LINUX Unplugged - Episode 234: Behind Canonical’s Curtain | LUP 234

Episode Date: January 31, 2018

How does a SNAP really get created & why would companies like Microsoft, JetBrains & Plex want to be involved? We’re live from Canonical’s Seattle Snapcraft Summit with the inside scoop of where things are going.Plus the major concerns we have about the future of Gnome Shell, why the Client Side Decoration debate is hot again & how Wayland is putting pressure on all the things.

We are in an alternate location today. We are in a backup location. Super secret. Location A was right outside where they have this secret room that we can go into, but we're not allowed to tell you what is said in there, and we're not allowed to record anything. Oh. But outside that room, we set up camp where they have these tables, and that's where we did Coda Radio Live yesterday. Right.

Yeah. So we were at the Hilton in downtown Seattle at the Sprint event. And the problem was, is next to the sprint event, the room over, so the doors are side by side, they're two different ballrooms. And the door over was a practical marketing strategy training session. Oh boy. And so all of these people who have been sent to practical marketing training by their companies were coming out and seeing these two guys in apparently servant clothing, companies, we're coming out and seeing these two guys in apparently servant clothing asking us questions like, what's the bathroom code? Where do I get food? My favorite one is a lady interrupted code of radio to ask me if they have non-dairy creamer.

Like you're sitting here doing the live show. I have this headset on, which is the headset like the ESPN broadcasters are wearing, right? We're coming into, we have both, Noah and I are sitting there talking. We're loud because we're on the air. We're looking at our laptops. We have a mixer there. There's wires everywhere. So clearly you're there to take live questions about the hotel.

Non-dairy creamer. So today we're like, boy, we should really find an alternate location. We should really, because that didn't work well because we got interrupted a couple of times. And we arrive at, outside the secret room again to discover a bit of, I would say, hushed chaos. Like the room was tense.

And we're like, what's going on? Why is the room so tense? Did somebody fart? And so we're walking around. And I'm trying to get on the Wi-Fi. And I'm getting frustrated. And I give my laptop to Noah. I say, Brown Bear, will you get me on the Wi-Fi?

He's got the special touch. Yeah, I got a touch base with JB1. And as I'm walking around, a wimpy comes around the corner and says, oh, the Wi-Fi's down. What's that? Oh, no. The network's down here at the switch. The switch lost its config. And the APs that are wired into the switch are essentially without any network

connectivity right now. You get a little tipsy with the switch. You misplace it and you leave it behind at the bar. And it was this weird experience. So, right before that, I don't know if you were there for this happened, but I was just walking away. I was just going into the restroom or coming out of the restroom. And this guy, this weird old guy, and you all know what I'm talking about, right? Like he has his phone in his hand. What you're going to look like in about 45 years.

Yeah. But this guy, it's like you're not quite sure if he knows that the screen is on or off, and he's not quite sure what those icons are. And he's mumbling about, I can't get it connected. Are you on the Wi-Fi? That's when he comes up to Noah. As I'm walking away here, hey, are you on the Wi-Fi? Wi-Fi. So I said no.

And then I was afraid he was going to want me to help him with this phone or maybe want to be my friend or something else terrible like that. You don't have time for this. And so I was looking at him. I just kind of brushed it off. And then he mutters. He goes, we lost the switch configuration. So the whole Wi-Fi is down. And I was like, dude, Wi-Fi does not come from switches.

Wi-Fi comes from access points. And this further confirms that you are some old guy that doesn't know how to use this technology. Obviously, the Wi-Fi works just fine. I'm not connected. And then Wimpy comes out and says, oh, the Wi-Fi's down. And then it dawns on me that dude is actually the tech from the hotel. He's here to fix it. And he was probably right.

It probably was a lost configuration. So the hotel went into problem-solving mode because they got a big client here, Canonical, who is doing... Rented a whole bunch of rooms and the whole thing. Has industry from all over Washington and local areas here working on things that are all connected to the internet. And this is a big part of why they're here.

And so the hotel is a problem to solve. And I see these, I see all these different discussions going on. I got a little audio, but I might tell the story again. And Linux One plugged a little bit with the audio. But they decide we got to move hotel rooms. Long story short, we decide we got to move hotel rooms. Yeah. And so we're now in the backup location.

That's a big backup last-minute change. Yeah, but now here we have wired. We have wired Ethernet. We've got 300 up, 300 down. So we're sending uncompressed PCM audio for this live event from downtown Seattle. That's amazing. Live from downtown Seattle,

this is Linux Unplugged, episode 234 for January 30th, 2018. Welcome to Linux Unplugged, Jupyter Broadcasting's weekly Linux talk show that is live from a canonical Snap event in a fancy hotel ballroom. My name is Chris. My name is Wes. Hello, Wes. Hello, sir. Nice to see you downtown. So I came to you this week.

You sure did. You're so nice. Well, you know, it's the least I could do when canonical's in town. Yeah, we're going to give you a peek at what it takes to actually make a Snap package. You get a company on board with Snaps. We're at a Snapcraft Summit right now. But we also have some breaking

news 12 minutes before the show went on the air we'll talk about. We have big other community news we will discuss as well as a very frank discussion about GNOME, something I'm trying for a little bit. And we'll end it with my favorite terminal emulator

of all time. Whoa. Yeah. Those are strong words. Whoa. Yeah, yeah. Those are strong words, Chris. I know, I know. And Wes, if that's not enough, Noah's here too. Hello, Brown Bear. Hey, thanks for having me. Yeah, hey, thank you for running the board during the show. Yeah, no problem.

I've actually, I've gotten to, we'll see if we, we'll see if I do a good job, but I got the bottom two of your show now. That's true, yeah, yeah. You are a machine. Yeah, you know what you ought to do? I don't know, man, we should pull someone in for Ask Noah, because Ask Noah's coming up right after this show. We're doing that live from here, too.

Oh, yeah, sure. One of us should ought to crash his show. Oh, yeah, absolutely. Or both of you. Yeah. All right, well, let's also have the Mumble Room crash our show, because every single episode of the Unplugged program is joined by our virtual lug, even when we're alive on location in downtown Seattle. Hello, Mumble Room. Time-appropriate greetings.

Hey. Hello. How's it in downtown Seattle. Hello, Mumble Room. Time-appropriate greetings. Hello. Hello. How's it going? Hello. Hello. Hello. You know the old saying, a watched Mumble Room doesn't boil? I have no eyes on the Mumble Room today, but I am told that we have great.

They're stronger than ever. Yeah, we have a great turnout today. It just takes us being live at a canonical event. So what is this thing that we're at? These sprints is kind of a rough definition well canonical describes it as something similar to old style bar camps and it's an agenda that's totally free form the attendees drive the agenda and that kind of sounds like oh it must be free floating but holy hell if these guys don't hustle their asses off from actually 8 a.m.

Technically, it starts at 9 a.m., but they technically start at a.m., really, because they have a company breakfast where everybody starts networking and setting up expectations for the day at 8 a.m. They start on the schedule at 9 a.m., and they don't finish until 6 p.m. And it is go, go, go, go, go the entire time. So they work with a bunch of different companies that have come in, and what they say from companies from different domains, from GUI to electron-based applications and all kinds of things in between,

robotics included. They all bring these different companies in to work on a common set of frameworks and snap packages. So we are here to observe that. And this event is pretty different than the New York event. It is much, much more of a nose to the grindstone, work really hard, which I thought New York was. Because it's not like they were screwing around in New York. But this is at a whole other level.

Because there's some big names here, which we'll talk about more in a little bit that we've chatted with. And so I think the pressure is really on to produce something. And there's areas here that are off limits to us. But they're still very accommodating. But there's areas, right now, we're not in an area where there's any canonical folks around us because everything they're talking about,

they're being very careful because of the partners involved. Not because they really have a preference, but their partners have a preference. Strategic partnerships. You have to do it the right way. Speaking of strategic partnerships, this will be impacting you and I on TechSnap in the near future. Just before we went on air, now it's like 15, 16 minutes ago, Jim Whitehurst, CEO of Red Hat, just tweeted and announced that Red Hat has acquired CoreOS as part of their Kubernetes strategy. Is that right?

Yeah. Wow. Yeah. Isn't that Kubernetes strategy. Is that right? Yeah. Wow. Yeah. Isn't that huge? That is huge news. Yeah. So Red Hat is going to be acquiring CoreOS, expanding, this is the tweet, expanding their hashtag Kubernetes and hashtag containers leadership.

That's not even really an overstatement there. Like CoreOS is defining some of the core fundamentals about containerization. And they, I think, have sort of prevented the total Docker takeover of containerization to some degree. Yeah, I think it's also been really interesting. And it makes sense that Red Hat would be also interested is that they've really been approaching it from the platform. You know, Ubuntu has their own Kubernetes sort of spin and setup, but there's a lot of people just they're rolling it themselves on whatever infrastructure. CoreOS provided a whole stack. You had a lot of tooling built in

and a whole environment designed with that first. What about the, doesn't some of the cloud first initiative, what is that we talked about last week in TechSnap? Oh, the cloud native computing foundation. Yeah, that's part of the sub part of the Linux foundation. Isn't there some technology in that group that's from CoreOS? Yeah, the Rocket runtime. Right sub-part of the Linux Foundation. Isn't there some technology in that group that's from CoreOS? Yeah, the Rocket runtime.

Right. And now Red Hat owns that. They sure do. And so one of the core container technologies that the Linux Foundation is putting their weight behind is something that Red Hat owns. Yeah, it really gets them a piece of the market. Yeah, that sure does. And that just happened. I is, and that just happened, I mean, I am, the ramifications of that

are going to be really big. We just did, too, sort of perfect timing, go to techsnap.systems, and we just did an introduction to Kubernetes. Episode 353. Thank you. Yeah, episode 353, techsnap.systems slash 353, too many containers,

and it's an introduction to, why are people even talking about Kubernetes? Where did it come from? What is its open source heritage? And why is it getting so much credence and credibility right now in the marketplace? Why has it gone from like zero to 100 miles per hour in the last year and a half? What is all that about? And what does it accomplish? That's explained in 353.

And then if you listen to that, and then you hear this Red Hat news, it accomplish that's explained in 353 and then it's if you listen to that and then you you hear this red hat news it really puts it all in context it's really good timing on that episode i'm just saying it sure it was good there's gonna be a ton more you know as this develops we'll actually see what changes yeah oh man i don't i don't know no i i know that you work a big part in in virtualization That's a huge part of what you do. Much with containers. Are they at the alt of speed? No, not really.

And part of that is a lot of our businesses, we do subcontracting for some very large businesses like Walmart and Sport Clips and stuff like that. But the primary focus of our business is all in a lot of really small businesses. And so for there, if there is containerization, we're the ones that are doing the containering. But for the most part, we're just virtualizing that stuff. It's nice to have tools to manage all that when you've got a lot of them. When you get a chance, check your Discord chat.

