LINUX Unplugged - 283: The Premiere Shell

Episode Date: January 9, 2019

Joe joins Wes to discuss the state of Adobe's Creative Cloud on Linux and why the Fish shell might be your favorite new tool. Plus community news, a reality check on Linux gaming, and some shiny new hardware. Special Guests: Jason Evangelho and Peter Ammon.

So, Wes, have you ever owned a Windows phone? Oh, I was hoping you weren't going to ask me that. I'm ashamed to say yes. Okay, I mean, it was a significant other at the time, but there was one in my household, and I definitely played with it a little. But you don't have access to it anymore then? No, although I am curious.

There was some news out this week. Microsoft, they've just kind of abandoned that whole platform, right? But they were producing new phones. They were still getting sold. So there's still relatively recent Windows phones out there. If you had one of the Lumia-branded smartphones, which I did, there was already a jailbreak. So you could actually do something useful with this unsupported hardware

that you've probably either subsidized or paid directly for. But now, thanks to actually a secure boot flaw, which I think is hilarious considering how much work Microsoft put into making that standard. Thanks to that, now on any Qualcomm based, which is basically every single one, any Qualcomm based Windows phone, there's a jailbreak. Yeah, I'm sure that devs will be rushing to port ROMs to it. Yeah, probably not. But if you have something like an HP Elite X3 or HTC One M8, I did see that one out.

That was kind of popular. Maybe more so on, you know, a better operating system, but you get what you pay for. This is Linux Unplugged, episode 283 for January 8th, 2019. Hello, and welcome to your weekly Linux talk show that's put Chris out to pasture this week. My name is Wes.

My name is Joe. Thanks for joining us, Joe. Especially since I know you have a new toy you would probably rather be playing with. Yep. But you're here. You're helping out. Chris is traveling. He's already told us all about his adventures. But meanwhile, well, there's a whole bunch of community news to talk about. Some big changes

planned or not planned for the upcoming Fedora releases. And there's a shiny new kernel coming soon with a brand new major version number. Then, those old GNU projects, they've got releases too. It's just that time of the year. And one of them, well, it's that Bash shell, which we all know and love.

Speaking of shells, we had an opportunity to sit down and talk with one of the developers of the Fish Shelf, one of my personal favorite shells. We've also got a discussion coming up about with the success of Proton, are there still problems lurking at the heart of gaming on Linux? And contributing even more, Joe had a chance to sit down and get some information about

just what's going on with Adobe. Is there any chance of Creative Cloud coming to Linux? But before we get into any of that, well, we're going to need a little help from our Mumble Room. Time-appropriate greetings, Mumble Room. Hello there. Hello. Oh, aren't you guys the best?

All right, let's just get started. We talked about this back in Linux Unplugged 277, and there was some chatter going on about Fedora 31, and was it going to be delayed by a half year? Were they going to break the six-month release cycle? They've been discussing this since November. Turns out, no, they're not. Yeah, this was a bit of a surprise, wasn't it?

Because Matthew Miller came on and talked about all the good reasons they were doing this with all the retooling and the fact that they really needed time to just wait for the release and concentrate on that retooling. But it seems that maybe they're just going to have to get on with it and double up their work. Yeah, as Paul Frield explained, after talking with the council, as well as some of the folks who are depending on the cadence, like especially they highlight IoT here, it was clear we need to look at this option.

But for now, we're not going to have any cadence changes. There's no changes for cadence. There's no objective requirements from them. Unless more specific reasons are given, we're not adopting this at this time. Which, okay. You know, it did seem like there was a little discussion about, clearly there's a lot of underpinnings in sort of just how Fedora gets made, right? The build processes, the CICD infrastructure, all the stuff behind the scenes.

I know there were particularly concerns about unblocking more community management, you know, tying it less to maybe some specific resources that just Fedora team members have. That's all important work. I guess there just wasn't enough of an argument about really breaking from the release cycle, because even though it's not always perfect, it is something a lot of people depend on. Well, this does go to show that Fedora does everything in the open, including making major decisions like this. It was proposed. There was a lot of support for it. We thought it was going to happen. But ultimately, there was discussion and the consensus

was against it. I suppose this kind of thing would probably happen in quite a lot of projects, certainly in proprietary software, and we'd just never hear about it. Yeah, that's a great point, right? If it was in a private company and we didn't hear about the first part, well, we would have never heard about it. It would have just been released and no news story. It also goes to show that Fedora, well, they don't stand still. And that's a good thing, right? There's a lot of great innovations coming from Fedora. This next story is maybe an area where they're playing catch up. So we've already talked about all the data collection and Linux for a long time has kind of been an environment where, you know, you had a little bit of privacy

on your desktop. There was a big hullabaloo about Ubuntu adding more data collection to the installer. And now Fedora developers are looking at implementing a per-system UUID identifier leveraged exclusively by DNF, their package manager, in order to more accurately count their user base. So what do you think about this? Do you think that the controversy is justified or do you think that there's nothing to say here?

Move along. My mature opinion after thinking about it for a while is going to have to be the second one. controversy is justified or do you think that there's nothing to say here move along my mature opinion after thinking about it for a while is gonna have to be the second one but i think it's a question worth asking like i don't think it's nonsense to bring up the topic i think it the way that they've implemented it seems like it's gonna be okay though they're not they're not reusing these uuids it's not being their intention is not to try to identify you as a user. Mostly, they've seen a lot of problems of trying to count stuff

like, you know, if you install six-piece copies of Fedora behind a NAT at your house, well, that counts as one, even if it was your roommate installing that. They've also added

a couple controls around, is this just like a short-lived environment, something like a container versus an actual, you know, maybe a server install or a desktop workstation. That all seems like data that would help them a lot. And personally, I'm okay giving

up. Yeah, I don't have any problem with this. And I think they kind of deserve the metrics, don't they? If you're working hard at making an operating system, you kind of want to know how many people are using it. It's like counting the download numbers for a podcast. We know roughly how many people download the shows that we do but we have no idea about any of the people unless they write in specifically to talk to us and i do trust fedora to do this properly they will do it all in the open and they'll make sure that this data isn't misused but then they can actually use it to do good things like concentrate on the areas where the users are

and potentially even concentrate on areas where they want to get new users. Yeah, right. That is sort of the flip side, is that you have good metrics about what people are using or what user bases are growing that can help the limited resources that are available for development.

Yeah, and it's a shame to see so much controversy about it. People will just get up in arms about anything, won't they? Yes, they will. They also get rather excited about just about anything. And that's probably going to be true about the upcoming Linux release, Linux 5.0. Yeah, this was going to be 4.21. I thought it was actually going to be the 4.20, but it seems that they wanted to 4.20 represent. So yeah, Linus has decided that 4.21 will actually now be 5.0. And there's quite a lot that's actually gone into this

or that's going to be going into it. Quite a lot of graphics stuff specifically. Oh yeah, there is a lot of graphics changes. I also noticed that the Raspberry Pi touch display is finally getting mainlined. Hey, that's nice to see. Yeah, that's long overdue, but yeah, good to see now. The other one that stood out to me

is this new large version of the Terminus console font that you can optionally compile in. That's going to be useful if you've got one of those fancy high DPI displays and you'd actually like to be able to use the regular terminal before you get to your graphical environment. I wouldn't know about those high DPI displays, Wes. No, I wouldn't, but that's one of those reasons, right?

