 Okay, so Linux is the wild, so I'm Luke, I am Luke Smith. Oh, I should stand behind the mic, shouldn't I? So my presentation is entitled Linux is the Wild West, which of course was decided one, the night of the application being due for this conference and just sort of came to my head. It was only after that that I realized that Linux Fest generously gives an entire hour to talk about whatever I want. So I'm gonna talk about Linux. I wanna talk specifically about minimalist software, why it's, I guess, important for Linux, but not so much in itself, but sort of as a greater, there's a more general point that I wanna make that's sort of about the nature of decentrality, decentralism in design features and other things. That's relevant to my own work. So if you don't know me, I'm not any kind of IT person. I actually am a doctoral candidate in linguistics. So actually who knows me? Who knows who I am? Oh, okay, all right. All these people who were awkwardly walking by me and didn't say hello, but hello everyone. So I run a YouTube channel, maybe I'll talk about that in a second, but it doesn't really matter. So anyway, and I will say there's an alternate title to this and that is the future is decentralized, but I thought that sounded a little too much like a TED talk, so I'm just gonna skip that. So anyway, there's this traditional dichotomy that people in the free software movement have had, and that is between free software and proprietary software. And of course, when I say free software, I mean in the FSF since free Libre, open source, distributable, visible code against proprietary software. So this is traditionally how Linux users have sort of defined what they do as opposed to what Microsoft or Apple or something like that does. And that sort of defines the identity of people in a lot of ways. And don't get me wrong, I like free software, I'm a big fan of it, I have two Libre booted thing pads and I always support using free software when you can, but I don't actually think that's the most important issue or really the defining feature of Linux as we know it. And of course, I'm not talking about the kernel, I'm talking about the general environment of Linux as a aggregation of a bunch of different tools. And then as I really think what is the important thing in Linux is minimalist software versus bloat. I don't know what to call the opposite of minimalist software that doesn't sound bad. Maybe enterprise, big programs that have a lot of features that try and do a lot of things. And I think that's really, at the core of it, that's more or less the important thing that distinguishes Linux from, or not just Linux, but Linux and the BSD, that whole world of software from everything else. So one example, this is actually something relevant to my own work. Oh, why am I still wearing this if we're not using it? So this is Herbert Simon. Who knows Herbert Simon? That's exactly how many people I was expecting. So Herbert Simon was, it's hard to describe him. We can think of him as being an economist or something, but he worked on something you could call cybernetics. The study of systems and stuff like that. And he has a very nice parable in one of his works and it's called The Parable of the Watchmakers, which I have a nice watch here that you can look at. So the parable of the watchmakers goes like this. So there are two different watchmakers. One of them he calls Hora, which is the Latin word for hour. One of us, one of which he calls Tempus, which is the Latin word for time. It doesn't matter their names. But the idea is Tempus, his idea for constructing watches is, so there are 100 pieces that go into a watch and he wants to minimize the amount of steps in that process. So he takes the 100 pieces and he sort of fits them together into one big watch all at once. He sort of clutches it in his hand and finagles it around. And that's his way of constructing watches. On the other hand, the other guy, Hora, his idea is basically that's too difficult of an operation. What I want to actually do is make watches sort of from the bottom up. That is, take the 100 pieces, make a subcomponent of 10 pieces, another subcomponent of 10 more pieces, et cetera, et cetera. So you have smaller pieces that go together to make this one whole. So Simon's whole point of using this is there's a sense in which the more decentralized model when you're building watches from subcomponents, this is actually a much better way of organizing your life or business. And if you think about it this way, so if you're building a watch with 100 pieces all at once and you hear a knock at the door and you have to leave it, all your progress is gone. Whereas if you build subcomponents, those are all entities in themselves. If new technology comes, you have to replace pieces or something like that, you can easily replace them if you build subcomponents, but not if you have a bunch of stuff they'll fitting together at once. So this is the general idea of it and it's one of the foundational ideas in cybernetics. And that is decentralization, right? So here's an image from the first page of Google images or whatever. Centralization of course is everything going together at once, everything being operated by one sort of core. Whereas decentralization is subcomponents that work together. Now at the surface, decentralization is a little more complicated in that you have more nodes and stuff like that. But it's actually in this sense it's a lot more useful. So we're all, I guess most people are familiar with the cathedral and the bazaar. You guys know this most people, okay? So the distinction of course is between software like in Microsoft or something like that is usually designed as if a cathedral is working on it, a whole bunch of people sort of have a set idea they go through it, as opposed to a bazaar type construction where everyone's, it's all, it's a total shit show. Everyone's doing different things. Everyone is sort of working together to make different, yeah. Right, right. But it applies at different scales, yeah, yeah, yeah. So, but the point general is what I'm getting at here in that, so you have, there's a kind of construction which is not necessarily even organized by a central authority but comes sort of from the bottom up, there's an illustration but we can skip over it. So my idea here is basically what defines Linux isn't necessarily that it's free software, that is an important component of it sort of ties into this but the programs are small and predictable. When you have bugs, since those bugs are in