollens around solace so that i didn't have that problem so it's very hard for me to adopt something based purely on when it it was developed on the premise of a limitation that i don't have so So it really takes some serious convincing. But see, I look at it from the other end. There are certain – I mean if I knew today that anything I ever wanted to install would be a download and a double click, I would switch. Because for me, that time savings – Yeah, because you don't remember the thing we used on Windows. Yeah, I know.

But for me, that time savings – I know but for me that time savings because I am window shopping essentially not for windows but I am window shopping software I'm downloading three or four things and I'm trying them out and so yeah for me like the difference between like if I could grab a snap

of the new Lightworks beta with the new UI I would be trying it out today but because I have to wait for it to land in the different package managers, I'll try it out in a couple of weeks. You know, I'm not as motivated by that point. The news is old. It's just... If only I was going out for coffee with the product manager for Lightworks tomorrow morning.

If only that was a thing. Yeah. Yeah, that would be amazing, really, if you could give them a snap of that. I know. If he was working in the same town as me, wouldn't that be incredible? Yeah, that would be really handy. But I follow what you're saying, too, I know. If he was working in the same town as me, wouldn't that be incredible? Yeah, that would be really handy. But I follow what you're saying too, Ike. I don't really think it's too – there's not at a point where it's hurting you not to have it, so you have time for it.

Well, it would hurt me to have it though. I mean I kind of think we're trying to like shoot a man with a bazooka here because the problem you're talking about is one that definitely exists there is a continuous integration problem with getting new software there are some if if we use windows terms here there are some titles that we'd really like to get access to we want the latest kdenlive we want the latest open shot you know well maybe not open shot but you know there there are some things out there we really want right away. But the way that we're going about it is a model from the early 2000s and late 90s, which was double-click this file and install it on your system. The world has moved on from then, and people are used to using software centers.

You look at Windows, and you have the Windows Store. You go on Android, you have the Android Store. You go on Mac windows and you have the windows store you go on android you have the android store you go on mac os you have a store so i disagree with the premise of double clicking and downloading from a site i know it probably can be integrated into a store afterwards but i'm having a hard time being sold on this okay so we'll talk about this in more detail but with regards to the store store versus download from a vendor website snaps do both and there is a store implementation maintained by canonical but you can replace that with your own store implementation and there is a reference store on github that you can use if you want to do your own back end so both both paradigms are supported there. You're quite correct that one

of the things that Snaps addresses is the complexity of creating Debian packages. Absolutely. And I also agree, because I've looked at the packaging format that you use, Snaps and EOPKGs are very similar in the way that they're constructed. They're both YAML files. They contain a lot of the same stanzas, very similar. But there are some other benefits, and there's some other little things that are coming on stream, and I'll have a chat to you about that and see if I can convince you that it's maybe worth a second consideration

at some point for Solus. I mean, I'm not hating it on the sake for hating it, because at the end of the day, it's not just about me uh it's about the solos users and their access to software but i've got to do it based on technical merit and not the original i like that i like that as well i think it's refreshing yeah it is i also have to think about the maintainability of an upstream now i don't like to say this and i really don't mean this is in any way disrespectful canonical do have a history of dropping projects after a couple of years as a lack of interest i'm not innocent myself i've done the same now i'm kind of concerned to see how snap will play out and as to the longevity of it i do

think it's got more of a future than flat pack because i think flat pack went about it the wrong way bundling of run times and basically overriding the operating system vendor. But it feels like, don't you get the sense that Canonical's I mean, they're pretty committed. Out of all the things they've launched in years, this feels like one that company-wide, there's sort of a line behind. Yeah, I mean, it does. The signs are there. The signs are positive.

And I am starting to come around to it more, but I just need that little something to push me over the edge. I'm quite happy to admit when I'm wrong. And when that does happen, I like to be wrong. Imagine yourself at the edge of a cliff with a fluffy cushion beneath and there is me with a tickling stick gently coaxing you to the edge of the cliff. Come on.

Come on. I'm fine. Well, I mean, once I am pushed over, and at this point it's become an inevitable. I'm just doing the last childish resistance. Yeah, well, I mean, it's natural. It's natural. I like to hear about the thought that goes in behind the distro, too. That's nice. And there's some good adversarial process going on. I'm quite happy to stick it in the top for a second.