It's not quite working well enough, and since we are something of a terminal-forward operating system, it's nice to see this somewhat first class. That's true, but I think there are other high DPI concerns before that. Oh, that's true, but this one's a low hanging fruit. The other good part that probably needed to get some work seen in general is work to remove some of the performance hits from Spectre and meltdown bug mitigation. In particular, some of the network performance problems that came from Google's ret-pulling fix. That's just going to be good. I mean, as long as security isn't compromised,

we'd like it to be both secure and fast. Have we had a kernel release since the beginning of last year that hasn't had some sort of mitigations in it? If so, it was pretty few and far between. And I imagine 2019 is going to just be more of the same. Yeah, but it's very arbitrary, isn't it? As Linus says, it's really just

he ran out of fingers and toes to count the number on and that's why he iterated it up to five. But there is something psychological about that, isn't there?

When the new distros come out with a 5.0 or anything with 4.0 or whatever, it's going to seem old and crusty. Yes, it will.

And especially, I mean, the kernel has just had such an interesting history. There was the two days which lasted just forever, and now we're burning through version numbers. I think for us people familiar with it, it doesn't really matter. Once you realize it doesn't really matter, you just pay attention to the release notes and what's going on with the kernel, you'll be fine. Outsiders, though, it might be a little bit confusing, especially, you know, things keep changing.

We have the Chrome-style version numbers, we have more semantic versioning style, and then other projects are just all over the place. You know which project I love for version numbers? System D. It's just so straightforward. It's just a whole number and it just keeps going up and up and up. No points, releases, no confusion. It's just whatever the number is, add one to it. Yeah, that's a system even I can understand. Yeah, that's the one good thing about system D. Ooh, Joe. Ouch. Okay, well, speaking of confusion, we've got a little update from the corrections department today.

Yeah. So on the last episode, which I was nothing to do with whatsoever. Didn't touch it one bit. No, no. You talked about this blog post on the Clear Linux blog that was written by Ike and Aoki and talking about how Clear Linux was going to go and concentrate on the desktop and have all this remote desktop stuff. And you guys got pretty excited about that, specifically because Chris had been tipping Clear Linux to be the next big thing. Well, it turns out that blog post was, it was a couple of days old, was it?

Maybe a few hundred days old? Yeah, I mean, published April 22nd, 2016. Ah, right. So, yeah, almost three years ago. Yeah, after that, in 2017, they switched to GN So, yeah, almost three years ago. Yeah. After that in 2017, they switched to GNOME, obviously, as we know from using it and reviewing it. Yeah. So that's not very good of us, really.

That was kind of all of our faults. Chris said that we can throw him under the bus because he is ultimately responsible. But really, I'm going to blame Marius Kwarbeck because someone sent it to him. He sent it to me. I posted it in our Slack. You guys read it and then talked about it. And along that whole chain, nobody stopped to look at the date. So it's quite embarrassing, really.

It very much is. And something we try very hard not to have happen. And generally, we have a chain of trust, right? And we only refer links that you've at least somewhat read or vetted or thought this might be interested. And this time, no one did it. It had been discussed that there was excitement around it, and that was too much for us. So we're sorry about that. I also, I just appreciate the corrections on it, though. We want to be accurate and we want to make mistakes. Well, we've got to own it. I will also say a lot of the stuff that it talked about in the article, XFCE things, that's maybe the silver lining here is XFCE still does make a great remote server desktop. Yeah, it does. I think from now on, you're going to have to write some sort of

script to check all these links, go through and grab the date or something. You know, that was kind of my instinct as well. Just like a link verify or something, a procedure to go through and make sure we're always up to date. Yeah. Well, you've probably noticed, dear audience, that Joe stepped in today and we're very grateful. That's because Chris, well, he's on the road. And once again, you can follow right along. If you go to our show notes, linuxunplugged.com slash 283, you'll find a link to the rover tracker. That lets you be as creepy as you like. Surprise Chris while he's out shopping or just trying to live his life.

No, no. But it will help you find out where he is. Follow along. He's tweeting updates, real cute updates with some Levi pictures. So go check those out too. Yeah, he's in California at the moment. And I'm very jealous because the weather's probably a lot better than it is here. You know, I'm sure it is. I'm trying to get back at him in my own small way, though,

for abandoning us in the rain and the drizzle and the cold. And that's because, as you know, Chris is something of a nano fan, and I just can't abide that. I mean, we're all welcome to our editors, of course. That's just an intrinsic free choice that we all have inside of us. But when you have workstations that don't have Vim on them, it kind of upsets me. So while he's been away, I've been making sure every darn computer in this place has Vim installed. So I was thinking about editors, right? And I just, I'm faster at it. How do you get to the bottom of a file in Nano? Am I just like paging down like an animal?

I just don't get it. You know, I've got a great idea, Wes. While you're installing it, alias Nano to Vim. Oh, that's evil. I'm not sure he would forgive me on that one. Thankfully, he can figure it out

and fix it himself. Yeah, true. Although, how would he edit the config file? I think at least he's got a GUI on these machines, so he could probably pull up everyone's favorite, Kate.

No, gedit. That's everyone's favorite. Oh, yeah, gedit. Timeless classic. Speaking of timeless classics, another recent release caught my eye, and that's GNU Ed 1.15. As you probably already know, or probably don't, Ed is a line-oriented text editor. It's used to create, display, modify, and otherwise manipulate text files if you're a sadist. Now, of course, no one really uses Ed anymore, unless you're in a very particular, maybe an embedded

environment, but I'm sure there are thousands of shell scripts out there that still make use of it in some way and have been ported from a legacy Unix environment and probably still work. Well, should I be embarrassed that I'd never heard of this until you put it in the show doc? You know, honestly, I don't think you should. It just won't come up. It's an interesting piece of Unix history and trivia, still getting updates to this day, but you would just never have any occasion to use it, you know? Especially these days with rich text editors, you probably do a bunch of composition on a website, so Ed is about the complete opposite. It's not the only GNU release, though, is it?

No, no, it's not, Joe. This one I think you probably have heard of. It's the Bash shell. The born-again shell, a complete implementation of the POSIX shell spec. 5.0, they've got a bunch of new, interesting, supported variables. A lot of people had actually been using bash compiled before the official release because some distributions wanted these features, which that's understandable.