Got to scroll down to the bottom. Put like a thumbs up in there for me or something, you know? So Noah has, not only is he providing the connectivity, because he's, well, I guess technically the hotel is providing the connectivity. Right. But the gear he brought, he has whittled this thing down to this incredible tiny package that he can throw in a bag. Yeah, you guys can't see it, but from our perspective,

it's like a mostly empty table with a couple laptops and a magic machine making this whole thing possible. And by the way, the laptops are, right now, I'm simple-telped into a client's computer. So the laptop has nothing to do with an actual show. The laptops are just for our web browsers and our chats. Like, there's nothing, the reason why I bring this up is years past when we've done live events for the last Linux Action Show or this show, like, there's literally been events where we're building entire PCs hours before we go on the air. And it's like this complicated, thick, multi-computer,

three-laptop, huge sound mixer operation. You wouldn't even need two people to carry this setup. It's really remarkable. And it's all because this guy's been playing around with the actual radio. It turns out he becomes a radio professional. All of a sudden, he gets the inside secrets, stuff they don't tell the podcasters. I guess they already figured that out.

Who would have thought? I guess it's something they've had to solve before. Speaking of something you have to solve, the cellular companies in this country have gone crazy. Solve that high bill. Switch to Ting. Go to linux.ting.com. It's a smarter way to do mobile. Your average Ting bill is just $23 per phone per month. It's pay for what you use.

A fair price for however much you talk, however much you text, and however much data you use. More on that in a moment. They have nationwide coverage, no contracts. It's $6 for your line, and then it's pay for what you use. Well, when it's $6 for the line, that enables cellular connectivity at a level that you don't normally think about how you can integrate cellular connectivity. And I'll give you an example. If you go to linux.ting.com, check out their store page. It's like shop. But first go to linux.ting.com. You've got to go there.

$9 SIM card. Sure enough, this maniac across the table from us, Noah's got a couple of these Ting SIM cards in his bag. He's got them right on his belt, right? He just whips them out, ready to go. It's funny you say that, actually. He has them in his bag. He's got them right on his belt, right? He just whips them out, ready to go. It's funny you say that, actually. He has them in his credit card holder. So he's got a wallet credit card holder with one credit card and multiple TINX SIM cards.

It really is how he does it. And so a Canonical employee comes out, and he's from London, and he doesn't have any cellular service. He's got one of these cell phones that he brought a cell phone with him, but he doesn't have a SIM card. He was just a meet and greet the night before the whole event kicked off. We're just sitting around drinking some Mac and Jacks

and he mentions, I kind of wish they had Wi-Fi down here because this phone's dead without Wi-Fi because I don't have a SIM card in there. Noah says, would you like one? The guy looks at him like maybe Noah's a salesman or something. He's like, no really, I just carry these around. You know what, you don't even have to put it at him like maybe Noah's like a salesman or something. What are you offering? He's like, no, really.

I just carry these around. And you know what? You don't even have to put it on. I'll just activate it. It's $6. What do I care? I'll just turn it off after the weekend. So sure enough, he gives the guy the SIM card, activates it on the spot, gets it all going.

The guy comes back, pops it in. The only thing that slowed us down, the only hitch is that he didn't have one of those SIM remover pins. You need those tiny little pins. But actually, we realized we were at a hotel. You just go to the front desk. They'll probably have it. That was a resolvable problem. But the story just relays the flexibility you have with Tink. What other network

could you do? I don't think it'd be possible. And when it's your own device, when you're trying to troubleshoot something, it's that same kind of flexibility. If you wanted to buy a phone, they have a bunch of great devices. If you want to bring a device, they have CDMA and GSM networks to check their BYOD page. Now, I was mentioning the data. So they have a post right now. I've mentioned it on a couple of other shows, but I haven't mentioned it here, and I just want to draw some attention to it. They have a hack your phone plan blog post, which I love because they've essentially written up everything I do as a

Ting user, and it's how to have a smartphone plan under $20. And that's really just because Uncle Sam's cut. Otherwise, it'd be like $16. I mentioned this on Linux Action News. I'm probably going to mention it one more time just because this is it. People ask me all the time, how do you have three Ting phones, and you're paying like around $40 a month.

How are you doing that? Right. Well, this is how I do it. That's the thing. They're even telling you. They want to help you out and help you save money. They're honey badger, dude. Exactly.

So check it out. Linux.ting.com. It is a better way to do mobile. linux.ting.com. Big thank you to Ting for sponsoring the Unplugged program. So I was kind of planning to have somebody join us today for the show to talk about the Snapcraft event, just because why not get it from the horse's mouth. However, they're really busy i was gonna

say sounds like they've got a lot on their plate and so uh these guys are so dedicated to this thing that we are scheduling our interviews after hours tonight and we're going to be conducting interviews into the evening when they're done working because they got a lot of stuff they're trying to accomplish and they're moving and then the other thing the other wrinkle is everybody's moving hotels around because of the connectivity issue. So that added a wrinkle to our plans. But I have been collecting audio snippets here and there and interviewing people, doing sidebar conversations. So we still have tons of stuff to talk about.

And I want to back up just a little bit because we got an email into the show a few weeks ago. And it said, you're talking about Flatpaks. You're talking about Snaps. I have no idea what you're talking about. And I think it's because we've gotten so far into universal packaging. We've gotten so far into this discussion now. Forget that there's some users out there who just haven't used them at all. Yeah. And so Snaps are containerized software packages that are supposedly supposed to be really simple to create and install.

And in modern terms, when you say simple to create, a lot of times what they're talking about is automated. So you can connect it to a GitHub repo, and every time you upload a new build to GitHub, it also generates a Snap package. Some CI server somewhere goes, does a new build, runs tests, publishes it out there. Yeah, and then the Snap software on the client itself

checks in every 24 hours and can auto-update if set. And so it can be a way for, say, Telegram to just constantly be publishing new versions of Telegram, auto-update like you would a smartphone app only. And the risks of doing that are minimal because all of the dependencies, all of the changes are contained to the Snap. The Snap actually mounts as a file system, and it's a runtime in there, and they have these portals out. I think that might be flat-pack terminology, but it's essentially they have these portals out that allow them to get access to certain things i think that right there so

like as a normal user i think right now i'd still be asking all right so i can package them how is this different i already make devs today why would i why would i want to try one of these new certain new types of dependent you know new types of systems right well so let me let me tell you who some of the people i've talked to uh while we've been here to kind of put this in perspective. Because I think that kind of makes the difference. Because you've got to view it from their perspective. Was it last week? Or two weeks ago, we looked at MB and Plex. And I remember one of the things that I didn't expect when I started looking into, well, what does Plex do that is open source?

And I discovered they have just this monstrous amount of open source projects. Their GitHub page is like three or four pages deep of open source projects. You just next, next, next, next, next, and there's still a full page of open source projects. And they target my QNAP. They target my Linux box. They target Windows, Macs, Android devices, all these different NAS devices. All the smart TVs in the world. Anything that can either be a Plex client or a Plex server, they're targeting it.

So they actually have a team that is just packaging up software constantly. This is what this team does. And how do you suppose they do that? They do it with a series of scripts and systems. Different for each different platform. And every time it's time to add another platform, it's this whole, geez, can we do another batch of work? Can we do another platform?

Can we deliver it on a regular basis? What are the esoteric issues? Because this is ARM version this, and this is an x86 version. And then, oh, by the way, on some of these systems, we're shipping something that opens up a web server and listens to incoming connections and has an API. Yeah, right. Yeah, exactly. Which is candidate number one for containerized applications.

It's anything that's taking remote connections. So you can start to understand how a universal packaging format is extremely appealing to a company like Plex that is contemplating, well, geez, imagine if Chris's QNAP that runs Linux actually has Ubuntu on it, could just install Snap packages. And then they could write one package. And the thing about Snaps is you can have one Snap name,

and you can have a 32-bit version and a 64-bit version and an ARM 32-bit version and an ARM 32-bit version and an ARM 64-bit version all published to the Snap Store. And when the user searches Snap Search Plex, they just get the one that matches their... Right, the client picks the architecture you're on and... That also solves another huge problem for them. Yeah.

So Plex is one of the companies that's here. Slack is here this week. Folks from the Azure team are here this week. Electron, specifically the builder side of Electron, so maybe output to snaps from Electron builder. Right from the two-link. .NET team is here. Skype team is here.

And Visual Studio Code team, as well as several, I mean, those are just the companies that I can remember talking to. But there's several other companies here, probably another handful of companies that are here, all talking about how to solve these problems, or this is what it can't do for me yet, or this is what we'd like to see have done with it.

And then Canonical and the people here in these rooms have to sit back and listen to all these people and go, yeah, that's something we could do, or boy, I'm not sure if there's a use case, but maybe we can make a compelling argument. Their stance is, you give me a good enough reason, we'll see if it's possible, but maybe we can make a compelling argument. Their stance is you give me a good enough reason, Plex, and maybe we'll add that because we're in that

phase right now. And that's an important step to solicit input from all the different parties that might use your tooling so you can make the right design decisions. And I feel like it is a competitive advantage that Canonical has for specifically driving this kind of adoption.

So I talked about, like, what does it take to build a Snap? Well, it takes more than having a technical idea and a way of building something. It takes a business calling another business and saying, you have something of value that a user base wants, we'd like to work with you, and we'd like to solve a problem that you're having. And that business goes, we would really like that problem solved. We will send resource X, Y, and Z to location B that you have chosen business, and the business is very good.

We will send personnel A, B, C, and D to this location, and they will resource together. Right. The kind of really boring, everyday details that just needs to happen to make this make progress. One of the few things that bureaucracies are actually particularly good at is this particular thing. And when you're trying to get outreach and adoption from folks like Electron and Unity and Microsoft Azure, you need somebody

that's picking up the phone and calling somebody else on the other end of that phone. And as awesome as open source volunteers and other things can be, sometimes you just need someone whose job it is explicitly to, nope, you'll be there these regular hours and you're making the phone call. And you call it, well, you're on developer outreach. That's what you might call it. You're on developer outreach.

And so you're going to be on developer outreach. Well, what that means is it's your job to get snaps adopted. So go out there and find people that have problems to solve and start solving them so that way we get adoption. But somebody up ahead looking at the 50-foot level

said, you need to drive adoption for this. How we're going to drive adoption for this is by listening to what they need, adding it back into the product, and then working with them to solve the problems. For Plex, that individual is going to walk back pretty much knowing this is going to work for us or not. There is real

tangible value for Plex. We spoke with insiders. We've got commitments or not. The other thing is they're writing real code here, too. You got an idea? Let's write it out. We can't go in the rooms. We can't talk about it, but at the same time like they've invited us to their slack discussion so i'm seeing the things that they're producing oh interesting and uh i you know there i can see things they're experimenting with they might try adding and you see it's uh hey you know i'm just making i'm not

going to give any specific examples but but Bob from Microsoft has this idea of how we could scale this this way. And I just tried this, so check out this branch and give it a go and let me know what you think. And then you see somebody else go, well, yeah, that's a great idea, but what about this dependency over here? And to me, what it feels like a little bit is,

sort of putting it directly, you get a bunch of heavy hitters together in a room for a little while. And when they're out working in the remote field, they're good, they're producing stuff, they have control over their own, they are masters of their own domain. But sometimes for it to really hum,

you've got to get all the heavy hitters in one room, and you've got to connect them with partners that you've created these relationships with, and they've got get all the heavy hitters in one room and you got to connect them with partners that you've created these relationships with and they got to send their heavy hitters. And it, to me, what I've noticed now here and at NYC is it seems like you can get eight or nine months of work done in, in four or five days. Yeah. Right. You have the right incentives and you have synchronous communication, right? All the people you want to talk with are good. Knock down walls for you. they're right next to you. And they're going to breakfast with you.