Mikey and I, half an hour ago, were singing the chorus from the same song sheet, and now you can see us slightly offset from one another, slightly grating. So this competitive disagreement can happen cordially, and it's fine. Yeah. I wish that some of the online press could maybe hear this discussion that Ikea and I are having now where we're agreeing to disagree to some extent, but we're prepared to talk about it a bit more after the fact. saying that open source projects are um fiercely you know competing with one another at the expense of collaboration because it's simply not true um that there is so much collaboration that silently

goes on in the background between large organizations and small projects and the commentators simply don't appreciate that it's happening so hopefully this evening or this afternoon or this morning however you're listening to this we've pulled back the curtain on that a little just a little bit and let you know that actually when you say these desktop environments are not working together and are not pulling in the same direction we really are because the one thing we all want is quality linux desktops to emerge to build a user bait and pull more um users and gravity to to our project my sense is it'll have to be a sustained campaign of of continually highlighting in a way that is genuine and also interesting.

Because I feel like that narrative comes from the need to write a quippy register headline or something like that. That's where some of that narrative comes from is the way that the model of that business is so click-driven. We all need each other at the end of the day. As much as we might all compete with each other, we need each other to be doing well, because then it keeps us on our best form. It forces us to look at ourselves, which sometimes is very hard. You don't want to look at yourself, but it forces us to.

And we have to keep improving. And once we do that, whoever you're competing with, they've got to do the same thing as well. Now, some people are going to look back on this conversation, and in a couple of weeks, someone will join on the Solus channel and say, I heard you, you know, you was on Linux Unplugged, you was very angry. Wait, what? You know, because people don't understand this.

You know, we have to disagree. We can't agree on everything. There are things we are going to agree on, but if we didn't disagree, then everyone gets a sub-power watered-down product with a unified vision. We need those competing things. We need those things. You need to create friction to create fire. And that's what we're doing.

This is going to sound very odd, but we all rub against each other in slightly the wrong ways. When it creates a fire, we need to move Linux forward. Yeah, we all think we're right. We all think that our direction is the right one, but we're all building on common technologies. Right. That's the thing. That is really the core thing. That's why it's so – it is so funny that – I think it's lazy too.

It's really easy to see a new application announced or a new desktop environment announced or a new distribution announced, and it's super easy to go, oh, well, look at us just spinning our wheels again and not taking the time to go, well, what are they doing that's new and actually interesting? Because a lot of times, okay, legitimately a lot of times there are people that are just sort of maybe they're rage forking, and that has contributed to this as well. But a lot of times there's real actual good work happening and it's not just about force it's about sharing code it's about working together it's about building on common technologies and that's not the discussion and the problem is is that every time there is a fork it's often portrayed as a rage fork without

necessarily understanding the motivations behind doing that you know marty was perceived as that you know when that was forked from GNOME 2 now it was forked out of frustration but but it was also forked to preserve a desktop paradigm that was well thought out and well tested and I would like to see that desktop paradigm um preserved yeah Ike you also talked about you know when we see other desktops doing things and it drives us forward um when i saw daniel in the summer and he showed me what at that time was an alpha or a beta of elementary and the work they'd done on high dpi i realized just how far behind Marte was in respect to high DPI because they really stepped on considerably,

considering that they were based on a desktop environment that didn't have all of the high DPI stuff at the outset, and they've had to implement it, and Marte is coming from the same position, unlike IKey, who started in a position of sort of kind of outset and they've had to implement it and marta is coming from the same position unlike ikey who started in a position of sort of kind of having a lot of that stuff in the in place in the first place so i realize that in order to keep up with elementary high dpi is a top priority for for the marta desktop oh yeah preach it i'm quite happy to acknowledge you know that elementary's got the

superior solution there and so is um budgie right now but you know we'll we'll catch up and we'll catch up because i can learn we can learn from budgie and uh pantheon yeah well i love you cheated um yeah one last confession to actually add there speaking of budgie speaking of desktop and speaking of brisk so one recommendation i would like to make to any budding developer who's listening to us now i would say that i'm quite an experienced software engineer by now you know i've been doing this for a while uh when i started budgie there was a lot of things i could get away with because i was making it for solos and all of that every single one of those decisions has come back to bite me on the ass in the most horrific and