But now they're official, they're out, you can go get it. They've got some nice new substitution variables. The random variable works pretty nicely these days. Every time you call it, you get a new random number. Hey, that's nice. And I think it's just nice to see that such a fundamental tool that makes up, you know, even in the days of containers running on virtual infrastructure as functions that live for like 10 seconds, there's still a whole bunch of Bash in like half of those. Yeah, but common ways, who is still using Bash at this point? I think a lot of people, you know, as a, let's say, pro-zoomer on the desktop,

shell switching, it can be something that's pretty fun and useful. But when you're, you know, in the admin server side, there's a lot of pressure against that sort of thing, mostly because, well, Bash is everywhere. It's already installed. I think that's less of a problem now that, you know, Linux is kind of one. There's a homogeneity about the marketplace that you might not have had in the days of Solaris or HPUX or any of the other bigger Unix distributions. But if you're just a user, maybe you're just a developer,

you don't have root on a bunch of these boxes, Bash is going to be the shell you have, so you probably have to know it. Yeah, I mean, I was only joking, really, the thing is, it's ubiquitous, isn't it? It's even on Macs and on the Windows subsystem for Linux, it's just everywhere bash. Yeah. And you know, there are some more minimal shells, things like things like dash, of course, classic sh that you could use. But I think we're stuck with bash for the time being, but you don't have to be, you know, on your own. And I don't frequently change shells on servers unless it's a server I either use all the time or use very infrequently.

But for home machines, you know, if I'm doing some processing, if I'm really getting some work done and I need some specific Bashisms or I just know how to do that really well in Bash, well, you know, it's one command away. You just type whatever your shell, you just type bash and hit enter and you've got a bash shell. That's why on a lot of my machines, well, I'm running the fish shell. And as rightly pointed out by many people, fish, it's not, it's not POSIX compliant. It does things a little bit differently. And for the old guard, you might not be comfortable with it because it'll break a lot of assumptions. But I think for someone new, someone you're trying to get on Linux, maybe they're interested in learning a little bit about the command line,

maybe they're interested in development or learning administration or anything like that, they just want to play with their computer, but they haven't had, you know, 10 years of using a command line before. Phish really is friendly. It's called the friendly interactive shell and I think that it really does live up to that. You've got stuff like easy, you know, easy history searching automatically. It's also got like smart directory memory, so it knows if you run a command frequently inside a certain path,

well, that command will just surface up automatically. I'm curious, Joe, do you use Phish at all? I have tried it before, and I did find it very useful, but it's not something that really stuck with me. I think that it's just a case of Bash is just there. It's always there no matter what machine you're on and that extra little step to get Phish installed is just a bit of a blocker. I just forget to do it and then I just haven't got into the rhythm of using it.

But you've kind of pushed over that hump, haven't you? And it's kind of your daily driver shell. Once you get used to it, it's not on all of my machines, as I said, but for just like simple common command line tasks, it's really hard to beat because it is friendly. It'll remember commands. So if you don't, if you haven't taken the time

to make a whole bunch of complicated aliases on that machine, it's got your back on those. And well, it just had a release. This was just at the end of December, right before the new year. Phish 3.0.0 was released. Now, there's some backwards incompatibility, so you're going to want to go read the release notes if you're a big Phish shell user. That said, they obviously have their eye on stability, so those are just kind of only the things they had to change. There's also some good news.

A lot of people, as someone in the IRC just pointed out, previously Phish wouldn't let you use double ampersand for like an and command as you would traditionally do with the bash shell. They wanted you to use the literal word and. They've added that syntax. So there's actually a lot of new changes that are really friendly for people coming from POSIX. So I think now is probably a better time than ever to talk about Fish Shell and to give it a try. Because of that, we were lucky enough to be joined by Peter, one of the developers of the Fish Shell.

He came on and chatted with Chris and I before Chris left. Let's hear a little bit more from that. Anybody that's listened to Linux Unplugged for a long time knows that Wes and I are both daily drivers of the Phish shell. We love Phish. And we wanted to bring on one of the core members of the project to just talk about a little bit after their new release. Peter, welcome to Linux Unplugged. Thanks, Chris. It's great to be here. Oh, no, no. Thank you for making it, especially sort of last minute when this just crossed our mind after

the new release. But I thought, Peter, instead of me saying it in any kind of ham-fisted way that I might explain it, from your own words, how do you describe Phish shell? What is the Phish shell? Sure. So Phish is a command line shell like Bash. If you are a Mac user and you launch the Terminal app or iTerm, or you're a Linux user and you launch Xterm or Kiti or whatever, you have a program which draws the window frame, and it has some text in it, renders the text, and that's called the Terminal Emulator. And the shell is the program running inside the Terminal Emulator

that will receive the text that you input, and then based on that decide what commands to run. So a lot of users are initially confused about the distinction between the terminal emulator and the shell. And Phish occupies the same space as Bash, as ZShell, as some other ones like TCSH. I like that explanation. That's probably much better than the one I would have given. Nope, dude. So I guess I have to ask you the biggest question on my mind, and that is, how do you get people to use Phish Shell? Because everybody's just got Bash by default now. Bash is sort of the default, and default is sort of supreme. And Phish Shell is an alternative to that.

And in most cases, or maybe in a lot of cases at least, people might not even be aware that they can switch a shell. So how does Fisch Shell fit in with that? How does that work? I guess the better way to put that is, how do you feel about the fact that Bash may always remain dominant and supreme, and some people may not even know they can switch? Does that affect the way you work they can switch. Does that affect the way you work on the project? Does that affect you in any way? That seems like a major elephant in the room with a project like this. Yeah, that's a great question. Driving adoption is certainly a challenge for our project. For the reason you said, Bash is very entrenched. And shells are

like text editors. They're very personal. When you develop a lot of techniques and muscle memory to use these tools day to day, that makes it a challenge to switch because when you try a new project, it feels unfamiliar, right? The things you're used to don't work. Not even muscle memory applies or less of it applies and there's surprises. And so one of the ways that we try to address that in the fish shell is we try to identify the features

that users miss from Bash the most and then we add them in. And we're conservative about that. We do that slowly but that's one of the ways that we can help make users feel more comfortable initially. It also helps that Bash is missing out on a lot of user-friendly features. Features that are part of the Phish shell.

We say that Phish shell is a command line shell for the 90s. That illustrates that there's a lot of ways command line shells are stuck in the past, but we think that Bash is a command line shell from the 80s. It doesn't have features like syntax highlighting. It doesn't have features like autocompletion and so forth. So one of the ways that users learn about the phish shell is they just see it on other people's screens and other users' screens. They're like, wow, that looks really cool. How can I get that?

And it spreads by word of mouth. Yeah, that has definitely been my experience. Not only how I found it, but how others have found it after they saw it on my screen. And you, you know, you described it as it's a personal thing. I use the word when I was chatting with Wes before we sat down. It's also like maybe the most intimate software I use on my computer because I'm not just interacting with it and clicking it, I'm communicating with it. Yeah, it sees everything that you run and it enables you to use your machine.

And there's not much software these days that I have like a very strong emotional opinion about where you could use language like love as for how strongly I love Fischl. Like I really feel strongly about it. I don't feel that way about a lot of software anymore, but Fischel is still one of them. And so there's something special about building a shell.

So the project's been around now, what, since 2005? The project is around since 2005. That was the first release. I was not one of the original authors of the Fischel. That would be Axel. I won't attempt his last name, but you can put it in show notes. And Axel started the project in 2005 as the first release. And I think development started to slow down around 2010. That's when I picked it up.