They're going to dinner with you. And they're at the lunch. That's the whole point of every one of your conversations. Yeah. And, you know, they're breaking off into, like, beers on this night and stuff like that. And you know what happens? They talk about work. In an excited way, right?

That's where you can really have those good ideas. That's what they all have in common, right? They have work. And so same with the folks that they're bringing in from out of town. And so same with the folks that they're bringing in from out of town. I feel like part of this discussion, too, should be that a lot of these are companies or projects that haven't fit into the typical Linux dependency story. They're not part of Fedora. They're not part of a standard Debian distribution.

They're really third-party packages, and we haven't had a consistent answer. And so that's why it's important that we have businesses and major linux stakeholders who are pushing forward on this yeah and uh it doesn't necessarily detract from app image or flat packs either because it i think what it really is doing is it's just defining each of their marketplaces right like this is going to be the area where snaps are especially competitive but i was having this conversation with no be the area where snaps are especially competitive. But I was having this conversation with Noah, an area where I think flat packs are more competitive is when you want to subscribe to an entire repository of available possible software that you could install and update at any future point.

And so as just an example, as flat packs, the number one competitor to snaps, essentially all of the development builds of GNOME software are available as flat packs. In the past, to get development builds of GNOME, like, if you wanted to run the next version of GNOME, you would either have to run some distro that was way out on the edge, but even that would generally be delayed,

or you could get these builder scripts that GNOME had created to build GNOME from source. Right. Then install it. It's not just apt-get. It's like, okay, go figure out everything you need to know to be a GNOME developer just to try this out. And that's one thing when you're building an application.

When you're building in a desktop environment and it's something that's getting frequently updated, that's crazy. Multiple layers of dependencies, multiple applications. And then imagine if you're a GNOME developer and you just want to test against the latest version of G-Edit. What a nightmare. Because you don't need all of GNOME,

but that's how you get it. So what they've done is they've broken out all the development packages of GNOME into individual flatpaks. And so you subscribe to the flatpak development repo, and you can get nightly, you can get different betas, and I'm just giving kind of a rough example, but you can subscribe to the software that you want, and then fromly, you can get different betas. I'm just giving kind of a rough example,

but you can subscribe to the software that you want, and then from that repo, you can install to choose Polari or G-Edit. It's just individual components. Interesting. And they come down as Flatpaks, self-contained. They don't mess with your main GNOME desktop's GTK dependencies. It's all packaged, compartmentalized, contained for you. And I think that's something that Flatpak will continue to be more competitive at

than Snap packages are. Because Snap packs are like delivering an individual application. And Flatpak is capable of delivering an entire repository that could show up as a browsable software menu in GNOME software. Whereas Snap packs show up as individual

Snap entries. It also kind of says, you know, this is still, and we're still really fleshing out what the full problem domain is. What problems need to be solved, and in what approach will you take? Yeah, they're not sitting still. No, right? It's very exciting. They're like, well, we're going to figure out our stuff right now.

So I think, Bashful Fruit, you in the mobile room said that your team in engineering is looking at using Snaps. Tell me about that a little bit. Basically, here. There he is. Hello, Bashful. Hey, hey. How it goes?

Okay. So we are basically looking to use Snaps to package up a bunch of our, I don't know, I'm going to call them enterprise-y apps. It has to do with remote access, things like that. But we have Linux agents, Linux clients, all that other kind of stuff. and on an offhand chance today in a meeting i made a mention about them and now all of a sudden i've got like two or three engineering teams coming over to talk to me about it because they didn't know and now they uh kind of are peaked and interested and looking to go you've been experimenting snapping up a few things haven't you yeah i got a few things i mean up you? Yeah, I got a few things.

I mean, up on the forums and stuff, I just finished Restic. Popey took over the Mumble Snap that I started, mostly just because I got sidetracked. And other than that, yeah, and then we're also looking at how we're going to leverage it in Budgie as well. So lots going on. That's super exciting. Yeah, the Mumble one could be particularly helpful for us because sometimes people get hung up on installing mumble before they can join the show and some people have problems with that so being able to just say install this yeah

oh that'll be great i literally installed it right before the show so yeah you've you've had some you've had some experimentation so you feel like you're comfortable recommending it to your place of work, huh? Yep, definitely. I don't see any problems with it. Any of the little points to be ironed out, they're already on their way. So, I mean, that's exactly why they're doing this get-together. The main thing for me is theming, but they're working on that. Yeah, there's some of the server side stuff too so i know that there's a sort of a big push on that right now i believe i've seen some threads going on about it and they're looking at how they're gonna sort of make that look and work and i'd even seen another thread where they're talking about

being able to distribute themes as a snap so don't like 100 quote me on that but i've seen some talk about that as well but the uh it's the fix is the fix is coming hey look at that dan from elementary says he's using the mumble snap too nice uh dan actually has a little news from the project i think we're going to talk about here in just a moment so dan get your uh get your medium post ready we'll bring that up here in just a couple of minutes i don't know you have anything else you think we should mention about the event i i think i'll probably play some of the audio we have next week once we've had a chance to just review everything.

Initial impressions as our first? My initial impressions I guess would be that it's been interesting to get insight into the culture and community and stuff like that. And being able to have some of the discussions even though we can't talk about them on air, even

though we can't record them or anything like that, it informs our opinion and it makes us valuable as a contact network because we're the only broadcasters that are here. That's true. We're the only people that have this kind of insight. We're lucky it's in our neck of the woods. I think they recognize that Jupiter Broadcasting is more important than that because they brought us all the way out to New York for that too, right? In the same way.

Well, we brought ourselves out there, but we were invited. We were invited, right. So from the perspective of when people are looking for tech news, and this is something I talked about on a show last week. When people look for tech news, you want people that have informed opinions, people that are inside of the things that they are talking about. I can't make the case any better. I mean, we're sitting here with the Ubuntu guys at the same rally where they're figuring this stuff out. I don't want to give the impression, though, that we're also getting like – we're not getting the red carpet either. Like we're working it.

Like we have to work for what we're getting too, which is fair. Fair enough. I mean they're here doing their thing. We're here doing our thing. That's how it should be really. But yeah. OK, Mumbler, I thought I heard somebody speaking. Does anybody want to jump in before we cover the last bit of... Yeah, go ahead.

I wanted actually to say that even though I can enjoy that Canonical is doing all of this, I just wish that they actually had a more concerted developer documentation set. And when they decide to build something, to build that something, actually maintain it. As a developer, being a company or being just an open source developer, I want to be able to go to Canonical's page or Ubuntu's page and go to developer docs, click there, click there, learn about that stuff that there are the recommended tools and be able to build, not need to attend an event

to figure out how the things work. I understand that they're doing research, and that's what they want to get out of that type of event, so that they can build it better. But if they just actually do that part well, I don't think we would even be spending that much discussing these events. Every company does these events. I'm not taking this as a wow thing.

I think I disagree. events. I'm not taking these as a wild thing. I think I disagree. So I would say, first of all, I wouldn't necessarily undersell docs.snapcraft.io. I would check out docs.snapcraft.io because it's pretty good documentation. But the thing is, and somebody who's had clients, it was an IT relationship, but I can tell you there are certain companies because of their status, because of their workload, because of their funding structure, there is a trillion different reasons, but I can tell you for sure that some of them just have to have, by the way, Noah's got a call. That's Noah's emergency bat line phone right there, if you heard that.

I can tell you there is some companies that really have to have their hand held. And I don't want to sit here and Microsoft bash, but sometimes if you want to get Microsoft to play ball a little bit, it just seems to be the way it goes. I agree, though. It would be great if everybody could be involved or if they could live stream this. That would be really useful. Look, no.

I'm not even asking for that. I'm not even asking for that. This is one of the things that I've pointed out multiple times that I believe that elementary has done better than any other distro. It has a set of tools that you can follow. Like, I'm not a particular huge fan of everything elementary does, but this is the one thing done right.

You can look at it and know, okay, this is what I need to pick up to build for elementary. And you go to Windows, Windows says C Sharp is the way to develop for Windows. Apple says Swift. No, hold on a second. First of all, I don't know that Windows is – I don't know that you can compare the two. Can you really compare Windows with Canonical?

No. I mean, I would say just check out docs.snopcraft.io. I'm not here to defend them. I'm not here to say that one is better or how they do it is better. From my perspective, this is sort of the way the business industry works, and they're not doing it because it's not documented. And again, all of this, even like the YAML files and all of that is pretty well documented.

So I feel like it's more about connecting with certain companies that need their hands held, that need that one-on-one or have a special pony problem. And then the reason we're here is to relay that that's what this is about. There's a

motivation from Canonical's standpoint to have Jupyter Broadcasting here. And it's not because they're big fans of Linux Unplugged. I mean, maybe it is. But I think it's probably because they want somebody relaying to the community what happens here. And what they want... Right. The partners here may not be

talking about it, right? But Canonical wants to be shown that they're building these and right they're not telling us what to say but talking about it right but canonical wants to be shown that they're building these things they're building these things and this is what's happening these events so i think because i think what the narrative could be if this wasn't if we weren't here and there wasn't a third party sort of arbiter eyes on these somebody could say well they're having these secret meetings with these secret companies and they're not even saying what companies are there they're're not even disclosing any of this. They're just doing all these things behind the scenes and pulling the canonical strings so that way Mark Schallerworth can

orchestrate his master takeover of the Linux landscape because Mark Schallerworth wants to dominate Linux via snaps. And that is the narrative that gets perpetuated when they do these kinds of events. I think they've gotten hip to that and they realize that they have to have some eyes on it. They have

to have somebody watching it, otherwise They don't have an agenda. Yeah, otherwise it just gets spun like crazy. Right. You know, I mean, I can tell you that if this was like a, if this was like one of those, we'll have them come in here and we'll just have them just tell people what we're doing and we don't have to, you know, we'll just, this will

all be taken care of. We wouldn't start the show with a Red Hat story buying out CoreOS because that's something that directly goes against what Canonical's interests are in the enterprise and server market, you know. I think it's so they understand that we don't really have any, we don't have a particular skin in the game, but they still want people to know what happened. Right.

And because otherwise it does come across as some, like, it comes across as if Mark Shuttleworth is pulling some sort of puppet string. Yeah, exactly. He's not even here. It snaps to still feel like, you know, it's a community if Mark Shuttleworth is pulling some sort of puppet string. Yeah, exactly. He's not even here. It's nice to still feel like, you know, it's a community effort by an open source company. Yeah. Even if you don't get to see every last thing. Right.