splendid of ways that's a very honest review i bet it has with the brisk menu i started out knowing almost immediately that this is going to be something that not just Solus is going to be using this. I can't get away with cutting corners here. And as a result, I'm doing it properly. And I had to go digging through the source code of GCK for some very, very badly documented functions that claim to get and instead do and send and then return so it forced me to rethink how i was doing things and to truly investigate and that's how buddhi benefited because i actually went and did it properly because i knew it would have to be something that would stand up not just on solos where i can get away with including it but

everywhere so if you start with that mindset then everybody benefits and you create better software yeah and the other way that um solace and budgie will benefit from this is that ubuntu mate are embarking on a multi-month funding of this development work and and that money IKEE can use to invest into whatever initiatives are important to the Solus project. Alt-Tab, please fix it. Oh, I don't remember

which machine I have set aside, but I do think I have a machine I could put Solus on. I think something happened to the Apollo, though. I don't remember what, but both my XPS and my Apollo are not currently booting. That is some bad luck. What did you do to it? I think the Apollo was the Fedora 25 install.

That broke. I think that's just a software thing, but I can't remember. And then the XPS, I think my little M.Dot drive died on me, which just drives me crazy because that's like the second one I've lost now. But the XPS – You should be making notes as to what manufacturers are behind these drives. Yeah, you're right. We need notes for a rant later. I'm a little disappointed though because I feel like the XPS 13 with that infinity display would just make such a beautiful computer to run Solus on.

I would really like that. I'm just so sick of dealing with broken drives. Yeah, it's so frustrating. With what you're sick of dealing with? I just have machines like I had a not only did I have the M.Di in the XPS and now for whatever reason

the Apollo isn't booting, but also my OG, my first bonobo won't turn on now they and it just dies even if i remove the battery so i've actually had three laptops die on me recently that's tragic man yeah i know i am son so if if you've got dying laptops what sort of laptop is there a gap for in your laptop requirements right now? The range. I mean, the actual only laptop I have that I can reload right now, ironically, is the Librem. That's the one machine.

Hey, it's finally coming in useful. Took a while. So that's my, that's my, and then I have the MacBook running Arch, which has been my, which has become my daily driver now, though, which is not ideal because the battery life is really horrendous on that thing. And the other thing about the MacBook Pro that's really frustrating is just browsing the web causes the fans to go in that thing. And under macOS, it's silent. And then so under Linux, having the fans run all the time, but it still feels warm to the touch. It's all off. It's all off. Just slowly melting.

Yeah, it feels like it. Like I'm slowly killing my macbook uh so yeah but it's a quad core with a terabyte and a beautiful screen yeah so it's my kind of my go-to like i'm going to be doing some work and i want a 15 inch screen right and you can plug in describe describe the laptop that you want right now it ranges i mean it really does it depends on the job i job. I think there's a spot for something like the XPS, but I was kind of, after talking to you about the XPS 15, I was also kind of thinking about maybe making that the next Linux box. I don't know. I'm all over the place, really. Because I could make use out of a really high-end machine doing OBS and shenanigans like

that, or I could make use out of a little low-end machine doing show prep and web browsing. So it's sort of... And what about something to test distros on? Have you got a machine to do that, or is that a machine that you need? That's the Librem right now, and the downside to that, and it's nice, because I can test Intel drivers, but like with Fedora, and

often there's little issues with the NVIDIA driver that I just don't know. I currently don't have a machine to test on. You don't have one. Because that was my Bonobo. But now the Bonobo don't boot. Bonobo no boot. Yeah.

Yeah. But I was able to play on with Wayland. So it sort of worked out. Yeah. It's hard because you kind of need a diverse set of chipsets. Yeah. Yeah. Kind of need a lot.