And I was a user of the fish shell. And I was very happy with it, but there were some bugs I found or things that I thought could work a little differently. And I just decided to scratch my own itch and set out to fix them. And I figured it would take me a week. And then, you know, a year later, I was pretty embedded in it at that point. So that's how I got my start, just scratching my own itch. Well, isn't that interesting? How many stories have we heard like that,

where somebody starts scratching their own itch? You guys just recently had a big release too. Was it version 3? Is that the new version that just came out? 3.0 or is it 0. You guys just recently had a big release, too. Was it version 3? Is that the new version that just came out? 3.0 or is it 0.3? It is 3.0, so this is our third major release. The first major release where I was on the project was 2.0 and 3.0 signals that we've made

some fairly radical changes in this release, and that's why we think it deserves a new major version number. Can we do a little bit more about those? One thing I've always liked about Phish is, you know, I have it on some systems, not all, unfortunately. But whenever I use it, I don't,

I'm not concerned usually that things are going to break. It's mostly just work. So are there some things people really need to look out for when they upgrade here? There's a few things that are different, but we have a, you know, we're certainly, we don't want anyone to have an experience where they install it and their software breaks. So a lot of the changes are backwards compatible.

So, for example, we support now in 3.0 the double ampersand for AND or double pipe for OR, a very long requested feature. But the old syntax, the explicit AND and OR commands are still supported, so none of that breaks. Very good. Yeah. But the old syntax, the explicit and and or commands are still supported. So none of that breaks. Very good. Yeah. We also have a way of staging breaking changes called feature flags, which allow users to opt into it. So these are changes that we plan to make in the future when the current syntax is deprecated. So this includes, for example, question mark as a glob. This often breaks users who paste in the URL,

and this will be interpreted as a wildcard. So with Phish 3.0, we plan to eventually remove the question mark wildcard. But in 3.0, you can opt into that removal by just setting a feature flag in one of your variables. So there's a lot of forward-looking features in the 3.0 release like that. Is the feature flag new to 3.0, or has that been around for a while? The whole mechanism of staging these deprecations and breaking changes is part of 3.0. Wow, interesting.

It signals a bit of a philosophical shift where we're trying to be a little less dogmatic and a little more practical and move a little faster, right? And it's very rare to have any program, but especially a shell, remove features. Normally, they just create features over time. So we want to have the capability to have the right feature set, not just the sum of all past feature sets. Right. But how you actually pull that off is a little new in the world of shells and getting it right is important. That's right.

It is a little risky so we'll see how it works out. Yeah. Oh, I'm excited to hear that. Kind of talking, let's keep talking about the practical stuff for a moment.

One of the things I love about Fish Shell, to be honest with you, is I can load it everywhere. Fedora, Arch, Ubuntu and even macOS and probably the subsystem for Linux

if I was on a Windows box. And I really appreciate that fact. Just, it's one of the, Ubuntu, and even Mac OS, and probably the subsystem for Linux if I was on a Windows box. And I really appreciate that fact. Just, it's one of the, I mean, obviously Bash is everywhere, but I don't want to use Bash, right? I want to use Phish. So I'm curious, kind of like roughly, is Linux your largest install base? Do you have, are there a bunch of Windows users?

Like what's the user base look like for Phish? Yeah, honestly, I don't know that. I mean, Linux and Mac are both huge, the largest users, of course, but we don't track how many users there are from any particular platform. There's no telemetry in the Phish, so we don't know.

But we do have a lot of actually surprising use cases that we discover. The team that is on Haiku, which is the BOS recreation, they have a lot of Fischl fans there, and they have Fischl running on Haiku. That's awesome. It makes me really happy. WSL, a Windows subsystem for Linux.

There are also a number of users there. I actually feel like WSL is a big opportunity for the fish shell because users who have WSL are less likely to have a long history with Bash. So they're coming at it with fresh eyes. I run it on my Raspberry Pi. I run it on my Raspberry Pi. I run it on AWS. We even have a user running Phish on the TH2, which is a 33 petaflop supercomputer in Guangzhou, China.

Right on. I bet Phish runs just great on there. I bet it does. I bet it does. Phish has always been one of those projects where you install it, and the defaults have just felt right. I've been able to install Phish and then run it for six months before I even changed a single thing.

I really love it. And I think it's one of the projects that you can install on a new Linux user's computer and make the command line a little less intimidating. And that has massive potential. I agree completely. a little less intimidating. And that has massive potential. I agree completely. One of the design goals of Phish is that it be very user-friendly. And the way we design features is we start by thinking about what do we want the user to see, to

encounter, and then we work backwards to the implementation of the feature from there. So for example, setting your prompt can be done either via the command line or via a graphical UI. There's a command fish config, which spins up a little web server and opens a web page that connects to that, all local. There's nothing on the network here. It allows you to set your colors and your prompt and so forth just with a graphical UI. So that's much more convenient for new users compared to,

for example, setting the PS1 variable for controlling your prompt like you would do in Bash. Which is how I had been doing it. You know what's funny is I've been using Phish for years and years and years and it was only until last week that I ran Phish config for the first time and did it through the webpage. I've always just done it the old way, the old school way

because I'm used to the way Bash does it. That's how you assumed you would do it. But that web page was great because there was pre-suggested themes and looks and styles for my command line that I love, I love. And it's an example of making, again, the command line accessible to people

that are maybe transitioning from another platform where you don't use the command line as much. And I think one of the things that makes Linux more usable as a platform is if we just acknowledge sometimes it you don't use the command line as much. And I think one of the things that makes Linux more usable as a platform is if we just acknowledge sometimes it's okay to use the command line. And people don't like that, but we can make the command line a better experience. Yeah, right, if we just embrace that and say,

we have a great command line. Well, Peter, what else should people know about? Is there anything you want them to check out for the project or things you guys are working on or any destinations you want to send them? Well, the project is hosted on GitHub. We have five contributors on the project other than me

who are very regular and very passionate. I'd like to give them some credit here because many of them are even more active than I am. In particular, I want to give credit to David Adam and Fabian who are both very long-time contributors.

And we have Aaron Guise, who's done a ton of great work across all parts of the shell. And we have Saeed Shouar, who was with me from the beginning. And Mahmoud is our newest contributor, who's very active in porting fish and getting it running well on WSL. So I'm really grateful for all their contributions. It's actually pretty easy to get started contributing to the phish shell. A lot of contributors just, they have a command they want to run

and tab completions aren't working for it properly. So they can just add a tab completion and that's a very easy way to get started or improving the documentation. So if you'd like to contribute, if you're a phish shell user, it's very easy to get started contributing to the shell. Yeah, I think especially in this case, that seems that's more so than what is even, I think, common with an open source project like this. There's some stuff in

there that is really user facing like documentation that anybody can participate in. Well, Peter, thanks so much for coming on the show. Thanks to you and the entire team for working on one of my absolute favorite pieces of open source software. All right. Thanks, Chris. Thanks, Wes. And thank you to Peter for coming and chatting with us about that. Peter was great and really friendly. And it's an interesting tale because as we talked about in that interview, at around 2012, phish development had stagnated.

