LINUX Unplugged - Episode 270: Stratis Pulls it All Together

Episode Date: October 10, 2018

Red Hat developer Andy Grover joins us to discuss Stratis Storage, an alternative to ZFS on Linux and its recent milestone. Also Google subtracts Plus, some KDE and GNOME news, and a bit of forgotten Linux history. Special Guests: Alan Pope, Alex Kretzschmar, Andy Grover, and Martin Wimpress.

This is great for us now that we're really an audio show. There's a pair of sunglasses out there that would have made a great sponsor back in the day when we had sponsors. They block out all of the screens around you. And it's not actually all of that tricky. There is a company out there called Casper that makes, not the mattress company, I guess it's a different company, that makes a screen blocking film. that makes a screen-blocking film. And one entrepreneur, with a last name of Blow, I think, or Blue, I can't remember his last name,

he popped out the lenses in these cheap old sunglasses, put a pair on with this film over it, and then set off into the city and realized that he was blocking all of the televisions around him. All the annoying signs that use TVs, all the TVs in the bar that you're going to, and you don't care about that sports game on there. You don't have to see it anymore. So his name is Blue, B-L-E-W.

And they're the IRL glasses. Yeah, they've even formed a company, and you guessed it, a Kickstarter. They've got a Kickstarter. Wow, they've raised $5,000 since we went on the air. That's not bad. When I first loaded this page about a half hour ago, they were at $5,000 since we went on the air. That's not bad. When I first loaded this page about a half hour ago, they were at $75,000. Now they're at $80,000 with 1,207 backers

with 21 days left to go. You can see everything except screens, they say. Yeah, I guess, actually, anything that is just a polarized lens rotated 90 degrees and then flattened, can produce the same effect. So you don't actually need this one company's special film. So you can go play with this right now. And it won't block OLED screens.

If you ever try to lie on a sun lounger and read your phone with sunglasses on, sometimes you get the same thing. Have you ever gone swimming with polarized sunglasses on? It's pretty cool. This is one of those kinds of things that I wouldn't mind, but you'd have to bring a pair

for everybody. Like, you'd go sit down at a restaurant and you're wearing these sunglasses. It'd be distracting for everybody at the table. I think he used to it just a couple times. Oh, that's that weird thing Chris does. No, you've got to bring a pair for everybody. This is Linux Unplugged, episode 270, for October 9th, 2018.

Welcome to Linux Unplugged, your weekly Linux talk show that's drying the tears of those G Plus users. My name is Chris. My name is Wes. Hello, Wes. We do have some Google news to get into today, but not too much. We won't overwhelm you because then we have some big Plasma desktop news, some upcoming GNOME announcements, as well as a bit of a GNOME history lesson. We're going to go back to the good old days when Linux on the desktop wasn't a joke, it was the hype.

And it's actually where that saying, the year of Linux desktop, came from. We're going to go back to that era and look at some fascinating times that resulted in software that we're still using today, but you might not know its heritage. We'll dig into that. Then, later on the show, we've made no secret about it. We are big fans of Red Hat's Project Stratus. It's a tool, or a layer really, that's bringing a bunch of existing technologies together on Linux to compete with CFS and offer some of the same feature sets as ButterFS, but using established technologies like XFS, the DM layer of the Linux kernel,

some of the same exact tools that LVM uses today. And Andy is one of the lead developers behind Project Stratus. He's going to come on and give us the high-level pitch of what it does, what it is, and then answer some of our questions that get down into the weeds. And after that, I'm getting organized, and I am using this week's app pick to do it. I'll share the app pick for you that I'm using to get my life in order, for goodness sake. But before we go any further, we've got to bring in that virtual lug. Time appropriate.

Greetings, mumble room. Good evening. Morning. Greetings. Wake up and smell. Howdy, y'all. Love it. Hello, Mr. Badger, Mini-Mech, Popey, Sean, Spazzy C, and Wimpy.

It is good to have you in the virtual lug today. Let's just get the elephant out of the room. Big story this week. Google Plus is shutting down in the wake of a bit of a data leak. And maybe the bigger part of the story is that Google tried to keep it under wraps. And we're learning about this from a Wall Street Journal report that got memos from within Google. And the really damning stuff about this is these memos make it clear that leadership, such as the chief executive, Sander Pichai, was briefed on the plan not to notify users, which he signed off on.

This is a huge story, and it seems that there was a, quote, not to notify users, which he signed off on. This is a huge story, and it seems that there was a, quote, glitch in an API for Google+, between 2015 and March of 2018, when an internal Google investigation discovered and then immediately attempted to fix the issue. This is one of the reasons why they didn't make a disclosure about it. It wasn't that somebody came to them and then got a bug bounty. It was that Google found this during its own internal audit, which is commendable in some ways. The memo reviewed by the journal, prepared by Google's legal and policy staff, and then

shared with senior executives warned that disclosing the incident would likely trigger immediate regulatory interest and invite comparisons to Facebook's leak of user information from Cambridge Analytica. They did have some considerations, though. So whether we could accurately identify the users to inform, whether there was any actual evidence of misuse, and whether there were any actions a developer or user could actually take in response, they didn't find those, and that was one of the reasons they gave for not telling us.

Which is kind of understandable. They believe it was never abused by any third-party application. It is kind of understandable. They believe it was never abused by any third-party application. It is sort of a gray area. I mean, it is our data, and I would like to know about what happens with it and how well-protected it is, but are we really entitled to that?

It seems like for the benefit of the industry, it's good. You know, it's good to be clear about this stuff. The way Google handled this announcement, I think is a lesson in brilliant marketing. I don't like it, but boy, did they nail the messaging on this. So instead of coming out and being like, yeah, not only do we have a vulnerability and not only are we shutting down Google Plus and not only do we try to cover it all up, they take a totally different approach. Instead, they launched Project Strobe, protecting your data, improving third-party APIs, and sunsetting consumer Google+. So they announced this buzzword project, Project Strobe,

that talks about auditing third-party APIs and clamping down on Gmail access. And then, as almost to make their case, even though Google, up until this point, had claimed that it was a platform, Google+, they cared very, very deeply about, now they actually use its failure as an excuse to move on. They say the consumer version of Google Plus currently has low usage and engagement. Get ready for this one. version of Google Plus currently has low usage and engagement. Get ready for this one. 90% of Google Plus user sessions are less than five seconds.

That sounds like, oh, I clicked on a Google Plus link. Get out of here. That's exactly what it sounds like. So here's what was leaked or potentially leaked. Here's what was capable of being leaked if third parties took advantage of this API flaw, which we don't know if that actually happened. It appears that the data is limited to static optional Google Plus profile fields, which would include your name, email address, your occupation, your gender, and your age. It doesn't include other data from Google Plus or other Google services like your Gmail or any G Suite content. They say when they discovered it,

they believe that it was a result of subsequent Google Plus code changes that caused interaction with the APIs that led this to be possible. So they probably just didn't have that many people doing very much good work on Google Plus anymore and some botched update. They mentioned that the review did highlight the significant challenges in creating and maintaining a successful Google Plus anymore and some botched update. They mentioned that the review did highlight the significant challenges in creating and maintaining a successful Google Plus that meets consumers' expectations.