small sub-components you can easily figure out where they're coming from. And the system is decomposable. You can take a piece out, do something with it, do something else, replace it. And there are other benefits that slide on resources. It's easy to maintain stuff like this. So I'm a big proponent of this kind of minimalist view of software. And you might be familiar with the UNIX philosophy which sort of ties into the same thing, right? Programs should do one thing and do them well. And if you have a program that's doing something a little, a little more advanced than that, it gets a little confusing. So a lot of people will come to Linux nowadays. A lot of people want to be Linux evangelist. Now I'm not actually a big Linux evangelist but I've always had the perspective of letting them judge you by your fruits, so to speak. Show what you can do on Linux, don't necessarily throw it in people's faces. But I will say one idea that I think a lot of people have when they're dealing with how do we make Linux accessible to other people is, what we should really be shooting for is making one cohesive environment, like one Linux. Like maybe even like one set desktop environment or something like this, but really one big thing that is Linux that is a brand that we can market and that's one thing that makes and everyone's doing sort of the same thing on it. Now I'm very much opposed to this. First off, there's the problem that we already have that and it's called Windows or macOS in the sense that it's sort of hard to compete with something that's basically the same thing and already established. So originally I actually, I got my undergrad in economics. One, and I worked on game theory and stuff like that which I actually still do. So game theory is just the analysis of strategic behavior or something like that. And one common example in it is sailing. So you of course can have these sailing races. And the interesting thing about sailing races, let's say there are two people in a sailing race, you can tack left and right to go for, to sort of catch the wind and go in certain directions and basically your strategy works out in this way. So if you're the leader, what you wanna do is sort of predict what the person behind you is doing. So if you go right, he goes right. That way if you run into some turbulence, he does too or vice versa. So he can't get ahead of you if you always make the same decisions. On the other side, if you're behind, the person who's behind in the race, always wants to make decisions that are different from the person in the lead. Because if they always make the same decisions, they're always gonna catch the same waves, they're always gonna get the same wind. And I think the same thing is true of Linux. Like a lot of people have this idea that competing with macOS or something like that or competing with Windows is an issue of making similar design decisions that they've made in the past. And I think that's not incredibly right. Now of course that makes, that some users are attracted to easy to use features, but I think the important thing is actually making Linux distinct from these. So a bad idea would be making a Windows or Mac clone or something like that. I'm not, I'll explain this in a bit. But my idea is basically leave the chaos in place, leave these small decentralized programs, let them be there, but give new users tools to make sense of it, make it easy for them to do that. Because you might have a chaotic world, but if people have the resources for navigating the chaotic world, that's a good thing. So anyway, how do we do this? So, well I don't know how to, I don't have some authoritarian idea of what exactly we should do, but I'll tell you what I do. So those of you who I guess know me from the internet know that I have a YouTube channel, and it's usually noted, because I have these colorful thumbnails with memes on them that aren't related to any of the content on the channel or something. It's just sort of a joke. So my thing, I actually started using Linux I guess three years ago or something like that. And really just when Windows 10 came out. And my perspective then was just, I'm gonna start putting up videos on YouTube, but it's just gonna be things that I happen to be doing today. Or silly stuff like that. And you can make a bunch of thumbnails and people start watching. And my objective has always been making just little tutorials that are maybe five, seven minutes that show you exactly how to do something, not droning on about how I feel about Linux or something like that. That's what Linux Fest is for. But anyway, so another thing that I really like, that I think bridges the gap. So in addition, I think that one of the things that I think is important to remember is the software is out there. It's just an issue of making it sort of accessible to new users. So I like making tutorials. And the other thing I like is wizards. I'm not actually, I don't think of myself as being a computer programmer. I don't really write programs. But what I do like is wizards. I love wizards, they're fantastic. Because the thing is if you create a wizard for putting software together, you're not adding more onto the stockpile of all the stuff we have, but you're making that stockpile easier to navigate. And that's what I'm a big fan of. So one issue that I had when I started using, so well I should say, if you don't know my channel, I don't just use Linux. I use basically a very highly configured setup with using a Tiling Window Manager, i3 Gaps. I use basically all terminal programs besides my browser and stuff like that. I don't use like the TTY, I use it in a graphical environment in i3. And my objective is pretty much using keyboard, programs that are easily accessible by the keyboard. Don't really use the mouse. And basically minimizing the time that I use, the number of key presses it takes for me to do something. That's basically my idea. So one program that I ran across sort of getting into this was MUT. Anyone know what MUT is? Anyone know? Okay, oh, most people. Great. More people than Herbert Simon, of course. So MUT is actually a really fantastic program. It's an email client. And the thing about MUT, it is a huge pain to get working. It is, it's on a, well, actually, who uses MUT? Okay, wow. Look at that. There's like at least two and a half people. Well, I do too, so I should put my hand up too. I love MUT, it's fantastic. But if you've, well, actually, let me ask this. Who's tried to use MUT? Okay, okay, more people, okay. So that makes sense. So