It's very obvious if you go look at the commit analysis over on GitHub. Peter stepped in and reinvigorated the project and I've just had a recent release. He wasn't the one who released it. There's enough other people in the project that there are several main contributing developers.

It seems healthy. It's probably not for some expert users. It might not even be a forever shell if you're going to go off and become like a terminal expert. But I think it's a shell that thinks about what the interaction between a user and a computer should be in a terminal environment. They're exploring interesting grounds. He almost made me switch to it. He almost convinced me, but almost. Well, you do need to be the curmudgeon, Joe. What would we do without you? Okay, so let's move on to another area I've been excited about recently.

That's gaming on Linux. And it's just some of my friends who are, you know, not Linux desktop enthusiasts. They might not even listen to the show, but they make a lot of comments about how great Proton is, how they can use their Linux desktops to game. And I've certainly experienced that. I've been advocating for it more. But while we were getting excited about that, we talked a lot about it here on Jupiter Broadcasting over our various holiday and year-end review episodes.

There was a discussion right around the new year over on Twitter that I missed until now. It started with a tweet from Ben Golis, who works over at Uber Entertainment, actually local here over in Kirkland, Washington. They make Planetary Annihilation. And, well, the discussion is basically they shipped Planetary Annihilation on Windows, Mac, and Linux. Linux users were actually a big

vocal part of the Kickstarter for the project. But in the end, they accounted for less than 0.1% of the sales and greater than 20% of auto-reported crashes and support tickets. Most of that graphics driver related, I guess. His opinion would totally skip Linux. So does this mean, is this a fundamental problem? Is there something we're missing here? Or does it just mean that much like Electron apps, we're going to get games, we just won't get them natively? I think that that is the reality of the situation

because if you look at this Twitter thread, it's easy to kind of come down on the side of, well, it's just not worth supporting Linux. But then people have chipped in on the Hacker News thread about this that surfaced this week saying that, well, they did it wrong. They didn't go into it with their eyes fully open. They kind of thought it would be easy to port it to Linux. And the reality dawned on them that it was actually quite hard. easy to port it to Linux. And the reality dawned on them that it was actually quite hard. And I think in that Twitter thread, there's a very, very depressing and telling phrase that he says, adding Linux support ended up likely costing Uber hundreds of thousands of dollars for a few

hundred dollars in sales. And that just doesn't look good for anyone trying to port any software to Linux, never mind just games. Yeah, that is true. And that's part of the problem is even if we have a more successful game development story in 2019, you might not know about it. You might not know the right choices. You might not. That's hard to find.

There is no universal development kit to target. You can be an expert at game development and still not be able to develop for our platform and have it be a success, and that's concerning. When the whole Proton thing broke as a story, there was kind of two ways to look at it. One is that, okay, well, this means that no one will bother developing natively for Linux anymore. But the other way to look at it was,

well, it will enable a lot of Linux users to use these games, play these games that they otherwise couldn't before. And it will potentially make people who are on the fence about switching to Linux make that switch. And then where the users are, the development goes.

I think that realistically, it's probably the former rather than the latter though, isn't it? That you're much more likely to get just Windows games working and tested with Proton and that's good enough. Why bother doing a whole native port

when you can just use this off-the-shelf solution from Valve? Yep, that is true. And then even more so, we're still, even with it working really well just out of Valve, you're still going to buy a computer

that probably has Windows on it, and if you're a gamer but not a Linux fan, there's not going to be a big reason that you should ever switch to Linux. Now, you might have a better switching story if you eventually happen to listen to this program, say, but we're just still same old, same old, I i guess i don't know though

where's my evo 970 didn't come with windows on it it came completely blank and i had to install whatever i wanted onto it and that's the reality for most gamers who build their pc i mean most enthusiasts who want to go full rgb and all of that they build their machines so this argument of that they come with Windows, I don't think is necessarily true. Obviously, laptops come with Windows, but if you're doing a self-build, and no self-respecting gamer would use a pre-built desktop, surely.

Yeah, sure, sure. Yeah, of course. But they just consider Windows to be part of that build. No, that's a very interesting point. Like, there's this cultural element. So if we can maybe change part of that build. No, that's a very interesting point. There's this cultural element. So if we can maybe change some of that, then if you're already an enthusiast, you're willing to do research, you're willing to maybe tweak or tune

Linux, Linux does have a better story there now. Yeah, but you only have to look at big gaming and PC YouTube channels like Linus Tech Tips. He had this video, Linux gaming doesn't suck because the assumption was that it would just suck. And then he actually tried some Linux games and said, actually, they're all right. And that was a surprise. And that's that culture, isn't it?

That gaming really is all about Windows. That's just been the assumption for 20 plus years that if you want to play games on a PC, you're using Windows to do that. Yeah, I suppose you're right. And unfortunately, there are still many workflows that you might need a Windows or another proprietary operating system to get your work done. And one of those for a long time and still remains one is the Adobe platform, their whole creative suite of applications. I know, I mean, I've used them over the years. I don't have them now, but they're obviously the standard and Linux users for a long time have been wondering, do we have even a shot of ever getting these applications? You've done some hard work and found out a little bit more information

though, Joe. Yeah, I've been chatting with Jason Evangelo, who has become well known in the last kind of few months, I suppose the back half of last year for writing for Forbes. He kind of switched from writing about gaming and hardware to writing about Linux as he got into Linux. And it's just become really, really popular. Everyone seems to be following on twitter and reading his articles and more importantly it seems that he has driven a lot of people into our community people who are into gaming or kind of on the fence or whatever he is really doing good stuff for the whole linux community as far as i can see and he's very pragmatic i know okay some people are going to say oh we don't want that proprietary software.

Why don't you just use Kdenlive or whatever, or GIMP or Krita? But the reality is that if we are going to grow the Linux user base, we're going to need some of these huge proprietary commercial bits of software like Creative Cloud. And he is very much of that belief. He's very much a pragmatist.

We need both, right? We should have good open-source technology, but we also need to use what the market's using. Well, in an ideal world, we wouldn't need any proprietary software at all, but it's a vicious circle of the proprietary software isn't there so the users don't come to the platform, and then the developers don't develop open source software for them to use so i think it's it's hard for me to say that i want this

proprietary software to come to linux because i don't think i would use very much of it if it did but i think that an awful lot of people would and that would be better overall for linux because this is all additional stuff it's like the the Proton stuff. It's like Steam. It's like the graphics drivers. You don't have to use any of that stuff. You can run Debian or Triscoll or whatever and just run completely free software. That's a good point.