And given these challenges and the very low usage of the consumer version of Google Plus, we've decided to sunset the consumer version of Google Plus. Well, Chris, where are you going? What else is there? I stopped using Google Plus a long time ago,

but the reason why I wanted to include this on the show is, like, the people I have noticed that are still using it are people from the open source community. There are people. That's all that's left from what I can tell. It's photographers, Mike What's-His-Face, that used to be on Twit.

Elgin. Yeah, Elgin, Mike Elgin, and Alan Cox and other open source users. Linus used to host there. Like, that's who's left on Google+, right? Well, there's also a whole bunch of communities, inverted commas communities, that they call. The Ubuntu one is one of the largest on Google+. It's got 288,000 people in it.

It's pretty huge. Unfortunately, every post in there nowadays has a lot of spam in the comments because Google clearly haven't cared enough to clear that stuff up. And if I go in there and have a look at the admin interface, it's just like constant spam all the time. It's been a really mixed message from the very beginning. They say they're going to implement this wind down over a 10-month period, slated for the completion by the end of next August.

10 months is a weird time frame. Looking at my G Plus feed today, it's full of people saying, you can now find me here. People choosing where they're going to go instead of G Plus. And a lot of times it seems to be Mastodon, and Diaspora seems to be getting a little bit of resurgence here and there as well. I'd say the one I've seen the most actually is Twitter. But the second one I've seen the most is Mastodon. Seems to be a good moment for Mastodon.

Now, if Mastodon steps up and produces something or continues to produce something that people find compelling, they may have an opening now. Although I've thought from the day they launched it, there just wasn't a lot of room for a third social network, at least not here in the States. The States is a Coke and Pepsi country, right, left. Like we seem to get two brands and then we're good. We got Twitter and Facebook. Three is confusing, Chris. Then you got the Instagrams and you got the WhatsApp. It's just too much.

There was no room for Google Plus in its silly circles. There just wasn't any room for it. That's why they had to cram it into YouTube to try to force usage. But that didn't really work out too well either. You know, the real tell there was that Google's own executives stopped posting to Google Plus and started posting to Twitter about three years

ago. Seriously. As soon as I noticed that, I'm like, ah. They don't believe it? Not dog food. Yeah, that's not good. But I, as much as sort of I seem flippant about it, for those of you who got value out of it, which is probably someone in our audience, I'm sorry about your loss.

It's just one of those dangers of these closed services. If this had happened two years ago, I would have been beside myself. But it's been waning for a while, and certainly since Mastodon sort of caught people's attention about a year ago, or people within the open source community about a year ago. been waning for a while and certainly since Mastodon sort of caught people's attention about a year ago or people within the open source community about a year ago. A lot, it's very active over on Mastodon now. And I think that I'm, you know, I've been posting there where I haven't

been posting on G plus for a while now. I've resisted Mastodon just because I'm over, I'm really, I don't, I don't particularly like social media. I like chatting with audience people. So that's why I'm on there. And I like tweeting back and forth with audience members, but the rest of the people that are on social media, I just, I generally find them to be obnoxious. Like I just, I don't know. I don't like any of it. I don't like, um, I generally am not a big fan of group think. So when everybody starts freaking out about the same thing, I immediately kind of pull back and of groupthink. So when everybody starts freaking out about the same thing, I immediately kind of pull back and go, oh, what's going on here? And that's totally

common. The mob mentality is super common on social media. And I'm just not, I'm not into it. It bums me out. So I've pulled back a lot. I mostly just go for the list that I have now on Twitter, where I can follow my friends and my at replies and I can reply to audience members. and follow my friends and my ad replies and I can reply to audience members. Have you seen in England that this week there's been a program called the circle and my wife's been going crazy about it. It's a social media quote unquote experiment where they take like it's like big brother sort of where they lock some people up in a hat in a apartment complex and they use nothing but social media to communicate it's really fascinating they've had you know men pretending to be girls and girls pretend to be men and straight people

pretending to be gay and vice versa and you know it's really interesting to watch how the different factions within that group communicate and like like what you said people do they're very quick to pick up the pitchfork and form a mob and it shows how uh when you know a lot of them one of the comments someone said was it's easy to forget that there's someone an actual person behind the screen when you when you get on that keyboard it's yeah and i think we see that all the time in youtube comments in reddit comments in some of the dross that's on the internet. And I think things like Google plus and Facebook or whatever, because they were real people, they had somewhat of a slightly less

crappy signal to noise ratio, but they were kind of one of the loudest voices on use your real name, use a profile picture. And that was, that was kind of one of the bigger things they pushed. You know, just in terms of online discourse, I, over the years of doing these shows, have gotten many very, very upset emails or comments. And if I respond in kind and I go in all guns blazing, then usually it just derails. But if I respond to them in a calm, rational way, I have to be in the right place emotionally to do that. But if I respond in sort of a calm, rational way, really my experience shows me nine times out of 10, I get an email back that goes, oh man, I was just having a bad day. I couldn't even really think about the fact that you're a

real person too. Like I just was super upset about stuff. I've gotten so many of those emails. Like I just was super upset about stuff. I've gotten so many of those emails. And I've been getting those since before Twitter even existed, when Twitter was a podcast app that was called something completely different. And my point is when I use these social media networks, I see that same disconnect that I have had interactions with over email and now other mediums where people are not really wired up to fully contextually communicate this way. And it's another reason why we often come on this show and say, if you can

even just once a year go to a open source style Linux style community event and meet some of the people that you talk to over the internet in person, it will change that relationship. You know, and it, and the reason for it is I think that the meat sack that sits between our ears is a little more tuned to those types of interactions where it can get body language context and the tone of your voice and the context of the conversation and all of those nuances that help you make an informed decision about what the person is saying. And when it's all over text, we project a little bit. We read intention into it. We think, are they mad at us? Did I do something wrong? I made this mistake

where I sent somebody a file via Slack on Sunday. And it was just a co-worker who is down in Texas. And, you know, she doesn't work seven days a week like I do. So for her, she's getting these messages on a Sunday and she's like, okay, thanks, dot, dot, dot. And then that was all she sent. I'm like, oh man, did I, did I piss her off? I, I probably made her phone buzz on a Sunday afternoon and it's evening down there. And I start feeling like all these, all these waves of guilt. And I even thought about writing her apology Monday morning. I thought about writing her apology right then, but then I thought, no, that'll just send her more messages.

I shouldn't do that. That'll upset her more. I'll wait till Monday morning, and I'll write her an apology on Monday morning saying, sorry, I didn't think about your phone might beep when I sent you a Slack. All this is going on in my head within 30 seconds of sending her that.

It could be nothing, right? It could be just- She might've meant thanks. Right, yep. I mean, I guess it's not surprising. It is a new form of communication. So maybe it isn't weird that we're struggling with it. We're figuring this out.