I mean, the people that are working on it that are interfacing with these companies, like the Azure team and Plex, are wimpy and popey. And the people from Canonical you've heard of before, they're just doing it in a setting that's sort of like common ground. Yeah, and I would be great if you could have pop-up, you know, how about this as an idea? Like pop-up Snap events where you need different locations at a Lug meetup or at a Linux event. Maybe there's other ways to do this. to do this uh but i don't think with i think i think if i guess what i get back to is if abba if app image had been doing this for every six months for the last couple of years they'd probably be way ahead if if flat packs had if red hat was cutting i mean fifty thousand dollar checks every six i mean new york was way more than fifty thousand dollars probably but this you know

this is probably a 50 25 that000, $25,000. You're paying all these people. You've got hotel rooms. You've got dinner, lunch, drinks for everybody for a whole week. It really shows an investment by the company. You're flying these people out from the U.K. I mean, the majority of the staff are all coming from London and other places all over the world. They're not coming from the States.

Some of them are, but most of them aren't. And that is a huge financial investment. Some of them are, but most of them aren't. And that is a huge financial investment. I got to believe that if Flatpaks did that, if another packaging format were doing that, it could do that. They would be doing that because this is what it takes. It takes a sort of industry collaboration because you're reaching out not to other,

you're not reaching out to businesses that have been making an open source product for the last 15 years, right? You're not trying to get crossover to package as a snap because eventually they're going to see the light because they're smart folks and they work in the open source community. You're reaching out to companies that have never done this shit before. And if I'm going to pivot the packaging of my product, I want to know there's someone I can call on the other side and be like, we need to get this fixed. You said you would help us. You want me to retask this 25-member team to package in your package format? How is this not a play for you to take over Linux? You've got to make it as easy for them as possible.

What problem are you solving for me to even want to consider this, right? But there is also a comfort level in, oh, I can call up this canonical company and I can tell them they've got to fix this problem or else, you know, we're out. I see both sides of it. I do see both sides of it,

and I see why they have us here. I know what their motivation is. They want to make sure it's sort of understood what they're trying to do. It's a complicated thing. It really is. And we're here to observe like flies on the wall, Wes. Searching out bandwidth at every opportunity. Linux-based flies, let's be clear. Yeah. Hey, you know,

there's one nice thing about coming to these kind of events. It's just about everybody. It's like a 95% Linux operation. You got a few people coming in from different companies, but even the people that are coming in from different companies, a lot of them are running Linux. And it's a lot of MacBooks running Linux.

It's a lot of MacBooks running Linux. I can take that. You walk around, you see the front of it, and it's like, oh man. And then you start walking around behind them, start looking over their shoulder, and they're like, ah, those are all Mac. Yeah, it is. It's all Unity. You want. And then you start walking around behind them, start looking over their shoulder, and they're like, ah, those are all Mac. Yeah, yeah, it is. It's all Unity.

You want to see a tricked-out setup. Oh, I know what you're going to say. Wimpy's XPS 15 with Ubuntu Mate. He's got... Is that Ubuntu Mate? Yeah. Or is it Windows? No, no.

Or is it Mac? I don't even know. He's got all these different desktops set up with different stuff. Plus, he's got virtual machines, and he's got experimental high DPI builds where stuff is dialed in on this version of mate that is like he's getting ready to you know put out there i had not seen it in person and it is all the new hotness right there it looks amazing i think he's switching people to it just by just by using it

being around it yeah popey's switching yeah popey'sy's like, I'm going to give it a go. I actually switched, not to Mate. I switched to something. I got rid of my desktops. I'm going to save it for the show, but I'll give you a little hint. You can see right there. You see that right there? Oh, we're going to do an ad.

We've got to do an ad. You're right. I just got distracted. All right, so let me tell you about DigitalOcean, then we to do an ad. You're right. I just got distracted. All right, so let me tell you about DigitalOcean, and then we'll get back there. DigitalOcean. I didn't forget. I was just getting there.

I was just getting there. DigitalOcean.com. Go over there and sign up. You know what? In fact, I'm going to make this up right now. We'll just use a promo code. Let's just say it's DL Unplugged. Let's just say that.

DLUnplugged, one word. You apply it to your account, and you get a $10 credit. Let's just say once you have a $10 credit, you can run the $5 rig two months for free. Let's just say that. D-O-N-plugged, one word. You apply it to your account, you get a $10 credit. Let's just say, once you have a $10 credit, you can run the $5 rig two months for free. Let's just say that. And then maybe you discover that this machine's incredibly powerful, and you didn't know Linux could do that much for $5 a month. And then say you get hooked, you want to deploy another machine. Maybe you got an

application you want to deploy. You can get deployed, I mean, you got deployments for days. Deployments for days. You know, and if you want to go play with some of that fancy new Snap technology, go spin up an Ubuntu machine on DEO. Perfect. You can set up an automated build system, or you can just play with some Snaps. Boy, can you.

The nice thing about DigitalOcean, too, is you just don't have to deal with any of those complex pricing structures. It's just simple. You always know what you get. It's predictable. It's scalable. It's reliable. 99.99% uptime.

You have an SLA for all services. You can deploy an entire application stack. percent uptime you have an sla for all services you can deploy an entire application stack like the base linux system the software like it's a git lab everything in one click or you can deploy the base ubuntu system or base core os system or centos system and then build on top of that you get ssh access in you get html5 console access you can just get started faster if there's some new open source project you want to try out, see if it's worth your time, you don't want to muck up your system. Oh, it's the best for that.

I go to this, and then the thing is, it's so fast, it has easily increasable storage, it has eight data centers all over the world. DigitalOcean's got data centers for days. So I just leave it in production. It's perfectly just... And they've got all the little things you might want to make that work too.

Firewalls, monitoring, build right in. Yep, yep. Easy, easy peasy. In fact, it's really easy now with their new DigitalOcean Spaces to do automated backups. If you check out their community page, they posted on my birthday a blog or, no, I guess a tutorial. Because blog would be underselling it. Because these are well crafted, well written tutorials, beautifully done, elegant tutorials, Johnny Ive White Room tutorials,

and they're great. And so this is one about doing automated backups with DigitalOcean Spaces. Go there, try it out. DigitalOcean.com. Use the promo code D-O unplugged. I wouldn't be in this hotel ballroom if it wasn't for DigitalOcean. Thanks, DigitalOcean, and thanks to you guys for sponsoring the show. DigitalOcean.com Linux boot. Let Linux do it is their tagline. And that should be everyone's tagline. Just let Linux do it. That should be Ask Noah's tagline. Yeah, right. So I think the big

deal here is that from the user perspective, UEFI works okay, but it's a giant, bigger-than-the-kernel inscrutable code base. It's a big mess, and we already have a code base that we know and love, and it's called Linux. Yeah, and it does a great job. It does. And so they write, are you tired of reinventing the wheel by implementing drivers for firmware again and again? Not with Linux Boot.

To the rescue! Yeah, I know. Just set it and forget it. So Linux boot is a firmware for modern servers that replaces specific firmware functionality like UEFI DXE phase with a Linux kernel and a runtime. They say it will improve boot reliability by replacing the lightly tested firmware and drivers, improves boot time by just removing unnecessary code, makes it 20 times faster to boot, which is actually kind of nice on a server, especially when you're trying to get up in a noutage. And they say it's a proven approach for almost 20 years in military, consumer electronics, and supercomputing systems,

whenever reliability and performance are paramount. I don't actually know what that means, because the only supercomputers that I've had experience with are actually often initiated by other computers. because the only supercomputers that I've had experience with are actually often initiated by other computers. So an entire standard PC boots up and then initiates the cluster of supercomputers. So I don't really know what that means. I'm a little curious what they mean.

But I'll go with it. I'll roll with it. I mean, Linux has been a data center for a while. And they may just mean the approach, too. They may just mean simplifying and relying on one shared code base. Linuxboot. Linuxboot.org. See, I'm pretty excited because you know me, Chris. I just love me some KExec.

And that's a whole integral part here. You are a big guy. So that's how they're doing this? Well, yeah. So the idea here is... Wes will literally go two weeks and not even install a bootloader on his machine. He's like, I just KExec it. I just load in the one that works in KExec.

Yeah, it works so well. Yeah, so the idea here is you still have some firmware, and that'll vary. Maybe it's a little bit of UEFI. Maybe it's core boot or, you know, uboot on one of those ARM systems. That does some of the firmware initialization, the bare sort of stuff to get the CPU started, the system running. And then you just boot a kernel because the kernel already has a whole bunch of code for driver and device initialization. I see. So you get that.

You have a little helper in it, RamFS, if you need. There's a couple different implementations of that. And then you get your root device supported. You do some pixie booting if you need to. Once you get the real kernel, kegsec in there. Away you go. Bob's your uncle. Yep.

That is great. It's funny how quickly, I guess funny isn't the right word to say, but it's awesome. It's so badass how Linux really gets its shit together quickly on this stuff. It's like, oh, we've got these three or four technology stacks that we could leverage to do this now. Where stuff in Windows comes along, and it'll literally sit in Windows for 20 years unused. And it just never gets taken advantage of. The system never really improves.

But we come up with these three or four disparate projects, and we all combine them together, and now you've got a head-secure runtime. You're using nerfs for just nerf and uroot and initramfs and all this crazy great kernel functionality that are all being written by different people. And smoosh it all together, and I think you've got something great that seems like it doesn't happen on the commercial platforms as much.

And it's interesting here, too, because you see some of the big names involved here people like Google and Facebook and they're seeing some you know real real problems they have in their actual businesses that they're trying to to solve and they're doing that by leveraging open source technology working together to extend it which is pretty pretty exciting. Well we'll see we'll see how long it takes to actually get out there. Yeah, right now there's about two test devices you can use. So whether or not this sees real-world data center deployments, I don't know. I just want it on a laptop. Yeah, exactly.

Ooh, you know, that would be so great. Something on there. So yeah, I guess I'm feeling this because I've been reloading machines. Again, this is the cross that I bear. I can't help myself. I can't help myself. Yeah, you stopped being the Linux switcher, and yet you're still the Linux switcher. Somehow. Somehow.

Yeah, so Mark Howe is just joining us. Mark Howe joining us live over at jblive.tv. We do Linux Unplugged live on Tuesdays. This week we're on location in downtown Seattle, and we are live at a canonical Snap event. Their January Snapcraft Summit,

I think is what they're calling it. And we're here sort of being flies on the wall. And so Wes is joining me, Noah is joining me, and the Mumble Room is joining me, and now Markow in the chat room is joining us. Excellent. You can join us, too. Go to discord.me slash jupitercolony.