It's all additions to Linux and free software. So I see it as a good thing. But yeah, he had a chance to talk to some people from Adobe some quite important people actually and so I thought that I would catch up with him and get him on the show and speak to him about how that went and so should we ever listen to that? Let's do it. Thanks a lot for coming on the show Jason. Yeah you bet happy to be here. Yeah we've been wanting to get you on Linux Unplugged for a long time and finally the timings have worked out. Yeah, you know, and before, when you asked me the first time around, I was really shy,

and now I'm not, so. So just a little bit of background. You've been writing for Forbes for quite a while now, and in the relatively recent past, you started writing about Linux, and those articles have really taken off. And you've used your Twitter following as well to kind of help that. And it's just kind of a bit crazy, isn't it? It has. I mean, in a very unexpected way, you know, when you spend like six years at a, you know, on a major platform talking about, you know, AMD and NVIDIA, you know, and Intel and reviewing, you know, big products and things like that,

or just breaking news. And then six years later, you kind of stumble into Linux for the first time in like 13 years and fall in love with it. And at that point, I gradually decided, you know, I'm loving this. I'm loving the discovery and really loving just writing about it and being so engaged on Twitter and Facebook with people who are enjoying the content. It's been a repeatable thing where, you know, I will take, say I'm writing two articles in one day

and I'll say, okay, hey, Introware is launching their Aries all-in-one desktop PC with Ubuntu and Ubuntu Mate, right? And then the same day I'll write something about NVIDIA and the Introware article will get more traffic and it blows me away every single time, but it is becoming a consistent thing. And I think that there's a really, at least to me,

a surprisingly large audience who is hungry for that content in kind of a, I guess, non-technical, like more conversational manner, if that makes sense. Yeah, well, it seems to be working for you. I've seen you post stats about the Linux articles being the most popular of all time, almost. Yeah, I wrote, towards the end of the year, I wrote about Deepin. And it was just kind of an opinion piece, you know, talking about why I think that it's a more beautiful desktop experience than Windows and Mac OS. And within a few weeks, that climbed to the number three most viewed article since 2012 we're talking about like beatings you know stuff like pokemon go when it launched and things like that i mean yeah incredible just incredible i i'm i'm still blown like months later i'm just still blown away

and it's been it's been a blast but just before christmas there was a bit of a blow up on Twitter about Adobe and Premiere, and people requesting that for Linux. And you managed to send so much traffic their way that they shut down the topic. So what was all that about? They did. I think that, you know, I actually heard about it through a couple of my Twitter followers and also Michael at Foronix. I want to make sure people understand he's the one who wrote it first. I just kind of elevated it and gave it a bit more exposure. But I do want to make sure that people realize he kind of broke that news.

But anyway, I wrote an article saying, hey, Adobe wants your feedback on bringing Premiere Pro to Linux. And the tweet blew up, the article blew up. And I think within three or four days, they had somewhere north of 6,000 upvotes for that and a lot of really healthy discussion in there between both the Linux community and a lot of the Adobe staff, you know, like, like I think it was Patrick Palmer, who does, he's their senior product manager for professional video editing for Creative Cloud. And I, at some point, someone alerted me to the fact that it was just gone. They're like, I'm trying to click the link in your article. It's just not there. It's nowhere. It's gone. And Adobe Help actually got in touch with me and they said that it had received so much, like a surge in upvotes that it had triggered their spam filters and they had taken it down. They just had no idea that it would be so popular.

So that got resolved. And yeah, that got resolved soon. And that was good. And then people kept on voting. So after that, sometime after that, Adobe's help account on Twitter actually reached out to me and they were like, hey, Bill Roberts and Patrick Palmer would love to just chat with you about your point of view as someone in the Linux community, looking at the possibility of Adobe products being ported over to Linux. And I said, great, let's do it. And so, yeah, we got on the phone like four or five days later and we had a really

great chat. But before you had this chat with them, they did officially respond and say something along the lines of, you know, we take this interest very seriously, but don't get your hopes up. There's a lot of fragmentation with Linux and it's not just supporting one platform. It means supporting loads of different ones and basically saying, thanks for the interest, but don't hold your breath. Yeah. And I think that that discouraged a lot of people, but I, you know, in a way, Joe, I kind of appreciate them for not getting people's hopes up. Yeah.

You know, although through the course of the chat we had, it became apparent to me that what matters right now is that this is a possibility. And this is what they told and this is what Bill and Patrick said. They're like, this is a possibility now. They weren't entertaining this idea before. Oh wow, so it really has made a difference then. It really has.

And I don't know if it was, let's be honest, it probably wasn't the user voice post and feedback that prompted that, but it may have helped, right? It may have definitely changed or influenced maybe their timeline or brought a few other important members of the team over to the Linux side of the fence. I don't know. But what matters is that they are looking at this as something that's possible now. And when they said that there was this fragmentation problem, everyone replied to

them saying, snaps, flat packs, snaps, flat packs. Is that something that they weren't aware of or what? They're very aware of it. And, you know, I kind of brought that up. You know, I said, hey, you can, you know, you can use stuff like Vulkan. You can focus maybe on just one distro and don't spread yourself too thin. But they're completely aware of all of the technical environments surrounding what it would take to port Premiere Pro over to Linux. But Patrick's problem is that there's a major difference between developing and testing, right? Because sure, they could develop it for Ubuntu, but it will get used on other distributions, right?

And there's all these combinations of distro and CPU and GPU and everything else. And the cost of testing those and scaling that up is what concerns them. But beyond that, they want the product on everything it could possibly be used on, right? And they want to come out of the gate flawlessly. Even though they're very aware, and they told me this, they're very aware that they could haphazardly release something on one distribution and maybe everything doesn't work. And they know the community will make it work. But that's not how they want to do it.

They want to come out of the gate flawlessly. And one other aspect of this is that the conversation isn't just about developing Premiere Pro. And I think that a lot of people in the Linux community think, why not just do that? That can't take a lot of resources. That can't take a lot of resources, it can't take a lot of time. But Adobe's issue is that their users are tapping into their entire ecosystem, right? They have a workflow that normally extends beyond just a single Adobe app.

Well, yeah, people want the whole Creative Cloud, don't they? Exactly. And so that is, bringing the entire creative cloud is a, would seem to me to be a monumental effort. And they just don't want their, their products coming onto Linux handicapped and, and kind of robbing users of, of what they've come to expect from their products. Right. And there, you know, another wrinkle to this is that obviously they want to achieve some, some, you know, another wrinkle to this is that obviously they want to achieve some, you know, levels of parity when we're talking about performance, right? They want to have parity between, you know, Mac OS and Windows and Linux. But that's only half the story. If all

the third party libraries don't support the format, having that great performance is kind of useless. So it really is that entire ecosystem that has to start coming together. Yeah, suddenly this becomes quite a large endeavor, sort of almost too much to take on. Yeah, but yes, I mean, yes. And they did have a lot of interesting things to say about enterprise. I mean, not in a lot of detail, but that's where they would focus first. I will tell you that. That's where Adobe would focus first,

is the enterprise. But here's what's really interesting. The path to having these Adobe products on Linux, where do you think that would happen? Would it be just a straight port, or is there something actually coming that would kind of get their foot in the door? What do you mean by something coming? I mean, I would assume that the path to getting it on Linux would be snaps essentially, or possibly flat pack. Yeah. And I

don't, I don't have an argument for or against that. Um, but I think when we're talking about like the, the architecture and the development and just the knowledge that their engineers and developers need, it's actually Premiere Rush. All right. So Premiere Rush, as probably you guys know, is kind of a mobile light version of Premiere Pro. And it's already on desktop. It's already on iOS. But it was strongly hinted to me that once it comes to android this year that is their kind of first step all right interesting and i wish i could tell you more but that's really all that