Yeah, we're still trying to figure out like the guardrails and how to make it work. Yeah. So that's why I think it is very true that the context in this kind of stuff really matters. All right. Well, let's shift gears now and go from the heady community stuff and let's move into, and I guess one more thing to that is I think that's why things like Mumble stick around because people get more context in voice chat

and stuff like that. Okay, so let's move into some new Plasma features. Yeah, the good news of the show. Plasma 5.14 landed this morning and I've already updated my neon boxes. It's got a couple of great new features. I'll just mention a couple because I know not everybody out there cares, but for those of us who follow this stuff,

there's a new display configuration widget for screen management, which is nice. You plug in an external display, little widget comes up by your system tray and says, is there just a quick screen layout? Do you want to go into presentation mode, which I think is pretty neat. Presentation mode will turn your power management settings so that your computer and screen do not turn off during your presentation. And then it can enforce in certain conditions presentation mode over applications like Chrome. Pretty nice. And the audio volume widget now has a built-in speaker test feature, which I can tell you how useful that could be. Oh, yeah. And thankfully, the network widget now works again

with SSH VPN tunnels, which is slick right there. It's one of those things that you don't think about until you have one, and you're like, why is this not showing up? I am so thankful that it's so easy to do standard, like, PPTP VPNs and things like that. So getting an SSH tunnel is going to be great. And then just, what was it, one or two episodes ago,

we were talking about Discover, getting FW update support. So now, if you're on the Plasma 5.14 and you use the Discover software manager, you'll also be plugged into LVFS and get firmware updates for your rig, and it gains support for Snap channels, which, if I am correct, is that, maybe Poppy Wimpy, correct me, is that different types of, like, correct me, is that different

types of like the beta channel versus the stable channel of a snap? Right. Very cool. Yeah, this was a conversation we had when I was back at Academy a few weeks ago. Alesh is one of the KDE Discover maintainers. And yeah, one of the most desired features was the capability for people to switch channel from within the GUI, so they don't have to use a command line like a barbarian,

but switch channels directly in the user interface, and they've implemented that in the few weeks since I was in Vienna, which is really good to see. No kidding. Talk about some quick turnaround. Also, some improved support for Flatpak. If you try to install a standalone Flatpak file, but the whole Flatpak backend hasn't been installed, it'll now ask, hey, hey, genius, do you want me to install the Flatpak support for you? Very nice. And then KWin, continuing its march towards excellent Wayland support, there is a fix in there to correct copy and paste between GTK applications. Oh, good.

I am so glad to see that. Yeah, because if you're, say you had like G edit open and you wanted to copy between G edit and K write or K, you couldn't do it before on Wayland. So that's been solved now. And they considerably improved

and polished KWin effects throughout, including a completely rewritten dim inactive effect, which is a neat effect you can set in KWin when a window goes inactive. So you take three or four windows up.

Only the active window is lit up. The other ones can have a dim effect. They can even blur a little bit. They can go translucent. And this is kind of a neat way to help your eye cut through visual clutter. So you need a couple of windows up on your screen. Maybe you arrange things like on the OBS machine, I've got three windows open right now.

It's pretty easy to tell which one's active because the close and minimize buttons are lit up. And so that's how I can quickly tell which is the active window. However, if you wanted a more dramatic visual cue, you could have it so that the only application that was its regular lit state was the active applications and all of the other windows would take a slight dim effect, and that's been totally rewritten in this version of Plasma, including a new scale effect and rewriting the glide effect when windows glide around on the screen. Just lots of little polish.

Love it, man. I just am so happy with it, and rocking the Neon 1804, and the updates came down this morning. It's been... I knew what I'll be installing tonight. And I notice, I notice almost every day now, there's some thread in RLinux or on some blog post about somebody's just discovered Plasma.

The popularity's real. It's not just us. Okay. Nobody has anything to add to that. Then let's change gears to... I do. I love KDE. You know I do.

I say it every week. Yeah, and let's do... Okay, well then, if Alex is going to chime in, then we have to check in on Popi, and Popi's laptop that was installed in the air running Neon at the time. Is that installation still standing, Mr. Popi?

That's the same installation, the one that I upgraded to 18.04 just last week. Boom! still stands. All right, ladies and gentlemen, that is your Popi Neon laptop update. We need a jingle right there. All right, now shifting gears to GNOME. There's going to be some changes coming down the road,

probably in GNOME 3.32. The application menus, or the app menus as they're often called, are the menus that you see up in the GNOME 3 top bar with the name and the icon for the current app. Well, these menus have been with us since the very, very beginning of GNOME, and they're going to be retired in the next released version of GNOME, GNOME 3.2. So no more app menus up in the top menu bar of GNOME Shell the applications

themselves that have menus like when you hold down alt for like file and edit those can remain but the ones up in the shell bar are going away

like one little area you could click that was never used very much anyway sometimes headquitter preferences in there they feel and this is

this is from Alan Day's blog post he writes when app menus were first introduced, they were intended to play two main roles. First, they were supposed to contain application level menu items like preferences about and quit. Secondly, they were supposed to indicate which app was focused. Unfortunately, we've seen app menus not performing well over the years despite an effort to improve them. People don't always engage with them. Often they haven't realized that the menus are interactive or even remembered that they are there.

I'm not too surprised by that. I'm guilty of that myself. Now, he writes, we're planning on removing application menus from GNOME in time for the next release, version 3.32. The application menus will no longer be shown in the top bar. Each GNOME application will move the items from its app menu to the menu inside the application window. If an application fails to remove its app menu by 3.32, it will be shown in the app's header bar using the fallback UI that is already provided by GTK. This means there's no danger of menu items not being accessible

if the app fails to migrate in time. I think I'd love to hear Wimpy's take on this, just because he's our number one Ubuntu Mate guy right here. Boom, right there. And Ubuntu Mate is kind of around because a lot of people don't like the changes in GNOME. I wasn't sure how far I could take that. So, Wimpy, what are your thoughts on these kinds of changes?

My first thought, and I'll let you take it from there, was I wonder if there was a discussion about how to improve them and make them more useful, or if the conversation immediately went, let's remove them. Who can say I wasn't part of those conversations? Obviously, GNOME have their own design direction, and they've iterated on that over many years now. And they've also not been frightened of going back from a decision they've made previously.

So I think, you know, we should let them do the things that they're the way they want to do it because if gnome isn't a desktop interface workflow that you are comfortable with or like then good news linux has got plenty of other desktop environments to choose from and i'm sure there's one out there that will suit you um more appropriately hmm i'm reading between lines there a little bit. Not really. Honestly, from my point of view, I don't actually care what Gnome decide to do or don't do. It's their project.

It's down to them and their designers to set the direction that they want to go in. And if some people love that stuff and it draws people to Gnome, then great. And if some people love that stuff and it draws people to GNOME, then great. And if some people don't like it and it pushes people towards KDE or Marte or something else, then that's fine too. You know, there's still plenty of well-supported desktop environments out there and all the major distributions. So yeah, just find one that suits you. Okay. That does make sense.

Now, here's what the counter-argument would be. The counter-argument would be, well, now, Wimpy, it's shipping on all of the major distros by default and new users coming to Linux, Wimpy. This is what they're experiencing out of the box. And so we must defend this pretend user that we're making up in our heads for a straw man argument. No, they're not.