And the chat room, there's a Linux Unplugged chat room in there that goes all the time. So if you want to join the show and talk about anything that you saw going across the chat screen, if you watch the video version, you can go in there and find it. You can get links. The conversation continues in there. So there's a really big conversation kicking off this week around client-side decorations again. I don't know if you guys remember client-side decorations, but they're like those header bars on GNOME apps. G-Edit's got it. Corebird's got it. decorations, but they're like those header bars on GNOME apps. G-Edit's got it.

Corebird's got it. Funny, most GNOME desktop applications have it now except for Nautilus and Terminal, which are the ones I use the most. But client-side decorations is sort of solving a couple of problems for the GNOME desktop,

and so they're pretty hot on it. It is solving some Wayland transition issues where applications would be responsible for drawing their own decorations. And when I say decorations, I'm talking like the title bar. I'm talking like the close, minimize, maximize. The title bar around your application,

the borders around your application are often referred to as decorations. And the GNOME project is pretty keen on the idea of applications drawing their own borders, drawing their own titles, their own decorations, and doing it in a way that sort of brings everything into that bar. So instead of having a separate file edit menu and a title bar, it's all in one combined bar. And if you go look at screenshots of Gnome Builder, you'll see exactly what I'm talking about. And in a blog post that they did recently, there is, Tobias,

I believe, did it. There is a blog post that shows like, what would Telegram and Slack look like if they use client-side decorations? And they're down there towards the bottom of the post. And I got to admit, it legitimately looks good. And this is a conversation that they've started before, that the folks on the Plasma and KDE side of the camp have rejected, saying we've come up with a cross-desktop protocol that we could use. We have these server-side decorations that we think are great. And, oh, by the way, if you've got client-side decorations, if that application crashes, well, it's responsible for drawing the title bar, so now the title bar has crashed.

So it's much harder to move it around. The windows froze. They've been iterating and discussing these problems and iterating on that discussion. What's wrong with client-side decorations? And so there was just sort of a stalemate. GNOME went ahead and just kept doing some stuff, client-side decorations, CSDs. But it kind of slowed down. Not a lot of projects.

You haven't even talked about it, it seems for seems like at least a year no although in the background uh so firefox 59 is coming out next and it's implementing client-side decorations as an option when you go configure the toolbars so i i thought i really need to get a perspective on this because as funny as it is this problem isn't going away and every it comes up, each camp gets even more sort of entrenched in their view of what's right and what's wrong. And it's, the language is getting a little more hostile now. It's getting a little more aggressive. And so away from a technical discussion and suddenly people have stakes now

and it's emotional. And so what is it all about? Does this even matter? Is this all a bunch of crap for very little gain? So I wanted to understand why the GNOME project wants this, because I get why the KWIN developers think it's not so great. We'll get to all of that here in just a moment. I'll talk about some of that.

So even though I thought I was going to be staying on Unity 7 until 1804 came around, I... You've been singing Unity's praises for months now. I thought, you know, I need to just have some handle on this. So, I gnomed up. I gnomed up all over the place. I started with that Arch machine that you had on the MacBook. Got it all up to date. Got my Gnome 3 all dialed in. I was like, all right, well now

let's move to another machine, another machine, another machine. It's how I do. And each time I sort of build on it and build on it. And I went all in. I got a beta build of Firefox 59. So I had Firefox with client-side decorations. I configured MailSpring to use client-side decorations. I loaded up CoreWord, which uses client-side

decorations. I went... You went full CSD. Yeah, I went full CSD. Full, full CSD. So I'm going to talk a little bit about that. I want to talk about what Martin, the KWin developer, why he doesn't like this, and where I ended up. And then we'll wrap it all up, like I said at the top of the show, with an app pick that I think is the best terminal emulator I've ever used. So that's where we're going next. So why don't we take a moment and thank Linux Academy. Linuxacademy.com slash unplugged. You go there to sign up for a free 7-day trial. Learn more about Linux and support the show. It's a full

featured training library with everything you need to learn new skills and advance your career in Linux and open source in general. Linuxacademy.com slash unplugged. Go there, sign up for that free 7-day trial and get a sense of the scale of the platform. How about that, huh? Get a sense of the scale of the platform. How about that, huh?

Get a sense of the scale of the platform. It really is. It's everything you need from hands-on labs to instructor mentoring, learning paths, which are a series of courses and content that is planned by instructors for specific career tracks. If you're busy as heck, they've got course schedulers so you can set a time frame and fit to it, and it'll help you set learning goals.

If you're a bad tester, they have practice exams to help you get ready for certifications, test what you've learned, and they also have courses created specifically to prepare you for certification exams. They have cloud servers that they spin up on demand that help you do the courseware in real-world production situations.

Shoot, that's worth it right there. Linuxacademy.com slash unplugged. And you can also go all Noah go full brown bear and go offline with their comprehensive study guides, their lesson audio, their personal notebooks, you don't need the cloud for any of that, you can learn Linux with that stuff too, offline linuxacademy.com

slash unplugged and check out the community section like the forums, you may find a fellow Jupyter Broadcasting member lurking. There's plenty of them. A vibrant community that's there to help you. They fork study cards. They really make it worth it. And then Linux Academy is revising content.

They're keeping it up to date, and they're adding new content and an incredible clip, consistently making your subscription worth it. Right. I mean, they can't make it no work. You're still going to have to put in some time, learn the content, play around with Linux. But they knock down just about every single other obstacle I can think of.

And they're going to help you get your head around it. Yeah. Yep. Linuxacademy.com slash unplugged. So let's talk about the KWIN side of things here. So GNOME likes it for the aesthetic value that solves some whaling problems for them. And the number one thing is it uses less vertical real estate.

You know, you're getting less stuff up and down. Some of the well-done ones, it looks pretty nice. I don't hate it. And on widescreen displays, it's nice to get more vertical real estate. But Martin on his blog writes that he's not a big fan of this, and he calls it a situation. He says that the Gnome Shell and Westman require client-side decorations, but KDE Plasma and Sway do not.

In fact, the Plasma folks, which I believe Martin was the one who drafted this, created a protocol supported by GTK, although it is broken currently, I've been told, that allows to negotiate with Wayland Compositor whether it's using server-side decorations or client-side decorations. In fact, he says the don't believe gnome, don't implement CSDs. You can ask them to support server-side decorations if that's how you feel. They could do it because they need to do it for xWayland anyways. That's a good point right there.

If you've got to do it for xWayland anyways. He says, and I totally don't understand. I totally understand you toolkit developers that don't want to implement your own client-side decoration. It's a lot of work, and it's difficult to make it look good and not alien on different desktop environments. I don't know. You know, I don't know. If you've ever seen Slack on Windows or Mac, I think that's using client-side decorations, essentially. So you can make it work, but they have a good point. You have to solve this problem for X Wayland anyways. You still have all of these developers who aren't going to want to create that

because it's a pretty big visual piece of your application all of a sudden that you maybe aren't even out for designing correctly. So I see both sides of the camp. So I say, I've got to just run these. I've got to run this. And I'll say, it's nice. It is really nice from a visual standpoint. But I ran it for just three days.

And in that time, sure as shit, Gnome crashed on me four times at least. And our friends at Canonical just recently iterated that one of the reasons they're shipping X.Org and 1804 as default is because when Gnome crashes, when Gnome Shell crashes on Weyland, it is catastrophic. Oh, yeah. I is default is because when GNOME crashes, when GNOME crashes on Wayland, it is catastrophic. Oh yeah, I mean that's it. It is the display server. Yeah, that's it. See, when you're on

X, yeah, when you're on X, you have the luxury of X still running. You can reconnect to X and GNOME just sort of pops back up and your applications look all shitty for a second and everything's fine. But on Wayland, it's a crash. And I'm sitting there going, this thing crashed on me four times.

Four times in three days. And oh my gosh, if Noah hasn't had some gnome issues here. You know, I... I don't mean to be a gnome bashing issue, but this matters. You've had full crashes. And if we were in a Weyland future...

Well, here's my issue. Like, we went to sit down to do show prep today. And I think I was sitting there next to you for, when I pulled my laptop out for 35 seconds, a gnome crashed. And you know what was frustrating

was I was, the reason I pulled my laptop out because I have the memory fish of a goldfish. Yeah, you have the memory of a goldfish. Yeah, and so I wanted to dump something down

and I opened up a text editor, got one sentence into it and the whole thing just dumps. Of course, I lose all my work. And you don't know what you're trying to write down anymore. Right. Well, it was frustrating. And then I went to recover.

And the system locked up so hard that I had to physically power the computer off and then power back up. And by the time I got through that, remembering my encryption password and getting back into Linux and opening up the text editor and getting back to what I was doing. By the time I did all of that, I forgot what I was going to say, which turned out not to be a big deal, because about five minutes after that, it crashed again. And then we closed and moved rooms, and I came back into a different room, and it crashed a third time. And so the question, what you and I were talking about is... What are they going to do about this?

Yeah, I'm a pretty dedicated Linux user, right? Like to the point that if my choice was between linux and and no computer at all i probably use no computer at all yeah and if i am getting frustrated and joking about well maybe i should use windows so at least i can get through the show and then i'll go back to using if i'm making those jokes what is the next guy that's like i was kind of in between but you know and here's what has me scared a little bit and i mean scared it actually really is something that i'm a little worried about here is the gnome team has a fork in the road ahead of them and what they're discussing right now up on their wiki and on the irc room and whatnot is gnome 4 do we re-architecture gnome and make it modular so we break out mutter we break out all the compositing stuff completely

separately we we break out the shell we break out the graphics effects we break out Mutter. We break out all the compositing stuff completely separately. We break out the shell. We break out the graphics effects. We break out the extension architecture, which also means we create a standard API. We have to create an IPC communication system for all these different modulized components to communicate with each other. There can't be any drop in state or delay. Or, and if we do that, we break every single extension in existence today. And mind you, today, the way the extensions work in Gnome Shell 3 is they're just hot swapping out JavaScript code in the Gnome Shell. It's not like these extensions are talking to Gnome and saying, I would like to put a meter up in the toolbar.

They're just slicing that code in their hot dog style, hot dog down a hallway. And that is dangerous. That's not how you want a professional desktop run. You know what I'm saying? But at the same time, critical functionality of GNOME 3 is provided by these extensions.

Right, and we want that. Especially if GNOME's going to concentrate on sort of a small core, you're going to need those for a workable desktop. And so the GNOME project's like, well, you know, we tell people if we get rid of the system tray to go to the top icons extension. Well, we tell people when we get rid of desktop icons

to go to the desktop extension. And now do we go break all extensions on people? Imagine if Firefox took away bookmarks. And so we'll just use an extension for bookmarks. And then two releases later, they broke the bookmark extension. So there's no more bookmark extension. Well, you just told me to get the bookmark extension. How can you go and break it two releases later?

That's the position they're in now. So they either break all extension in existence and make GNOME a usable environment to begin with, or they kick the can down the road, they create maybe an API or something, and they don't address this core architectural issue that, if you think about it, is a massive challenge. And so it's not that I...