i know i can just tell you that the the way that they were explaining it to me and the kind of the you know the the attitude that they were portraying and the enthusiasm don't take this as the gospel truth or anything but it's really i really feel like it's you know, the, the attitude that they were portraying and the enthusiasm, don't take this as the gospel truth or anything, but it's really, I really feel like it's, you know, once this hits Android, it's something that they can legitimately start to explore. Did the market share issue come up at all? Because that must be in their mind. Uh, it, it, it, it really didn't. Uh, and I don't know if, I don't know if that's just because they were, you know, trying to be polite. And I also didn't bring it up. I was really encouraged walking away from that conversation because they don't view bringing their products to Linux, they don't see it as a

technical barrier. They know how to get there. Right? And I think the larger issue for them they know how to get there, right? And I think the larger issue for them is having some reassurance that all of the effort that they invest into it, into not just Premiere Pro, but the entire Creative Cloud suite, into the testing and scaling for every distribution and making sure that the entire ecosystem is working as it should on other major operating systems,

it's encouraging to me that they know all that, right? And that they are in the process of developing Premier Rush for Android, kind of hinting at the fact that it helps them with potential Linux development. I was really happy walking away from that because it wasn't, a lot of the Linux community, at least who's, that has interacted with me about this whole Adobe thing, they feel like Adobe's just giving them lip service, you know? Like, ah, we're going to get your hopes up and we'll do this poll and, you know, we'll

make you think it's possible and then we're just going to slam the door shut in your face. And that may have been true in the past, but I genuinely don't believe that's true now. And I think it's something they're at least open to. And it's just a matter of resources and time and market share, probably. Let's be honest. Yeah. Yeah. Well, it'd be very interesting to see what happens with it, whether or not it is just lip service. But hopefully you'll be clued into what's happening and you can maybe come back and let us know. But it's been great to have you on and very much look forward to speaking to you again sometime. Yeah, absolutely. Thanks, Joe. Yes, thank you, Mr. Joe.

That was a great contribution and a great interview with Jason. You know, he has been he's been doing a lot of great writing these days. And he's definitely a friend of the network. Yeah, could you tell that I'd only just woken up and it was still daylight? You know, I wasn't going to bring that up. And no, actually, I don't think I could tell. So good job. It was a good chat. And yeah, he's a very interesting guy.

And yeah, I definitely recommend checking out his articles on Forbes. He's always got a positive spin on everything. That is actually what I was, I have a question for you about it, but that was going to be my takeaway from that interview is, all right, well, there's nothing concrete there. Nothing's changed really, but it's also not a bad thing, you know? I mean, obviously, Adobe runs a massive cloud platform. They clearly deploy a lot of Linux servers

and have a wide array of technical experts at their employees. So, of course, they know about Linux, but it's nice to hear that there's not this horrible bias or rejection or just developers being tired of hearing this very vocal minority of people clamor for support. So it seems like there's a positive relationship we can actually build on here,

and it might not be anything soon, but there are professionals who use Linux. Yeah, it does feel like it's going to be medium term rather than short term, but they are clued into the fact that Linux users do want to use their software, do want to pay them money. And as Windows and macOS seems to get worse and worse, and Linux seems to get better and better, then you never know, maybe we'll get that critical mass of users and maybe they will pull the trigger on it. Right, and I mean, they already have to adapt to a changing marketplace.

While Windows and Mac might still dominate the creative design markets, mobile is huge, right? It's eating the desktop market share all the time. So they've already worked on some stuff for that platform. So I've talked to a couple people inside Adobe as well, and they're actually, that platform's surprisingly modular. They have a lot of high-level code and infrastructure to sort of orchestrate on top.

So it was interesting to hear Jason talk about that the technicals aren't the problem and that it comes down to something like, you know, not wanting to cripple the product, wanting to have a good reputation. That's a very business sort of concern that I think sometimes we in the open source world forget about. Yeah, it is all about dollars and cents

at the end of the day, isn't it? Right, I mean, I guess it has to be because that's what keeps the whole thing going. And clearly there's a lot of people who use and rely on their products. So I can't blame them. Time will tell. Maybe go try some free and open source software and submit bug reports and see if you can make them a little bit better. Yep.

All right. Do you want to hear about my toy then? Yes, I do. We teased it a little bit at the Yep. All right, do you want to hear about my toy then? Yes, I do. We teased it a little bit at the top of the show. You were very gracious despite being woken up early by this delivery of something you've been expecting

for quite some time, I believe. What did you get? Yeah, I've been wanting this for quite some time and finally sorted it out. This is an Entraware Apollo laptop. Now, I have to say quickly, full disclosure, Entraware do sponsor my other show, my independent show, Late Night Linux, but this will be completely

unbiased. I can't do a full review because I've only had it for a matter of hours and I did go back to sleep briefly today as well. You deserve to, you deserve to. Yeah, I haven't had much time with it. So it's really just first impressions. But I'm pretty pleased with it so far. Turning it on straight away, the screen was just super bright, super sharp. And yeah, really, really impressed. This is their 14-inch Ultrabook that I'm talking about here called Apollo. Right. Okay. So Chris had an Apollo as well, and I've used that somewhat extensively.

But is this right? That was a 13-inch screen size, but the form factor hasn't actually changed? Yeah, it's the same chassis, but smaller bezels. Oh, I like it. Yeah, so they've squeezed that 14-inch screen into it. And that gives you that extra little bit of real estate. It's 1080p. We're not talking about crazy 4K or anything. And on a 14-inch screen, do you really need any more than 1080p?'re not talking about crazy 4k or anything and on a 14 inch screen do you really need any more than 1080p i certainly don't no no i mean i think i think you probably need a bigger

screen if you're gonna really leverage those resolutions but some people like it and you know you've seen even on some of the the low-end mac products some pretty high resolutions. Yeah. Well, there are some bad sides to this. It's got USB-C, but that is not Thunderbolt. So there's no external graphics to be had. But then again, I'm not going to go buy an external enclosure and a graphics card. If I want a serious graphics card, I'll put it in my desktop machine. Thank you very much. This is supposed to be light and portable. Was that a dig at Mr. Martin Wimpress on a day that he's not here? Well, I just don't really understand why you would use that.

If you're tethered to a desk and an external screen anyway, why not just use a proper desktop machine? I don't understand all this Thunderbolt business, but I know some people are into it and storage as well, obviously, is going to be faster over Thunderbolt. But it has got some full-size old-school USB, although they are 3.1, so that's nice.

And it's got DisplayPort and full-size HDMI and an SD card slot and gigabit Ethernet. You don't often get that these days on an Ultrabook. And that's important when you have a lot of WAV files to upload. Yeah, exactly. And it's also got a headphone jack, which you can't really take for granted these days, can you? No, you have to check. It's a little awkward. It's sort of like trying to be like, what's going on here?