New users aren't coming to GNOME 3.20. They're coming to Ubuntu 18.04. They're coming to Ubuntu 16.04. That's where those users are going. It's only the absolute lunatics who are using Arch or the tiny niche who are using Fedora or the Germans who are using Sousa who are getting these weird new versions of GNOME. Your average user is using Ubuntu 18.04

and they're using the user interface that you know and love today. Yeah, and I think really what you're saying there is that there is another layer of catch here where if something was to radically change in GNOME, there's the distributions that still can have vendor patches that can apply to it to restore that functionality. And we get flack for that. You know, people moan at us because we hold back versions of Nautilus or other things, because we have a large number of users who care about these things. And they tell us when stuff is objectionable. And it's all very well designing something for the future,

but we've got people right now who are using these systems, and they need to use a system that is familiar and intuitive and easy to use and all those other UX terms. We'll see what this looks like in 2004 when the next LTS comes around. It's interesting. I often don't object personally to a lot of the changes GNOME makes, but it's strange to see a project

that feels like a stable desktop that is changing so much and removing features. Most change I think that we see is accretive, right? We're like, oh, Crit, I got a new feature this week. That's really cool. GNOME does it differently. Yeah, it's simplifying, though. It can be a good thing.

It can be an easier code base to maintain. It can make moving to new display technologies easier. It can make re-architecting your desktop shell easier. So there's reasons to do it. You're right. They're not doing it for the fun. They're not doing it because they want to annoy their users or they don't want their desktop to be pretty and usable.

They're doing it because of perceived benefit to the user and the developer. So it makes total sense to do these things. It's just frustrating for users who have come to love icons on the desktop or menus in the menu bar or whatever it is that's being modified. It just frustrates users because we're all resistant to change. But you go back to what GTK 2 and GNOME 2 look like. I was going to say like five years ago, but I meant five minutes ago in Ubuntu Marte. But, you know, you look to what it looked like. People have moved on, like some people have.

And those who haven't, there are other alternatives for them. Well, why don't we go back in time a little bit? So you heard Popey there mention Nautilus. The history of Nautilus is fascinating. And I recently came across someone who had front and center experience with it in a completely unexpected environment. In a book that's out now that I'll have linked to the Audible version of the show notes called Creative Selection. And it follows someone named Ken Coesita, I think is how you say his last name. And he is a software engineer who has been involved in the industry for a really long time.

And he eventually wound up at Apple and he created the keyboard for the iPhone and the iPad and the Safari web browser. And his story about trying to get Mozilla's source code for Firefox back in the day to build is great. How they discovered Conqueror and then decided to go with WebKit, like that whole story is in this book and it's fascinating. But before you get there, there's a little history lesson about a small little company that you may have heard of before named Easel. Easel was founded by Andy Hertzfeld, who's also an Apple alumnus, in 1999 in a little town called Mountain View, California. It had 22 employees and had managed to raise about $12 million. That's not bad for a small crew. And in Ken's book, he goes into some of those early days at

Eazl. Before coming to Apple, I had a job at a startup called Easel. Our goal was to create an easy-to-use Linux system suitable for everyday computing, a free software alternative to Apple Macintosh and Microsoft Windows. The company was led by programmers who worked on the original Macintosh in the 1980s, including Bud Tribble, the first software manager for the Mac, and Andy Hertzfeld, the software wizard whose graphical user interface code helped to set the Mac apart from the text-mode personal computers that were the norm of the time. These fellows were my heroes, and I joined the company to work with them. The elegance and simplicity of their Mac software was my main inspiration for wanting to become with them. The elegance and simplicity of their Mac software was my main

inspiration for wanting to become a programmer. The inspiration for Easel came from Andy, and his vision for the company was fueled by the free software movement Andy identified with Stallman's idealism, and by the concept of developing a file and icon manager that would make Linux a fitting competitor to Windows and the Mac, its more established rivals. So their core product was really going to be based around what we call today GNOME Files, but GNOME Files is based off something that we all refer to as Nautilus. Andy called this program Nautilus, and it would help Eazl users find files, read email, launch programs like word processors and spreadsheets, and perhaps do cool new things like keeping track

of a few digital photos. Eazl contributed Nautilus to the GNOME project, a loosely confederated free software community whose members, both individuals and companies, would be providing the rest of the software for the desktop computing system we were trying to build. To be a part of GNOME, Nautilus had to be licensed under the GPL. This had important implications for Eazl as a commercial entity. Since people would be able to download Nautilus for free once we finished it, the company had to figure out some other way to make money. Not surprisingly, this involved creating proprietary software that Easel could charge people to use, a set of proto-cloud services including automatic software updates and online file storage. These cloud services would live in an Easel data center and would not be free.

The idea was to integrate Nautilus with these services and position our no-cost software as a lure to draw people to Eazl's pay-to-use features. The combination of dot-com fervor, enthusiasm over Linux and free software, abundant venture capital money, and our founder's connection to the Mac made it seem to me that Eazl might just be the next big thing. So you have to put yourself in the mindset of the time. This is the late 90s, early 2000s. Linux is powering the dot-com boom, and the hype around Linux on the desktop was really strong. And it was where that joke originated. It's the year of the Linux desktop.

That's where this comes from, and that's why as it goes on, it gets to be even more of a joke. You were doing some searching, though, during the clip, and you can still find references to Eazl in GNOME file source code? Yeah, Wimpy pointed it out, but it's all over there. Yeah. Yeah, it's pretty fascinating, and so they were like, we can't lose. We've got this cool new, what

he refers to in the book as proto-cloud services that we didn't call it the cloud back then. It was just online storage. It was a better day, that's for sure. Yeah. And they had some other really cool, like, update management software. And they thought, okay, this is going to be it. We'll create this software.

People can manage the updates across multiple Linux boxes. When they run out of file storage, they can save to our cloud storage. They didn't call it that, which we'll have integrated in with the file manager. And it's just going to be great. There was just a few problems. Number one, people didn't have that great of bandwidth. Number two, they fumbled the opportunity. But if indeed we had this chance, we fumbled it. We never lived up to any of our lofty goals. Chief among our missteps was failing to conceive of our software as a single product instead of as a set of separate projects. We never figured out how to integrate the pieces.

Nothing worked smoothly. Our software update feature was riddled with bugs that often broke programs while trying to update them. Our code to connect Nautilus to our cloud services didn't work at all. The Nautilus team our cloud services didn't work at all. The Nautilus team had persistent problems coordinating with GNOME. The loose structure and lack of profit motive of the free software community meant that they did not share our money-making goals or care to coordinate with us so that we could meet our delivery schedules. All these setbacks caused delay after delay. Yeah, and things didn't go too well.

After Eazl launched Nautilus 1.0, the company then laid off 40 of its 70-member staff, and then things kind of spun down from there. The company didn't really manage to pull it out, despite even getting a substantial investment from Dell in December of 2000. The company failed to successfully monetize or secure more funding before their venture capital ran out. And things had changed a lot in that dot-com boom bust time. You know, by March of 2001, things just were different. And the market was beginning to shift. Things just were different and the market was beginning to shift. But the sort of silver lining of all of this, not only did all of those folks go on to do bigger and better things, but we now get GNOME using stuff that was created in 1999 and has been refined and refined and refined over time.