GNOME is my favorite desktop environment, hands down. It's the one I... If it never crashed on me, it would be the one... You'd be there all day. I'd be at the top of the roof right now at this hotel, 29th floor, screaming about how great GNOME Shell is. I just... I can't, when i can't

even make it a few days without it taking a huge dump on me and now having a little more insight into what's causing these problems i see that they have a massive technical hurdle to overcome and it feels very different right like for a while especially after like 320 or something we can talk about gnome felt stable again, right? It was like, okay, I see what they've done. As any software project grows though, right? And as you add more extensions and the answer can't be no extensions.

It just loses, it just, I think also Wes, the thing is, it's the Wayland pressure because now crashes have way more consequence. It's not just a repop of the shell. It's you lose everything that you're working on, and you reboot the machine. And we're in that awkward middle ground now, where before it was sort of like, oh, if Wayland's a crashy mess, it doesn't matter. No one's really using it.

But people are starting to use it for real things in real production applications. And if it's not up to the task, that's a serious issue. Yeah. Well, they're making hard decisions in Gnome Shell that if they can't figure out how to do it in Wayland and X, and they can only figure out how to do it in Wayland, they just drop the X version. Right.

Some of the things we're seeing happen to Gnome Shell are really about getting ready for Wayland. And just a couple weeks ago, we talked about how the KWIN project is a Wayland-first project now. Yep. And X11 is now in indefinite maintenance mode for them. So we are making that transition. It's happening. Fedora is shipping it. Ubuntu 17.10 is shipping

it. And GNOME is making a huge ask with these client-side decorations when they're in sort of a tedious position. Because the tedious part of it is they have bigger distro support than they've ever had. All of the major

distros are shipping GNOME desktops right now. All of the corporate desktops. It's kind of their heyday in one sense. Unless you're Blue Systems, everything you're shipping is a GNOME desktop. So they are in a great position, but the tedious part of it is they can't really use that leverage because there's these fundamental architectural

issues that everybody behind the scenes is going this is a big problem. I don't know what the hell we're going to do about this. What's going to happen? We don't know. Yeah. And so it was with that background that I was like, okay, this is an interesting time that they're choosing to bring this up. Dan, I don't know if we have Mr. Dan Rabbit's audio again, but I wonder if Dan has thoughts on the client-side.

Yeah. Dan, what do you think about this client-side decoration conversation This conversation has come up again, especially from like elementary OS desktop standpoint. Well, we've been all in on client-side decorations for a couple of releases now. So, I mean, they work fine. And like you said, the space-saving thing is really good. One big thing about client-side decorations that you get is they're actually integrated with applications that want to use the dark style instead of server-side

decorations that aren't aware of the application's content, so they can't adapt correctly. We also have a facility in Elementary OS for developers to be able to set brand colors in their applications. A lot of the title bars for applications in App Center are branded colored instead of just plain,

you know, the same for every app. Like, there's a lot more, like, interesting color in the UI. Yeah, I have seen that, and I do really like the way that looks. I've also seen people make certain buttons stand out more that, you know, like, my eye is drawn to it more because it's, like, a certain color of blue that really pops or something.

So that's, you know, and the red buttons really stand out. Don't click that unless you really want to lose this document. I like that. So, um, do you have any kind of sense of where the K win project is coming from and, uh, see like that we'll do the server side decoration thing instead,

because you still got to solve this problem for X Wayland. I mean, I don't know. I, I, I, got to solve this problem for x wayland i mean i don't know i i i'm not sure i fully understand the argument against having client-side decorations to be honest it it kind of seems like you can do everything you could do with server-side decorations with client-side decorations i i know i see some people saying like oh well that means that all the apps draw where their

own window positions are. And I don't want my apps to have different window button positions. That's not really how it works on the GTK side. Like it's a widget and you as an app developer just use the widget and you pack everything in between the window decorations. But GTK itself decides where the buttons go. So, you know, if it was solved in that same way with a cute widget on their side, like, that completely removes that argument, too.

Yeah, it does seem solvable on their side. In fact, it almost seems like it'd be easier for them to support the client-side decoration stuff. So I teased it a bit. Do you want to also touch on the switch from CMake while we have you? Yes, that would be great. We'll circle back to this.

Tell me about the switch from CMake and what's that all about and what are you switching to? We're actually really excited about this. If people don't know, CMake is a build system. A build system is what you use to take your source code and turn it into executable binaries. So every application has to have a build system of it as compiled code. We're using CMake for a long time, and CMake's pretty good, but one of the things we have to do to make CMake work for all our use cases

is we carry this big set of modules that helps us do things like internationalization or using G-Resource or all kinds of little plugins like that. Well, we recently... I sort of hacked those on. Yeah, and so developers kind of have a choice where they either add a dependency package that is only shipped in elementary OS, or they shove all these modules into their

upstream project and try to stay in sync with our module package. That's not a super great experience. But we've been looking a lot into the Mason build system, because more and more people are adopting it. It's in SystemD now. It's in SystemD now. It's in GTK+. And so we talked with their developers upstream, and we were actually able to get some patches in there to solve some issues we had with Bala support

and translations and things like that. And they've been super awesome and responsive to us, and we're totally all in on Mason. We're migrating all of our projects over to it. That's awesome. Wow. Has that been, from like a project standpoint, is that like something you guys have like a powwow on?

Or is that like somebody decides on high? Like what's the structure on making a change like that in elementary? It kind of started off as just like a, hey, have you heard of this Mason thing? I'm like, oh, yeah, I don't know. Yeah. And then like I think I think actually it was me who first used it. One of my little side projects. And then I showed somebody and I was like, hey, check this out.

And it was like the amount of lines of Mason it took to build the app was like such this tiny like neat nice little file that everybody was like holy crap i gotta look at that you know and so it just kind of kind of spread out from there yeah and then we're all kind of like okay let's just do this for everything now that's the best that is great i gotta say i really appreciate that your you know blog post had such an easy example because i've seen a lot of other projects right about switching and it's easy to say oh yeah you know it's an easier config but your example really lays it out like it's just simpler yeah yeah i will link that in the show notes that is a very good point i uh i want to now uh i want to shift back to let's wrap up this uh plasma conversation real

quick thank you dan for the insight on that i was curious i've been i've been really getting the elementary os itch recently trying to hold off off the itch before release just so that way I don't... But boy, do I... Every time you come on here, it gives me the itch. Or that's the rash. I'm not sure. But either way, I appreciate it. Thank you very much for giving us the update.

Thank you. This was sort of... I sort of did this too because Wimpy's like, you need to try out Plasma on Wayland and see what that's like compared to Gnome on Wayland if you're having all these troubles. And as is typical, Wimpy was right. And so I tried out KDE Neon. I got the user edition.

And I loaded it on the AirMaster broadcast machine that's having some issues. I loaded it on my laptops. And I've been sort of changing my perspective on Plasma Desktop. I've always liked it initially, and I tend to sort of trail off because I tend to start to get annoyed with having to change all of these settings.

Every time, the first time I launch any Plasma Desktop application, I just have to go through all these configuration changes. Everything, Dolphin, every app on the desktop. You know, if you have, like this, my XPS 13 has a high DPI screen. Yep. I got to change.

I have to go into the display output, and I have to scale the display output. I have to go into the font settings. I have to scale up my fonts. I have to go down to my panel. I have to resize my panel. And I have to go into the mouse settings and change my mouse cursor to support high DPI. So it supports high DPI if I change four different settings in four completely different locations in the system settings then i get high dpi uh after i've done all that you know it's nice it's fine it's great

i don't need to change it again that's true for everything dolphin console everything's like that those knobs are nice on one system if you really want it to be just the way you want but when you have five systems you kind of just want the nice default. I was laughing. I got down and I'm like, yeah, I'm configuring how many lines my screen scrolls when I move my mouse scroll wheel one notch. I'm like, this is probably at a level of detail I don't need to be at anymore. So that in the past would piss me off. Like, what are these guys doing?

Why don't they just add some sane defaults? Why doesn't anybody just add a sane default? I'm flipping on that now. And I have a different perspective after getting just dumped on by GNOME a lot. This is an ongoing issue. It affects our OBS rig at the studio. It's a constant problem. And I've come to the realization that having to change defaults for six months

is a lot better than having my desktop crash on me all the time. And so if I'm going to live with one annoyance, I'll live with the fact that I can change the defaults, and if nothing else, maybe this will finally get me off my butt to actually sync my configuration amongst my machines. I've threatened to do it for years. You have threatened to do it for years.

You know what I mean? Yeah, absolutely. Every time I get into Plasma, I start to threaten even more, but now that I'm sitting around Wes doing text-nep all the time, I think I may finally get off my duff and sync this config around. Wes, are you a

C file fan? You know, I have used it, but I don't use it currently. Oh, okay. Are you Dropbox? Are you a sync thing guy? I have a sync thing for some things, and then some chef-based config for other things. Okay. So what would it take to get you to try C file for syncing

these configs? Just some time to do it, I suppose. Maybe we try it tonight if we get back to the studio. I'm telling you, here's a Noah Redbook prediction for Chris. Once Chris tries C file, he will never go back. Because here's the thing. We should try it tonight if we have time. I feel like I've said this a couple of times throughout this weekend, but there

are projects that I like because they're open source projects, and I just think they're really good open source projects. Yeah, yeah. And then there's good on the level of this is such an amazing product. The license, be damned. It's just a good product. And that's how I feel about C-File. It's just, it is

one of those things I installed one time. I haven't looked at C-File in ages. I installed it one time on one server. Installed the sync client to all of my computers and I change a lot of data. I've got video files and audio files and all the source files for this show it integrates with collabor online server too yeah it could be cool syncs all over the place and i never once not one single time have it bite me and the closest i ever get is let's say i leave a file open on one computer and i open up on another computer and i it it up on another computer. You get a conflict copy?

Yeah, I get a conflict copy. And it timestamps it for me and says, here, boss, here's all the files you have. Which one did you want to actually... Amazing. So their pricing is based on amount of users, not amount of storage? Well, C file is free.

You can pay for them to host it for you. Oh, I see. But you can also just run it yourself. If we can make it back before 10 p.m. tonight, we should try to set this up. All right. Because I've got to try this. Because I've got to tell you, now that I have this plasma, I was going around and showing some of the canonical guys, too.

I'm like, let me just show you this. Right? I've got a pretty sweet desktop. Yeah, yeah. I've never heard of it. Yeah, and I'm like, and you know what? The consensus is, once you get it set up, these plasma guys really know what they're doing. Like, they have built this thing real well.

And it's perhaps better positioned for our Wayland future than I think. It's kind of been the dark horse, right? We talk about Martin's excellent blog posts here and there. But they've just been churning away in the background making stuff work. Yeah. making stuff work. Yeah. And it sort of strikes me that it's actually visible.

Just Noah and I were sitting in the studio. We had AirMaster running Unity 7, which is good. You know, it's pretty performant unless it's been running for a week. But if you reboot and it's been like 48 hours, it's good. And we moved. We just sat there. We didn't even stand up. We didn't move. We just reloaded the machine to KDE Neon with Plasma

and he and I both look at each other and go, this is noticeably faster. This is just, we can just look at the screen and you can just tell it's noticeably faster. And that was with the open source driver. Then we loaded the NVIDIA blob and all of a sudden it's like, boom, snap. Yeah.