Because I – yeah, exactly. I don't really have anything in the very high-end performance range right now. You don't have any AMD. No, but most things in the mid-range are usually powerful enough these days. I'm all over the place, Wimpy. You can't nail me down, I know. Yeah, this is a tricky conversation. Well, because basically there's a spot in the lineup for all types of hardware.

Because if you think about it in terms of review, both are valid. I would like to review on the high end of hardware and on the low end. For my daily use myself, I like something as powerful as possible. I would say the Apollo probably falls a little short as a two-core. You want four?

Yeah, but then I also don't want it to be... Well, you want battery life. Yeah. And you don't want it to be loud. Yeah. I'm all over the place i've um i still use my apollo as my lightweight you know go go out and use computer but when i need to travel and i need a workhorse to do work on um but i don't want something heavy i've got a

fully full tilt um x 13. Oh, 13. Yeah, because that's small enough and light enough that I can stick it in my backpack. Oh, for sure. And pack my pants and socks and my toothbrush and toothpaste and all the rest of it. Yep, yep, yep. But if I want to, you know, just go down to the pub on a Saturday afternoon and do some work, then I take the Apollo because that's got a larger screen and it's lighter weight. And yeah, the battery actually lasts longer

because the XPS 13 is a UHD display. So it burns through the battery a bit faster. So, yeah. So, you know, it's difficult. So I found that I need about five laptops. Yeah, that's kind of what I was... Find all of the I need about five laptops. Yeah, that's kind of what I was... Find all of the niches that I need. Yeah, that's kind of what...

I hate to say it, but like if I'm doing reviews and sometimes I'm doing really high-end production stuff and sometimes I'm just browsing the web, reading emails, and those are different machines. So my sort of test machine is the XPS 15 because it's got both Intel embedded and NVIDIA discrete graphics. So you get to test both on that, and it's got a 4K screen and all the rest.

That's why I was kind of looking at it. You can do more than 16 gigs of RAM too, right? You can do up to like 32. I've got 32 in mine, yeah. That's it in the sweet spot for me. And how is it thinness and weight-wise? It's great. I've got it just on the floor here.

Yeah, it's not heavy. It's about the same thickness, slightly thicker than the XPS 13, only slightly. It's got the same keyboard as the XPS 13, exactly the same. It's just got a wider palm rest area. Is the trackpad bigger? Is the trackpad? No, that's a good question. I haven't measured it, but it's big. Oh, bigger is the track it's the same no it's uh that's a good question i haven't measured it but it's it's big oh no it's much but actually it's much bigger

thinking about it yeah it's much bigger yeah yeah i would like that to be bigger yeah and you know the other nice thing about that is i'm pretty sure dell still accepts bitcoin so that's very yeah i bought mine with bitcoin wimpy are you playing it all with libre vault or what's uh what's your latest experiments in syncing not not recently so i was using it oh i don't know some months ago three four months ago i did a little test but what i was using it for i've switched to a different tool so i don't have it right now uh so have i heard of the tool you switched to? Probably not. It's called Yadim. Y-A-D-M.

I haven't either. It's yet another dot manager. So it's a dot file manager. And this is what I use to manage my whole roaming profile thing now. And Wes, you want this in your life. It's amazing. That does sound amazing. you want this in your life, it's amazing.

That does sound amazing. I feel like we had, that might have been, we had a.file manager on our PIX runner-up list for the last recently. I don't think we put it in the show. So it works like Git. Instead of typing Git, you type Yadim and then all of the Git commands follow as you would expect.

So it kind of just wraps it. Yeah, but you can, it's got crypting in there as well. So, for example, anything sensitive that you want in your.files, you can encrypt those when you store them in the Git repository and then decrypt them when you pull them out. This is the dream. So I've got things like my VPN keys and things are in my.file Git repository,

but all of that stuff's encrypted. So what are you not syncing the config of? Is there any application that's caused too much trouble or it's not worth it? Is there anything you just don't sync? No, pretty much everything that I rely on is synced, either the.file that it uses directly or I export certain G setting branches for applications

and then have a script. So in my Yadim.files, I also have a scripts directory export certain G setting branches for applications, and then have a script. So in my Yadim.files, I also have a scripts directory. And one of those is to restore those decomp dumps so that when I check out that, I can run a script and everything just comes back.