And it's been changed a lot, but it still has that heritage. And it was kind of ahead of its time in a lot of ways with its cloud storage solution. I remember Ubuntu had one too. Ubuntu 1 was originally going to be a cloud storage and syncing solution. That was a big thing that people wanted to get into for a while. And it took really until Dropbox in that modern era to get things right. But that GNOME files goes way back. It is neat that it's still around and it's still useful.

It's not a bad file manager. Yeah, yeah, yeah. As long as you're using the Ubuntu version 1, apparently. Heyo! Alright, well, so there you go. There's a little Noma history. If you have a project that you would think would make really good for us to go back into, especially if there's like

some audio we could pull from a YouTube video or a book somewhere out there, let us know. Linuxunplugged.com slash contact. That was a proto-segment, sort of like a version 00.00001. Put it out at the wall, see what it was like.

Yeah, we'd like to expand that later on and do something a little more and do projects that people are fascinated about that would apply to more than just GNOME users and things like that. So if you have a suggestion, linuxunplugged.com slash contact. And let me check. While I'm doing a little bit of admin work here, going to bring up the jobs page.

It looks like there are still two AWS. Oh, oh, no. Only one of each now, it appears. So there is an AWS training architect position open at Linux Academy, and there is a Microsoft Azure training architect position. Both are full-time with full benefits and all of that, and are remote. So I'm going to put a link to that in the show notes.

I've got to remember to do that right now. I'll put a link to this in the show notes if you would like to apply for those. They're also looking for a full-stack Node developer, a Ruby developer, which is also remote, finance controller. There's several jobs actually open. Man, that Linux Academy.

Growing like cray cray. Moving and shaking. Yeah, they are. So I will put a link to that right above the Stratus stuff. So if you're interested, go check that out. LinuxUnplugged.com slash 270. I figure we get the audience hooked up with the good stuff. You know what I'm saying?

Looking up for the audience. All right. So let's move forward now. So we just looked back at GNOME past. Well, let's look forward at Linux storage in the future. Now, for a while, we've been talking about Project Stratus, which is an effort that's happening over at Red Hat to bring together various components of Linux's structured subsystems for storage. And there's several of them that exist already, including well-established file systems like XFS that I'm a huge fan of, LVM, which relies on a bunch of fundamental technologies that we're all pretty familiar with. Stratus is a new approach

to constructing a volume managing file systems, whose really its primary innovation that Stratus is bringing, and what's going to kind of make it come to market pretty fast is it's reusing existing Linux components. And it really has my interest piqued because, A, it's based on XFS. Oh, you love it. You're already there. You're already there. And B, it's promising to try to drive feature

comparity with ZFS and ButterFS. It's pretty fascinating, and this last week, they hit version 1.0. It's a major milestone, so we thought, well, let's bring one of the lead developers from Red Hat on the show to talk about it. We asked Andy Grover to join us. So joining us right now is Andy Grover So for the folks that aren't familiar with it, give them the elevator pitch on what Project Stratus is. Well, I think we have a couple options for advanced storage features on Linux, but I don't think any of them, for various reasons, have really necessarily hit all the sweet spots. So I don't know if Stratus is going to hit all the sweet spots, but I think it's going to fill in. It really is meant to provide an easier way for people to manage their local storage so they don't have to worry about like

extending partitions or, you know, anything like that. It's just, you know, it kind of does all those things for you. And so you just get the benefits of these advanced storage features. So you just get the benefits of these advanced storage features. You know, there's like ZFS and ButterFS. They have snapshots and they have built-in, a lot of built-in features that we want to make easy to use for everybody. So that's kind of where we're trying to come from. What about some of the projects out there right now? I mean, ZFS, ButterFS, or even some of the up-and-coming things like BcashFS,

what about those made them not really seem like a good fit? Why a new project? That's a great question. Well, I think coming as a Red Hat developer, I think ZFS is just, the licensing terms are a real problem for the powers that be to go with that.

Not that other distros haven't said, hey, okay, we're going to go for it. But, you know, I think Red Hat and Fedora, you know, very conservative when it comes to like patents, codecs, and the licensing. So that kind of ruled out ZFS as something that we could do. Butterfest, oh boy, Butterfest. Oh boy, I can't believe it's not ButterFS. Yeah, right? I mean, it was in tech preview in RHEL 7

for a long time before it got pulled. I'm sure you talked a lot about that. So it just hasn't quite gotten there yet. I know some distros are shipping it, but even Ubuntu, they're shipping ZFS. So I don't know. It's just like it's not quite what we need it to be. It hasn't passed the test, the sniff test, essentially.

It hasn't quite got enterprise-grade stamped yet. Well, I don't want to say that it's not right for some people, but it's not right for everybody people, but it's just it's not right for everybody. So fair enough. But again, I think the core of the question is trying to fish out why not launch a 10-year project and build the next revolutionary file system for Linux? Why was that not the solution? Well, I think you kind of answered the question right there.

Time is imperative in this. Do you feel like maybe Linux is a little behind say what ZFS is capable of and so we needed a practical solution that was bolting together existing tech? I think that there were some and this has happened before with Linux. I think we were behind early on in our

USB support, and we kind of like other OSes kind of highlight some deficiencies, and then we kind of like rally in effort and catch up. So I think we're a little bit behind, and it's not that you can't do everything on Linux. It's just, you know, LVM, I mean, a lot, I think we'll talk about this more later. I mean, LVM and Stratus are both kind of leveraging the same core kernel capabilities, but LVM maybe isn't necessarily as easy as it could be to use. And some of the other options aren't

maybe an option either. So I think there's space for Stratus to come in and support these things and get there and have it in two years instead of 10 years. Yeah, that seems like one of the key things there. So what are some of those things? You touched on this just for a second there. It sounds like Stratus would be using some of the same plumbing that LVM's been using for years. What are some of those existing well-established components of Linux that Stratus takes advantage of? Well, I think people a lot of time, if they've used LVM, there's LVM and then there's this thing called device mapper. And device mapper is the kernel

code that LVM uses. So you can use device mapper through LVM, but you can also use device mapper in other ways. And device mapper, learning how to use Device Mapper for me was really like a kid in a candy shop. Because, I mean, there's so much there. There's so many capabilities. There's RAID. There's thin provisioning. There's integrity.

There's all these fringe Device Mapper targets that different people have written that do different cool things. And all of those things can be put together in different ways and used in different ways. And you can use LVM to do that, but there are also different ways that you can put the Lego box together. So that's what we're trying to do with Stratus and make those available and more friendly to use. It really seems in that sense that device Mapper is doing a lot of the magic and you've just been able to tie it together and add on a little bit of niceness with XFS on top.

A lot of it, yeah. So I mean, one thing that Stratus 1.0 supports is thin provisioning. So this is something that lets you create file systems and they don't actually use the amount of space that they're actually using, and you can have a pool that's shared amongst different file systems. Stratus supports this, but it's like we haven't had to implement it. The other smart people who have implemented the thin provisioning support in Device Mapper, we just get to make use of that.