And I know it's a fresh install, but we weren't replacing super complicated installs. These are pretty minimal. Pretty minimal machines, yeah. Oh, my God. It's smooth. It's clear. The transparency is the... To illustrate that point, and you'd have to be there to really understand it, but to illustrate

that point, the boot time was so fast that he literally thought we were waiting for the BIOS to post and then getting to BIOS. And it was actually, we were waiting for the log on screen to actually show up. Yeah, and part of it, I think, is I switched out LightDM, and I think that makes a difference. Because it doesn't flash as much and all that kind of stuff. But I know I do this. I know people that have been listening for a while know that Chris moves around. Chris is on that new thing now.

Yeah, no, I know I do this. I know people that have been listening for a while know that Chris moves around. Chris is on that new thing now. Yeah. No, I know. So I'm doing this specifically to sort of live client-side decorations for a bit, live the Plasma Wayland, just get some perspective on these particular issues that are coming up and get a feel for it. And so right now I'm running here on this XPS 13 Plasma, the current version of Plasma for the KDE Neon distribution, or it's not, don't call it a distro. High DPI, Wayland driver, or Wayland, using Wayland for display, and

getting great battery life, getting great performance, getting noticeably better battery life. I did all of Coda Radio. I never do a show on battery. I never do a show on battery. Well, you're a professional. Ever. And we got here in the morning, you know, Coda Radio is at noon. I did the of Coder Radio. I never do a show on battery. I never do a show on battery. Well, you're a professional. Ever. And we got here in the morning.

You know, Coder Radio's at noon. I did the whole show. I did the morning prep. I did the show all on battery. Wow. And then plugged it in later, just so that way I wouldn't be inconvenienced later on. Not because I had to.

And, yeah, I'm just relaying, if this is something that's crossed your mind, if you've been wondering about where GNOME's going, what's going to happen with GNOME Shell, I noticed, too, on, like, that's crossed your mind, if you've been wondering about where Gnome's going, what's going to happen with Gnome Shell, I noticed too on RGNome on Reddit, there are weekly posts about what's going on with Gnome Shell's leg? Why is Gnome so leggy? There's constant discussion about

it. People are starting to see it more and more because there's more people using Gnome than ever before. Yeah, absolutely. And so more and more people are noticing this stuff, but you can go to Launchpad and look at the bug tracker for 1710 and look at all the stuff on there for Weyland and Gnome 3 and see that there's... I could be wrong on this because I haven't looked for a month,

but I think they're getting more bugs related to Weyland than they've ever gotten on anything. Wow. Yeah, it's not even an LTS release. I'd have to recheck the numbers, because this was when 1710 was kind of new. Yeah. But that's not a good sign. So, this has all given me pause recently, and it's something that

I've been talking to you guys about behind the scenes. Like, I am really worried that we are barreling down the GNOME 3 highway. Canonical is shipping it, which means Dell is shipping it. System76 is basing Pop! OS off of it. CentOS and Red Hat are selling to corporate customers on GNOME 3. And the more and more esoteric edge cases that we bring into the fold,

and the more desktop use cases that we bring into the fold, and the more pressure we're putting on Wayland, the more cracks we're seeing, not necessarily in Wayland, because Wayland is a protocol. It is all of the software that's attempted to write their own implementations. Yes, exactly. Yeah. And it's fundamentally bringing up architectural issues that we've been able to punt for years because X helped cover it up, or XYZ helped cover it up. Or no one was really using anything that used Wayland yet.

And so when you see projects like Elementary OS that have their own desktop or Solus that have their own desktop, you go, oh, that's sort of why they did that. That's sort of why they must have been hip to this. They didn't want to say anything. It's not cool to publicly shit on another open source project, but behind the scenes they go, people are going to get hip to this

and we've got to come up with something else. And the Plasma project, to their credit, has credit has been since the five release has been working at this problem and making their architecture sound and cleaning their house for, for release after release, after release, they're now getting into the feature. You know, they're now they're getting into piling the features back on, which that'll be interesting to see how that goes. Cause it's been a lot about refinement and now now Plasma feels like 5.13. Man, they're adding some cool stuff. You're going to have full Chrome integration

with notifications, so downloads in Chrome will show up in the Plasma notifications. You're watching a YouTube video? You can click the volume icon on the Plasma desktop. You can pause the YouTube video. We're talking nice features are coming. These are like Windows-level features.

They just recently introduced the Vault. Things are good. It's good stuff. And they're doing a good job of it. But they have sort of crossed that threshold of refinement, refinement, refinement to, let's try doing this now. Let's add this. So now they're going down uncharted territory too.

But they do seem to have a better foundation. So how long did it take for you to get that up and running customized and to a way that you could use it that's a good question being deep in gnome land which has a pretty easy and like well-defined two days if i could interject if i could interject as far as how long did it take him to use it zero he used it fresh because he you installed it i think you installed some programs like to get telegram and stuff but you had to do show. We didn't have time to customize it.

So he did show prep. He did the first day he used it stock. Just stock, yeah. And then you started tweaking it. Yeah, and in all of the installs, I also timed how long it took, roughly timed. Not super anal about it, but roughly timed how long it took before I had to install GTK apps and a bunch of GTK dependencies. Oh, yeah, that's another factor.

It's about, on average, about 15 minutes. Oh, because sometimes I just don't know. Like, I want a GDEB-I. If I'm on an Ubuntu system, I want GDEB-I. It's my favorite. It turns out there's, like, essentially a cute DEB-I alternative, right? But I didn't know that. I didn't remember.

It takes time to find those alternatives. But sure enough, so one of the things I've done, Wes, But I didn't know that. I didn't remember. It takes time to find those alternatives. But sure enough, so one of the things I've done, Wes, is I've gone, only you can see this, so I'm going to just tell you, is I've just gone with, I've got a lot of windows open. I've just gone with a really simple one bar at the bottom, just a couple of applets here, the clock, my task launcher, my virtual desktop. Pretty simple, pretty traditional. Pinned a few apps.

Yeah. So I'm not going with like lots of plasmoids i'm not doing you know i don't have menu meters all over the place and and you know stats and big slideshows on my desktop like i've done in past it's just minimal plasma setup keep it nice did a dark theme it's handsome i actually installed arc dark and your favorite then i went and set arc dark on the gtk stuff because they have in the plasma settings you can go in there and set your gtk appearance settings they make it really easy actually yeah yeah yeah it looks better than it did under unity 7 when so i so i'll show you right here you know so when you're in the studio and the machines are rebooted as you know oh this doesn't

have this is intel but as you know in the studio on airmaster when we reboot you have to launch the nvidia control panel go reset up all your settings. Right, so that way the capture cards work and get all the refresh rates right. And it is a bright white GTK application that doesn't match the desktop theme at all because it's being run as sudo or root or whatever. Sure enough, with Plasma Desktop,

when you set GTK appearances and you say prefer dark themes, every GTK app fits. So I actually have my GTK apps, because I'm using ArcDark in Plasma, and I have my GTK apps set to ArcDark, I can open these apps side by side. You just get one consistent look. If you took out Sherlock's magnifying glass,

you could find inconsistencies. The OK button on a Qt app is a different shape slightly than the GTK version. But it's the same shading. It's the same theme. It's the same color. It's the same title bar because, you know, I'm not using client-side decorations.

And I opened it and I went, look at this, Noah. Look at this. You can't even tell this is a GTK application. If I was an average user, I would be completely oblivious that there were two different toolkits on my screen. And that was really nice because it sort of gave me permission to use a couple of GTK apps and not feel like I'm paying a penalty. You don't have to bend the world to jump into this new place. You can just use the software you want. Yep. So that's where I'm at.

Plus, you combine that with Firefox has a Breeze theme, a Breeze Dark theme that you can install that kind of makes it look nice on the Plasma desktop. A Breeze Dark theme that you can install that kind of makes it look nice on the Plasma desktop. Oh, while we're talking about it, the other thing I did this install around is no Chrome. Just not even installing Chrome at all. I'm just going all in on Firefox. And Firefox already has some of that integration with the Plasma desktop I was talking about for Chrome. Firefox already got some of that. So you get that stuff today with the current version of Firefox on the current version of the Plasma desktop.

This is a whole new Chris. No longer GNOME and Chrome. I know. I know. I know. It's weird. I know. Meanwhile, Noah's over there installing Windows.

Not really. You know what's funny? You really have to be careful because I still, to this day, dude, I still get telegrams and emails going, dude, how's FreeBSD working out for you? Oh, yeah, from that April Fool's. That's amazing. The whole episode, we talked about how it was a joke. The whole episode, we talked about how it was a joke. At the end, we talked about how it was a joke. And there are still people that are like, I talked about it one time for like 15 minutes

and people are like, oh yeah, he's a BSD guy now. You're not going to actually think I'm installing Windows. I do have a backup plan, as JJ asked. My backup plan, if for some reason the Firefoxing doesn't work out, is to maybe go with Vivaldi. I've been hearing a lot of people talk good stuff. So right before I wiped my boxes, one of them I put out in Vivaldi.

It is great. Vivaldi, it is slick. It feels faster than regular Chrome. I love the way it shows loading stuff. If you're on Linux and you want a browser that feels like it's turned up to 11 a little bit, Vivaldi is pretty good. And they just implemented sync support. They're getting bigger, I think. A lot of people are telling me Vivaldi.

Have you tried Brave? No. Joe likes Brave. He's talked to me a little bit about Brave. I haven't tried it. Seems like both have gotten some good buzz. Yeah. Now, Brave's on the desktop too, right? Because I think it's available as one of the universal packages. Yeah. Bashful makes's on the desktop too, right? Yeah. Because I think it's available as one of the universal packages. Yeah. Bashful makes a good point about Netflix.

Actually, no, I watched Netflix in the new Firefox. That's happened. I've done that. I've been there. I specifically watched 35 Seconds of Star Trek Deep Space Nine in Firefox. Ooh, excellent choice. Just to verify that it worked. It's like, okay, good enough. I don't have time right now. Anyways, I should probably move on. But I guess this snap event, having seen people here that really know their stuff talking about architectural issues that they're not sure how to solve, that they're not even sure they have the leverage upstream to have upstream work with them to solve. the leverage upstream to have upstream work with them to solve.

It sort of underscored some stressing that I've been doing about the future of GNOME. When Canonical announced that they were switching to GNOME, I was pretty excited that we were going to have, as the saying, what is it, a lot of wood behind a single arrow? Because that's not something we commonly have in open source. I think that's something these corporate people would say. And I thought, great, this is a great opportunity. Right, but once we have some focus. But what really has given me

tremendous pause, and maybe this is wrong, I hope, and somebody can correct me if I'm wrong, but what's given me tremendous pause as I looked into it before this show, at least according to the known wikis, is they haven't even decided if they're going to address these architectural issues.