You're holding it up to the light. Oh, yeah, there it is. Okay, so there's some questions in the IRC I just have to ask. As someone who has used one but never bought one, is this just a rebadge to Clevo? What's going on with Entrawear? Well, Clevo is the ODM, but there are some less scrupulous vendors, resellers, you might say,

who just buy Clevo stuff and then just sell it. Whereas what Entrawear does is they very carefully select the parts. It's a Clevo chassis, but they make sure that the parts that they're putting inside this thing actually are going to work with Linux properly. And they also make sure that the firmware

and everything's working. And that's why there are some vendors who punt out the new Clevo machines quite quickly without fully testing this firmware stuff whereas enterware sometimes wait a little bit and make sure that everything is going to work and you know that was my first impression of it um with ubuntu okay i don't really like gnome i think it's fair to say but um right yep yep it worked perfectly everything worked as should. It was a very good experience. It was very much a stock Ubuntu installation. You know, it wasn't long before I put Xubuntu on it,

you know, obviously. But from what I saw of the Ubuntu installation, it was very stock and everything just worked perfectly. And with Xubuntu as well, so far, so good. I've not come into any problems at all. See, that's a really good point because, you know, there aren't that many ODMs in the world, and so

really, whatever you buy, you're just hoping that the person who assembled it did a good job of choosing equivalent parts. So if you've got a little bit of a guarantee that oh, well, the people who sold me this are at least aware that Linux is a thing, and in this case actually optimize for it, well, it might not be the fanciest, it might not have the

smallest bezels, but it'll run Ubuntu. Yeah, someone's asking about suspend in the IRC. I've never suspended a laptop. Once you've got an NVMe drive in there, it boots so quickly that there's just no need to suspend it. Oh, really? See, that's funny. I feel like I used to have that.

That was my workflow for a long time, but now I just close my laptop and forget about it. I feel like I used to have that. That was my workflow for a long time, but now I just close my laptop lid and forget about it. I don't know. I just like to have it either on or off. I mean, it's not secure. That's totally reasonable, and it's nice to actually reboot your machine, that your update didn't fork anything, all of those things. I'm just surprised. I'm so used to... I just don't even... It goes probably weeks without rebooting.

Yeah, I don't know. I reboot my machines all the time because now that storage is so fast, this thing, you turn it on and before you've even, you know, taken a sip of your drink, it's booted. So, you know, that's the same with all modern machines with decent storage in them. So for me, suspend's just not really a thing anymore. The nice part too is if you don't reboot all the time, then you forget what you have to do when you reboot, maybe adding SSH keys or entering permissions or re-logging into something.

And of course, the moment that you have rebooted is probably the time that you really didn't want to. Yeah, exactly. One nice thing about this is it's got a 2.5 bay as well. So if you want to put spinning rust in it or a larger SATAsd you can do that i'm not going to do that i don't think i've got a 500 gig um samsung ssd nvme in there which i think will do me for now but it's good to know that i've got that option if i want to yeah right

that's something that you always want to look for is is this serviceable will it have a lifetime can i upgrade it when i want and need to? Now, most users might not, but in our audience, well, that's a great way to get more life out of your machine. Yeah, well, as for upgrading it, it's really, really easy. It's like 10 screws or something. Oh, yeah. And then the bottom just comes off, and it supports up to 32 gigs of RAM, which is really nice.

So you can get all the VMs essentially running on it. It's easy to take that for granted. I mean, I've owned several laptops where you had to take the whole main board out before you could even get to the hard drive. So the fact that it's 10 screws and, you know, Bob's your uncle, great. Yeah, you can upgrade both storage and the RAM.

It's very user serviceable. And you could even swap out the fan and stuff if you wanted to, you know, down the line if anything goes wrong with that. So yeah, pretty impressed with it so far. But yeah, I can't really say that I've lived with it long enough. So maybe I'll have to give you an update at some point. Is this going to become a distro review machine or just a mobile workstation for you? No, very much not. It means that my old, old laptop is going to become the full-time distro review machine. This is very much production.

I've got my main desktop for main production, but this is my take to LinuxFest Northwest machine. And so it'll be probably a Zubuntu, you said? Yeah, yeah, Zubuntu. No deviance there. All right, that makes perfect sense. Thank you for that mini preview of the Entraware. I'm sure we're going to find out a lot more as you actually use this machine. I'll be curious to give it a hands-on look when I get to see you at LinuxFest Northwest. That's a great chance. We should just talk about LinuxFest

Northwest. It's their 20th anniversary this year, and if you've never gone before, it's a great time to go. We've got links to all of that and everything we talked about today. Linuxunplugged.com slash 283. Now, Joe, I'm just ever so grateful that you came and hung out here with us today. Where can the audience go to find more of you? Well, Linux Action News is a good one. Linuxactionnews.com and error.show. If you want to listen to user error,

that's me and Popey and Dan from elementary talking about life, the universe and everything and a bit of Linux chucked in there. So yeah, check those two out. And also latenightlinux.com if you want to hear me effing and blinding, effing and a jeffing. And you almost always do.

Thank you to our wonderful mumble room, as always, and, of course, the IRC and everyone else in the Jupyter Broadcasting community. LinuxUnplugged.com slash contact if you want to send us some of your favorite things, either interesting finds, questions, or just your comments and feedback. We love it all. For all the other Jupyter Broadcasting shows, jupiterbroadcasting.com. Thank you so much for joining us.

We're going to get out of here, but don't worry. There's lots of good stuff to look forward to. I'm always excited about the next user air release. Just let me say, Joe. So good job. Thank you for joining us. And I think we're going to see you next Tuesday. Comment wise you can do better than that next tuesday So, I just got my father an iPad,

mostly so he could read some, like, journalism, you know, various apps from bigger newspapers than his tiny, dying hometown newspaper. And he used, like, a Windows... Oh, I don't know, was it an XP machine? I don't know, whatever his... Before he retired, right? And so he has this very particular interface with computers,

and I've got my mom on Linux now, and part of me wonders, he might just actually take to the command line because it's sort of a conversation and less of an exploration visually. And so, like, but Bash is kind of arcane. And if you know what you're doing with bash or zsh or really any shell you can obviously configure most of the stuff that you find but with fish it just like it really is friendly and that friendly only goes so far and once you're an expert you probably don't care about friendly but for new users like if you first got a fish shell and then you learn

bash there's probably a lot of bash things that you would think were so silly. Don't you find that Phish kind of does things by itself too much though? At least with Bash, you type stuff in, you press enter, okay, you've got tab completion, but if you don't know about that, you're not going to get it. Whereas you start typing and then just things happen and you're like, it's a bit scary for people who don't know what they're doing. Okay, I can see that. You probably do need, in a basic introduction,

you probably do need a little bit about that. You probably have to touch on that feature. It feels like to me you have to have used Bash and have some vague understanding of how it works before you could use Phish. I guess I could see that with the very, very present autocomplete. I can see that being confusing. Yeah. But I wonder if you started off from a blank shell, like with no autocomplete history,

and if it would make more sense that it would, like, if it was only commands that they had typed that got suggested to them again, if that would be better. I don't know. It would be an interesting experience in user experience research. Yeah. I can't see many companies putting money into that, though, unfortunately. No, no.

That's exactly what I was thinking. No one's funding that.