We get to make it easier to use for people. And that makes, I mean, that saves a tremendous amount of effort to leverage that. And like you said, to leverage, you know, XFS file system, we don't have to write a file system. We have a, we have a great file system that we can just pick up and use. So. Amen to that brother. Yeah. Amen. So speaking of great file systems, one thing I've always enjoyed about ZFS is, by and large, I can kind of move things around. If I've made a data set on one machine, as long as it doesn't have a lot of crazy feature flags,

I can use that on another machine. How's that going to work for Stratus, especially if different kernel versions, different user space tools exist? Will it be portable? I think it won't be entirely. If you're, it all goes back to the things that Stratus is kind of wrapping and putting a bow on.

So it's like, you need a file system, you need a kernel that supports XFS. Well, they're all, you know, that's all of them at this point. You need a kernel that supports the device mapper targets that we're using. So that's then provisioning and cache target. And that's pretty much all of them too.

But there's a lot of other device mapper targets on the horizon. One really exciting one is called VDO, which does block level deduplication and compression. So if in some future Stratus version, we incorporate support for video, then that would mean that, you know, we'll have checks and stuff to make sure that

if you then move that pool, that file system, to an older machine that doesn't have support in the kernel for that, that's a consequence of, you know, if it's not there, then it can't work. So that's kind consequence of, you know, if it's not there, then it can't work. So that's kind of, that's kind of where we are. Okay. I think that's probably to be expected, especially at this stage in the project. So let me ask you something else that's going to be a little more challenging to answer. If I were to have a conversation with say a ZFS expert, they tell

me that what part of what makes ZFS so dang great is that it's total integration from top to bottom. In fact, there's such fans of it. Andy, they'd say, don't even use one of those dirty RAID controller cards. Don't put a dirty hardware RAID in there. Let ZFS control the entire stack because of its integrated brilliance. And then Andy comes along and says, actually, we could do this with a bunch of different existing technologies and have something that works great.

What's your response to the it has to be integrated or it's going to be bogus? That's, yes. It almost feels like a religion war when I hear them talk about it. It's like it's a way of believing that like it's a handcrafted integrated solution. But in a way, Stratus feels more like a genuine old-school Unix solution. I think that's a good point. I mean, so Unix is about, like, reuse and combining, you know, things that do one thing into things that do multiple things. So, you know, ZFS has been called a rampant layering violation, and I think they did it

for good reasons. There are advantages to doing it that way. Um, some, I think that, um, I'm not sure if all those advantages have actually been, been fully realized. So I'm, I'm pretty optimistic about what we'll be able to achieve, achieve. Like, like one of the, one of the things is like, like if you have a disk that fails on a ZFS pool, it's like you can track like, or excuse me, if you have like a sector that fails on a disk

in a ZFS pool, it's like you can track all the way up through the, through the ZFS data structures to like exactly the files that are, that are affected. So that is going to be harder for Stratus, but I think it still could be possible. And I think that there are other advantages that having a more layered approach give us that can outweigh that.

So looking at some of your documentation, it really became clear that the API is a core part of Stratus and that there's a lot of design around making that like a prime feature. Why is that? Well, I think if you look at the state of computing now compared to 10 years ago, you know, we have this, you know, kind of, if you want to call it a renaissance, you have this move to the cloud, you have orchestration, you have Kubernetes, you have Docker, you have Puppet, you have Chef. I mean, individual machines are less configured by people and more configured by

kind of layers of automation on top to achieve more productivity. So I think it was very important for us that Stratus be able to work within an automated environment. And I think having the API is very important to making sure that that's possible. I mean, I'm grateful already just if I don't have to do any more text scraping of command line utilities. If there's a real API, oh boy. No kidding, right? I'm really not a fan of scraping of command line tools. And that's, it's, yeah. So I really feel strongly at that.

And that really kind of made the API be something that we really wanted to focus on. And I think that's going to be one of the leading factors of adoption for this. I think that's what's going to drive other distributions to adopt it. So let's talk about that for a second. I bet the intention here is to have this to be more than a quote-unquote Red Hat project. It'd probably be nice to have it be a general community project. That would be great. How is that possible? What thoughts do you have

there? I'm not asking you to solve the problem, but do you have thoughts around opening it up to a wider community? Well, I think it's kind of a if you build it, they will come sort of thing. Just like practicing as much of our development in the open and on GitHub as much as possible. This ongoing discussions and IRC,

that's what we can do and publicizing it and kind of stating our desires to work with other people. And I think it always, like like you kind of have to like have something and then people see that and you know they want to they want to work on something that art is not going to their efforts aren't going to go to waste so i hope that with 1.0 um we've shown that we're serious and that we'd like to work with other people and and and build it and the more the project goes on, like the more possible new things and interesting things that people could hook onto

and make their own in the code base. I think the number of those things has really gone up. So I'm hopeful that people will do that. It's a little bit difficult because we chose, so we didn't write it in C. We decided to write it in Rust. Yeah. And there's a lot of in c we decided to write it in rust yeah and there's a lot of reasons why we decided to do that that we can talk about but i'm the the big concern was that that was going to hinder our adoption of that was going to kind of push away developers a

little bit i hope it doesn't and maybe it'll it'll attract other potential developers. So it was like developers and then getting onto other distros that it's an additional build dependency that is there. So I'm hoping that things will work out. Right. I think the choice to use Rust was a really fascinating one. I think that's actually something

I've heard people mention about the project. So it's created some hype in some circles at least. So maybe it's done some good as well. Yeah, the little rust circles, they're definitely behind us. Yeah, passionate. Okay, well, 1.0 is obviously a big deal symbolically. Does this mean I should be using Stratus? Are you recommending it for general use or even just for playing around? Or should we hold off a bit? Mending it for general use or even just for playing around?

Or should we hold off a bit? I think 1.0 is important for a couple big reasons. You know, like putting a stake in the ground and stabilizing our on-disk metadata. I think that was important. Let's just say we're working on a 1.01 release right now that's going to fix some pretty serious issues. And we're also working on a 1.1 release. So I would definitely hold off. Even adventurous people, just give us a couple more months,

and then the less adventurous people, they can hear what the adventurous people say, and then they can make a decision at that point. Is the goal to get 1.0 of some form shipped in Fedora 29? Yeah, that's the goal. We're a Fedora feature. 0.5 was in Fedora 28. Oh, yeah. And was pretty rough.

But, I mean, it's more about, like, you kind of need these milestones for yourself and for people who are following the project to kind of understand where the things are. So 1.0 will be in Fedora 29. I think that's going to be probably the best place for a lot of people to try out Stratus. I heard you say we a few times. Can you give me like a rough idea of like how is the project structured? Is it more than just Andy? Is it a few people? Is it a whole division within Red Hat that has its own building?