Wow. That to me is, that just gives me a moment of pause. When they're like, yeah, I mean, we don't want to break extensions, but we also don't like desktops crashing. Seems like too totally, the thing is too, and the things I find slightly unforgivable, and I'll leave it at this, is it's not like multiprocessor architecture was a new concept when Gnome Shell came around. Gnome Shell is a relatively new piece of software, and the problems that they failed to solve were well understood at that point. They made a conscious decision to architect Gnome like this,

maybe to cut corners. I'm just assuming, I don't know why you would design a modern desktop environment like this, but they did. They made that choice. And now, they seem to be almost willing to repeat the mistake again. And the worst thing about it is, this is my last point,

and really, we are so screwed either way. If they don't fix the problem, then we have a joke of a primary platform for our desktop environment because it's as bad as classic Mac OS at this point. And if they do solve it, they're throwing out all of this established momentum

behind GNOME 3, right as all of this wood is behind the arrow, and it's the perfect time to seize the developer DevOps, SysOps market, except for, once again, we're going to reboot our desktop environment. I would submit to you that we are far more screwed if they decide not to fix the problem. Yeah. I really think, because I really think, like, even if everyone comes onto Linux, and I agree, it's a really terrible time for, like, the third time in a row. So it's like we have a habit of really terrible timing. But at least

there is light at the end of the tunnel, because right now you're going to have a bunch of people coming over, and Gnome 3 is not great to begin with. I love it. No, no, no. I mean, as far as, like, if you sit down at Stock Gnome 3, it's not great. I've got to at least put minimize and maximize back on. And you know how to do that. A lot of people won't.

So we're already starting behind the curve, and now we're going to add to the fact that at some point in the distant future, either we're going to deal with a crappy desktop for a little bit, or a desktop with a huge problem, or we're going to have to reboot it somewhere down the road after everyone's already switched. Those are not good options to me. I would have Dan chime in, but I know we've got to move on, because I didn't want this to be a GNOME bashing thing,

because I love the GNOME project, I love the GNOME desktop, but Dan makes a good point in the Discord chat room that this is really kind of a consequence and a way of designing for Wayland and trying to either have this IPC system that is secure and safe versus just not even bothering with that particular problem.

I can completely understand that. Either way, and you know what? I'm still very appreciative of the hard work they have done, even if we're at this point. I do really appreciate it. And when it's working right, nothing sings to me like GNOME 3 does. It's really great. So I'll leave it at that for now.

But if you've been thinking about that, I'd say go give KDE Neon a try. It's a pretty great way to get your hands on KDE Neon. And it's one of the strengths, right? As part of the open source Linux community, we do have options, and it's worth our time to explore those and take advantage. Speaking of those options, let's wrap this up on

one of the favorite applications I came on when I was using Gnome. I was looking for client-side decorated apps, and I came across Tilex. Tilex? It's a replacement for Gnome Terminal. I thought this was like a shower clean. It's a pod that you can eat.

I was like, how is Gnome Terminal not client-side decorated? What is going on? This is their terminal. I wanted to find one that was, and I came across Tilex. As the name implies, it's a tiling terminal emulator that follows the gnome head. It uses

client-side decorations. It lets you split it into multiple, like vertical, horizontal panels. You can synchronize input across all of them. It supports persistent layout. It has very good transparency support, unlike the regular GNOME terminal. Drag-and-drop support, notifications

that can actually go into the GNOME notifications area, custom links, which are nice. You can click on hyperlinks and add custom terminal hyperlinks to open things in your terminal. And you can layer the panes, if you will, the West panes. You can put them in any arrangement

that you really like. Oh, nice. And so what I did is I also combined that with an extension for GNOME Shell that allowed me to always open up applications on a certain virtual desktop. So Tilex always opens up on virtual desktop 3, for example. That's your terminal desktop.

Slack and Telegram on desktop 2, and they're tiled next to each other. That's awesome. Yeah, and so there was a nice workflow when you combined that extension. So I always knew my terminal was always here. Everything in its right place. Yeah, I always have Guake I can go to for a quick install.

But when it's time to sit down and live in the shell for like 20 minutes or something like that, or I want to kick off a job and just check in on it with a quick virtual desktop dropdown, this was it. Tilex, you can find a link in the show notes because the URL is sort of crazy. But you can also just Google T-I-L-I-X, the Advanced GTK3 Tiling Terminal Emulator that has client-side decorations, which is really what I was just trying to all in on CSDs if I could.

It's also kind of nice. It feels sort of modern, and I feel like a lot of times terminal emulators feel like somehow Mac still has one of the best ones, and yet we are the terminal people. So that's nice, too. Somebody could tell this project about the good word of universal package formats, though, because they literally have packages for Anagros, ASIC OS, I don't even know what, Arch Linux, CentOS 7.3,

Debian, Fedora, OpenSUSE, Ubuntu, and then they just make a targes available if you just want to pull it down. Old school. Yeah. I love it, though. Of course, it's open source. They have it up on GitHub.

And you can also find it on hashtag termics in the IRC free note. Well, you know what? I'll be installing later today. Yep. That is pretty cool. And it kind of makes me miss GNOME a little bit. Kind of makes me miss it. But console is pretty solid, too.

Console supports some of the stuff. I'm just saying. All right. miss gnome a little bit kind of makes me jealous but console is pretty solid too console supports some of the stuff just out of the box i'm just saying all right well that's going to bring us to the end of this program because we got to get the heck out of here so that way the ask noah show continues on so if you're joining live don't go anywhere because the fun you just got through the first part of the sandwich really and there now there is the other half of the sandwich and some homemade potato chips oh yeah yeah oh yeah there's a nice aioli on there. Oh, yeah, maybe a little chipotle aioli, a little brown bear aioli. That's on there, too.

So go get more Wes Payne. He's at Wes Payne on the Twitters. That's right. And techsnap.systems for our other program covering all kinds of stuff, including some meltdown and Spectre issues. And like we said earlier, Kubernetes. You can get some more Noah in your face

at asknoahshow.com. That's Noah's main landing page. I think he has a Twitter account. It's at ChrisLES, I think. That's right. Stealing my Twitter. At ColonelLinux, the show, at AskNoahShow. The AskNoah dashboard, asknoahshow.com. I said that part. Oh, did you?

Once you get me started, I'm like, you know, I just... And sure enough, if Noah didn't pull off a live broadcast from the car on the drive down here yesterday. He did a live... He did like a 15, 20-minute live show from the car on the drive down here. So, yeah. That piece of gear he brought with him is pretty... Only because you said I couldn't.

Yeah, exactly. It seems like if we just strap that right to his back, He can walk around and just be on air all the time. Just give him time. It'll probably happen. The whole network is at Jupyter Signal. You can join us live, jplive.tv, jupyterbroadcasting.com, slash calendar for the live time. And last but not least, linuxunplugged.reddit.com

and jupyterbroadcasting.com, slash contact for all the ways to get a hold of us. We'll be back next week. Thanks for joining us. See you next Tuesday. Yeah, that's what's missing. See, I don't have the music. Tuesday. Do we have any titles?

What do you think? How to Contain the Code, Noah's Magic Box. Containing the future? No. That's kind of... I wouldn't click that. Future is a snap. A pod you can eat. The crafting of snaps.

Hey, I'm not at the studio, and I just want to make sure that this is actually going to work. So can you just say hi to me and just make sure that this is working, Sarah? Hi, yeah, I can hear you. Thank you. Okay, thank you. Yep. This is a pretty crazy setup we have. You're talking with Sarah.

We're talking to a mumble room. Those people in the mumble Room are literally all over the world and we're sitting here in Seattle in a hotel room right now. And there's like, what, at best a 2, 3, 4, 15 millisecond, 30 millisecond delay between us? Wild. Remarkable. And none of us are using a phone. Well, Sarah's probably using it. Is Sarah on it?

Yeah, so one of us is even on the phone. That's crazy. That is amazing, really. Confluence of technology. Alright, the only thing we can't do is pick a title, apparently. So we gotta do that, because the FNO show is what we're worst at. There's also stuff going on in Discord. Now that he has his own show, isn't that interesting?

Now that you got your own show, you turn around on the title thing. Let's see how it is. I see it. Give me shit. 45 episodes in, and I still didn't care about titles. And it wasn't until you were like, here's what happened. Because he's really good at this. I know it's gonna be bad. No, he's really good at this. I know it's going to be bad. No, he's really good at this. So he knows if he just comes out

and says, you're an idiot, you should do this, I'll be like, I'm not an idiot. Let me show you why I'm never going to do that. Never again. And he uses that in the opposite way, too. You can't do a live broadcast. We do a live broadcast. You won't be able to do it. So he tells me, I give him this idea, I'm bouncing it off. And he answers all of my questions exactly how he has them. You won't be able to do it. Yeah, okay. So he tells me. I give him this idea.

I'm bouncing it off. And he answers all of my questions exactly like he has. And doesn't say, don't do that, do this. Answers all of them just exactly the way. Two hours go by. Then I get a message. I say this with love, but I just want you to. That's literally what I say.

I just want you to consider you could do all of that work. Or you could pay some attention to titles. And description. And description. And it would make a really big difference. And you could just try it. Just see. That's just a thought for you or whatever.

And what's funny is, see, he knows how my mind works. And this just floats in over Telegram, right? So you can see this friendly message. I'm sure this is planned in his mind. In my mind, I'm just like, oh, that was a nice little follow-up. Yeah, that's a really great idea. And then when I go back and read it, I'm sure this is planned in his mind. In my mind, I'm just like, oh, that was a nice little follow-up. Yeah, that's a really great idea. And then when I go back and read it, I'm like, wait a minute. Well, I mean, it's your call.

I'm not like, no. But so anyway, so then I did. And sure enough, you start looking at the download numbers when you actually pay attention to the title. Turns out it really matters. The guy that's been a podcast for 15 years actually knows something about podcasting. Turns out he's pretty darn smart. Oh, well. Damn you, Chris.

The bill's in the mail now. Sprint, Snap, and Core. Here's what I came up with. I came up with peeking at Canonical and behind Canonical's curtain. I kind of like behind Canonical's curtain. Yeah, that's nice. It's got a ring to it.

Okay. We should probably just go with that so that way we can get out of here fast. No. But thank you guys for helping. If people aren't interested in Canon in canonical then they probably aren't interested in this episode that's a good point that's a good point mac os terminal is the worst i guess it doesn't have plasma or no minute that's the only thing but uh yeah yeah also it's not that

like we talk about that all the time yeah that's we literally do uh i almost didn't even want to mention the fact that i'd move back to plasma because i'm like jeez i people got to be sick and tired of this but you're all over the place, Waffler. I know. The only reason, I thought I was moved in. I thought I'd moved into Unity. I set up my lights outside. I got my rugs down. I got my couches moved in. I put my

big screen TV up. I thought I was set. And then I realized that my neighbors were assholes and that my foundation was leaking and that my roof sucked. So I was like, well, now what am I going to do? Move to an RV? Yeah.