Okay, right. But maybe that's a little overboard. But Hat that has its own building? Like what kind of – okay, all right. But maybe that's a little overboard. But like what's it look like? Well, we have a – our core team is four people right now, including myself, all Red Hat people. And we've also had a number of kind of what you could call like drive-by contributions from a number of people fixing up things. So, yeah, there's four of us at Red Hat currently working full-time on the project, and hopefully we can attract some more small contributions,

and then those small contributions can lead to larger contributions. Hmm. Well, Andy, I'll have links to the project page, your Twitter. Is there anywhere else you want to send people to that might be interested in the project page, your Twitter. Is there anywhere else you want to send people to that might be interested in the project? Well, we had a G Plus page before they killed G Plus.

Too soon, too soon, Andy. I'll also just, yeah, make sure you put the Strata Storage Twitter. I think that's the primary one. Okay. We will have that in the show notes as well. Well, Andy, I hope you and the team keep up the great work. We're pretty excited to watch where this project goes, and thanks so much for

coming on the show and explaining some of this to us. Maybe we'll chat again in the future. Hey, thanks for having me, and have a good one. It's going to be an XFS future, I think. At least for a lot of us. Not like ZFS is going anywhere. But it'll be nice to see this sort of become an option for Linux users, I think.

I'm delighted to hear they've based it on existing technologies for the most part, and most importantly, XFS. I'm looking forward to the thin provisioning. I was talking to Wes about that. I have a one terabyte MVNE in my disk, or in my laptop. It's a one terabyte disk. And I am really looking forward to just saying, make home a terabyte, and I'll just figure it out later.

I have to look at the space eventually at some point. And LVM was always too much to manage. This is going to be nice. Yeah, you have a simple, I mean, we've already tried it. The command line interface is pretty clean. It was easy to get set up. It's easy to use. It's fast. So there's not a lot of blockers once you actually have a package to install.

Alright, well we'll keep an eye on it, see where things go and we'll have all the links in the show notes where you can read more, including get diagrams about how this thing works and what layer it's sitting in the Linux stack and all of that is there. Also, for your reading enjoyment, this is

a great benchmark that Wes found over at scalegrid.io, and they decided to compare XFS and Extended 4 on their AWS EC2 MongoDB instances. How about that for a title? And you know what? Both are great file systems is the long and short of it. And if you have a moderate to, you know, average rig, you're really not going to see much of a difference in terms of performance. Probably just doesn't matter, right? Pick the one that is easiest to use. However, if you have a significant workload

or you have a particularly fast rig and fast disk, well, this is how they summed up the difference. I thought it was pretty good. They say XFS is spectacularly fast during both the insertion phase and the workload execution. And they say down in the like really pushing type workloads that XFS maintains its lead and its throughput better than Extended 4 does.

And in general, in performance terms, XFS is indeed a force multiplier when paired with high-speed disks that can take a real advantage of a high-end PC. It is a force multiplier, they write, Wes. A force multiplier. XFS is just great. I just really love it.

That's why I was so stung when those guys and gals over at Dropbox announced that they were ending support for anything. It was rough. It was rough. It's such a good file system. I'm not over it yet. I'm not over it yet. Replacing Dropbox is one of the many things on my to-do list.

I got a lot going on. You know, living the VP life these days, Wes. I got wound care. I got a lot going on. I got three kids, Wes. You don't even sleep. That's how busy you are. This is actually true.

This is just a lot going on, and I need to manage all of it, reclaim some of my sanity. So I've decided to get my crap together. So you got just a big wall of Post-it notes? I assume that's what you're doing. I actually did start with a whiteboard. Realized that did not scale. Also was not portable and could not be done on the fly via my phone. So then I wanted something that would let me manage my tasks from the phone, from a lady tube, and from Linux. That's a big ask. Those are some integrations. You know, there's some really nice to-do apps for every phone out there.

They're really great, but they're like only available for the phone. Like they don't have a web component necessarily. I can't. I need something on the desktop. Or the ones that do like Wunderlist have never really stuck with me. It's never really held me down. It doesn't work with me. Well, I'm pretty pleased so far with Todoist.

And I've linked in the show notes it is an Electron app, but it is a wrapper for the official Todoist web app, and I tried a couple of them. There's a few of them out there if you search for this. But this is the one I like the best. It'll support auto-start and minimize down to the tray with a system tray icon. It has keyboard shortcuts that work more like the native application. It has just a native-like desktop feel.

It has the quirks of a web app, but they've gotten it pretty close. And it comes available as a dev or it's in a lot of different repos. Yeah, so it's easy to get going. And I've combined this with the app on the phone and the app on the LadyTube. And now I've got sort of ubiquitous task management. I can summon Siri or the Echo and I can say add a task to my list. I can sit down and I just have a Workspace 5

dedicated to this application. It's the only thing on Workspace 5. So I can go right to Workspace 5 and make it to do boom, boom, boom as fast as I can think about it. And that's key, right? You need something. You need to be able to dump them

as you come up with them. Here's the other thing that's great. It has natural language support. So if I need to, like, say, write down a task to finish the expense report for that lunch we went out to on Friday, I can write, finish expense report Wednesday at 1 p.m.

And it will automatically just, it'll understand that, and it'll create the finish expense report and at 1 p.m., and it will automatically just, it'll understand that, and it'll create the finished expense report, and then the Wednesday, and it'll itemize all of that out and create a reminder. Yeah, and put it on the calendar. It's nice. It's really smooth.

So getting a version on Linux that felt close to native, that had a tray icon and all of that kind of stuff, like a full-fledged application, that's what sort of pushed it over the edge for me from some of the other ones. So we'll have a link into Todoist for Linux. There's a lot of them out there, but only one of them passed the Chris Niff test. So you've got to check that out. I mean, I know you're picky about these apps, too, so you're pickier than I am.

That's probably a good bet. It's worth checking out. You watch out. Don't get me too hyped, because if I go all into it, then I'll sign up for the enterprise version, and I'll be getting you to install it. All right. I like it. You say that now. You say that now. It's pretty cool.

It does have good team management features. And the other thing that kind of made me want to use Todoist over some of the other quote-unquote services or self-hosted solutions, I want something that is a to-do application for a spectrum of my life. Like, it's not like a full spectrum to-do management. It's kind of like a narrow slice of the things I need to do in my work and personal life. It's kind of a weird way I have it.

And so I'm using Todoist for that. And for this particular slice, it's really great to be able to send things in and out of Todoist from different applications. So I've got like this OCR scanner for my phone. I hold it over a receipt. Oh, yeah, like just some document you want to – yeah, totally. It imports all the details and it figures out what I paid at the restaurant. And then I can import that into the expense management system. All of that routes through Todoist.

Todoist is sort of like the central piece that can coordinate all of that. So it's really nice to have something that has an API. I get from a self-hosted Task Warrior server. I've tried and I mean, it was not too bad, but just not that shiny. You know, I could see a place for like a separate task management system that was for like stuff with the family and the dog and, you know, like my medical stuff. Yeah, I've had some look for like shared Trello boards for that sort of thing. Whoa, full Trello.

Full Trello, Wes? That was for some bigger planning. Maniac. Maybe this works better for things where you're just like, yeah, firing off one-off tasks you need to not forget about. See, TechMav knows exactly what's up. TechMav in the chat room says, ScanBot and Todoist. This is what I'm talking about. This is what actually pushed me over is when I started using these two apps together.

Mind-blowing, man. Just mind-blowingly. It makes doing some of those tedious things that I hate to do and forget to do easy because when I have the receipt in my task manager, it's like when that task comes due, the information I need to complete the task is right there

and OCR'd for me, copy, paste, boom, I'm done. It makes it so much quicker. I love it. There's no blockers. You just get your work done. And they give you an iCal feed so you can sync it to a calendar

so you can get all of the tasks overlaid on your calendar. Perfect. I love it. So anyways, that's my pitch for Todoist. I actually signed up for their premium one too because then you get some features that they should just build in. I'm pretty. So anyways, that's my pitch for Todoist. I actually signed up for their premium one too

because then you get like some features that they should just build in. I'm pretty sure you just want an excuse to promote one more Electron app. Well done. Well done, sir. You are talking to the guy who runs Natifier on like all of the web apps he uses all the time now. I'm that guy. I'm like, I'm spinning up spaces on my desktop. I'm spinning up virtual desktops.

Just that way I can put another Natifier window on that virtual desktop. It's getting better. There's so many damn web apps these days, Wes. So many. All right. Well, go get yourself some more Popey and Wimpy and Mark II over at the Ubuntu podcast. Go enjoy. Another fine episode.

Is it season 11, episode 30 that just came out? It is. Yeah. Oh, it's getting close, isn't it? It's getting close. It's almost end of season time. Oh, no. Yeah, three months, and then we'll take a season break.

Oh, okay. Three months. That's not too bad. What are we going to do? Yeah, we finish at Christmas. Oh, okay. I mean, I can prep myself. That's enough time for me to let that soak in.

That's good to know, because now I can prep myself. I don't know what I'm going to do. What do you listen to when you need a break from your relatives around the holidays? We need a season break. You're right.

We do need a season break. I like that. Our season could just be like the Christmas holiday. It could be, yeah. You know, it could be like a short season break.

We don't have to like have this whole curry rule. That could just be something. We could just get curry on the side anyway while we're working. Yeah, man. You're doing it wrong.

You don't have curry. You're just doing it wrong. That's mandatory. So, yeah, okay, all right. That doesn't make sense the more we think about it, actually. If we can have an excuse to mandate getting Curry, why would we not take that? God, what was I thinking? They're brilliant over there.

You know, you just got to watch and pay attention and just pick up a few lessons from time to time. That's true. Go get more of Wes Payne over on the TechSnap show, TechSnap.Systems. He's also at Wes Payne. Why do I always doubt that? It's accurate.

You're spot on. It's Payne with a Y, too. Oh, yeah. Like an action hero, like an 80s movie actor action hero. That's pretty much how my life goes. You're like the podcasting version of Hulk Hogan

without the whole drama thing that happened with Gawker. Let's not bring thatker. Don't, let's not bring that up. Let's not bring that up. All right, I'm at Chris LAS. You can follow me

on the Twitter if you'd like. I don't really care. I don't use it a whole lot, but I do try to respond to tweets. So I try to be a good Twitter citizen that way. I'll also mention the Telegram group.

It's been quite a while since we've plugged that. JupiterBroadcasting.com slash Telegram. Go there and join the conversation and every now and then a random host pops in there and starts chatting with people too.

It's pretty fun. And of course, we encourage you to join this show live. We do it on a Tuesday at 2 p.m. Pacific. You can get that converted to your time zone over at jupyterbroadcasting.com slash calendar. Now links to everything we talked about today over at linuxunplugged.com slash 270, including the stuff for Project Stratus and more information about Eazl and their contributions to open source. Go read up about all of that, as well as the community news and bonus links in there that didn't even make it in the podcast. Thanks so much for tuning in to this week's episode of the Unplugged program.

Hope you can join us live next week. But if you can't, LinuxUnplugged.com slash subscribe for all the ways to get this show every single Tuesday. All right. 270 feels like a nice round one. Nice round one right there. Quality episode. It's really great. Thank you, Andy, for having you on. Great to get those questions answered in there.

Now, we just got to name this thing, and then we're going to clear out for the Ask Noah program, which is coming up on the live stream next. I've got a question for you, Chris. Yes, sir. Where's this review of the Precision Workstation? Oh, I know. I should have said something in the show. Well, you see, we have a particular kind of review scenario we're putting it through, and we're making a special episode that will be airing the week I go to meet BSD. And so it's all part of a bigger episode. So that's

why it got delayed. But I meant to bring something. I meant to bring that up in the show because I'll give you a little preview that isn't directly related to the review. But I decided 1604 is fine. But let's put 1810 on there. So I put 1810 Stockabuntu on there. When I first logged in, I sat there for a minute, just, just taking it in. I'll have to show you what I'm talking about. The, the, the new theme with this screen is it's vibrant.

It, it like, it jumps off the screen. It is, it is beautiful. I don't know what they're doing on this monitor and I don't know what's happening in that theme, but I mean, macOS and Windows don't have anything on this look. It was so gorgeous, and I don't know, maybe I was having a stroke when I saw it, but I couldn't believe how brilliant and shiny it looked. It just looked, it looked the best I've ever seen Linux look on a 4K screen, and it's just this edge-to-edge infinity glass, you know, on this 4K, super nice monitor. Man, I'll show you before you leave the studio. I'm excited to see it. Yeah, it does. It does look gorgeous. I've been running it in a full screen VM on my

XPS 15 with a UHD panel. And yes, it's striking. Yeah. Yeah, this laptop's been fun. So we had to come up with a particular challenge for it just to kind of put it into a workload that maybe somebody buying a machine like this would want. So it kind of delayed it because then we wanted to wrap it in with a wider subject. So it'll come out, but it'll probably be not next week but the week after

because I'm going to meet BSD, which is happening the weekend before that Unplugged, and then I'll be traveling home during Unplugged. But we'll have a show if all goes as planned. Now that I've said that on air, it's better happening. We have promised people. I do have good news. We have bad news and good news.

I didn't quite make it into the show. The bad news is that a Chrome developer recently confirmed that they have no plans to enable hardware acceleration on Linux. Even when you go turn the flag on in the config settings, it actually isn't hardware enabled. It lies to you. The flag is a lie. It's a false flag, as Joey put it on OMG Ubuntu,

which did make me chuckle. And in his OMG Ubuntu article, which I think we'll have in the show notes, maybe make sure we have it in the show notes for me, would you mind? I will look right now. There are builds that do actually have hardware acceleration. So when I saw that news, I thought to myself, well, crud.

We're never going to get that new game streaming tech that we just talked about last week on the show. Linux users are out. Not more than eight hours ago, I got confirmation from a listener who is successfully playing full screen, 60 frames a second, 1080p Assassin's Creed on his Fedora workstation using Intel graphics as part of the beta to the project stream. Wow. So it is doable. He's got a more modern Intel chipset, but it is doable. He says it's taking about 20 megabits a second of his network. So you do have to have a pretty solid connection.

They mean it. But it does look like this project stream from Google is playable on Linux. It might just all be happening on the CPU. It might not be happening on the GPU. He's just got a beefing machine. But it's happening.