MUT is one of those programs that's fantastic. I really like it, but it takes so much effort that you have to pour in to get it working. And after that, I sort of had the idea, actually, what's, oh yeah, well, here's what it looks like, but you can't really see anything there. So what I ended up doing is, it basically took me a year to get MUT configured the way I wanted. And that's relatively typical. But I realized after I was done with this, I could basically list out all the stuff I did, like all the stuff that you actually need to go through. Now, I don't just mean MUT configured. I wanted offline email. I wanted my passwords to be automatically accessible and encrypted, so not just any person could see them. I wanted indexing, I wanted all this stuff, all of these moving pieces working together. And that's what took so long to actually, configuring MUT itself doesn't take too long, but getting it working with all these pieces is a huge pain. So what did I end up doing is actually writing a wizard for this, like literally, and it took literally just like one weekend. So what this wizard does, it's just an incurses menu. And it's not just for MUT. So what it does is, all you do is go to it, you give it your email address, and it will auto-detect a bunch of information for that. But it'll give you an automatically configured MUT RC. It gives individual RC files for each email account you input it, so you can actually use multiple, you can go through it multiple times and add multiple email addresses. It gives you keyboard shortcuts for jumping between those boxes and accounts in MUT. So you can jump back and forth. So if I have five email accounts, I1 goes to my first one, I2 goes to my second or something like that. Or you can jump to inboxes and stuff like that. All of that is generated automatically. And this is actually a huge pain because different email services, of course, have different names for their sent mail or something like that. That's something very useful. So also, I have it configured with offline IMAP, which is a fantastic program that will mirror excuse me, an IMAP, the contents of an IMAP email account on your actual computer. So if you actually just configure MUT itself, it has to, you know, you have to be connected to the internet or something like that. And I was like, I don't wanna do that. I actually lived a year or so, or more than two years without internet, like at my house. So I wanted to have email, you know, I wanted to be able to respond to them at my home. So I wanted something like this. So the wizard does all this, configures MUT, it configures offline IMAP, it encrypts your passwords with GPG. And you can also set up a cron job so that it checks your email accounts at every so many minutes and also uses not much to index email. So yeah. I'm sorry, you said you did not. No, well, I would write email at home. So I would, you know, I'd get, usually get a bunch of emails from like my YouTube channel and then I'd, you know, download them when I'm at work or something like that. And I'd write out my responses at home. So internet at work? Yeah, yeah, that's what I use. It's not like I didn't live in a cabin totally alone. But, and by the way, if any of you guys are thinking about like, if your internet is about to expire at your home, I am a full proponent of just not having internet at your house. I think it's fantastic. It's really great. Yeah. Everyone who's done it is like, oh man, I'm so free. You realize how much time you have. Yes, he says no. Yeah. So, yes. Cell phone. Cell phone. Oh, I hate those things too. The only reason I have a cell phone, only reason I have a cell phone is because, you know, my parents still pay for it. So, you know, whatever. I have to have it. They'll complain if I don't. But yeah, I think cell phones are the same. Like, yeah, realistically, I'm gonna just have a landline when I have that ability. I hate those things, but I really just never use my cell phone. I could. Well, I will say, I basically just keep my cell phone at my house all times. Like, I happened to, I happened to, oh, I didn't think that's funny. I thought a lot of people did that. I don't know. I hate phones so much. It's terrible. And you know what? Okay, this is not relevant to my original thought. Oh, I got, okay, never. I have plenty of time, so I'll tell you. So, one funny thing, it's the state of our society. Oh God, I hope my dad isn't watching this, but he might eventually. So, one of the things about my dad is, he plays Pokemon Go, you know what these adults nowadays are playing? Like, it's crazy. Like, 20 years ago, 20 years ago, I was, my dad's probably watching this. Hey, dad. But 20 years ago, he used to complain of me playing Pokemon at the dinner table. And now, like, every time we go out, he's playing Pokemon Go and I don't know. Like, he came out, I recently moved from Arizona back to Georgia after a while. And he came out to help me in Arizona, but I secretly suspect it's to get, like, sand slashes and stuff that they don't have. Like, that's, he was so happy. He's like, wow, I've never seen a Lilith. This is great. So, that's our society. So, yeah, I'm against cell phones too. And it's also insane to me that, like, 10-year-old kids have cell phones. I don't know how this happens, but anyway, so what was I talking about? Wizards, wizardry. So, the thing about, so, yeah, what the wizard does is it configures MUT for you. It configures offline IMAP. It indexes your mail so you can easily search it with a keyboard shortcut. It gives you a lot of what I consider sensible defaults, encrypts your password, all this stuff. And on the user side, it just works. Like, you don't have to know what's going on, but it's totally fine. You don't actually have to think about it. Just follow the directions, which actually is sometimes hard for people, but, you know, anyway. So this, I actually, I've done other stuff, I've done other wizards, but this is the one I like the most, just because it saved me, like, months and months, well, it didn't save me months and months, but it saved other people months and months of time configuring this program that gives you something really, really sweet, and it gives you all the bells and whistles, and it's something you can do in, like, literally three minutes. So, yeah, and this is what I'm talking about. That is, you don't necessarily need to give people a cohesive environment so long as you give them, you help them along the process of configuring them, and, of course, all of the moving parts in MUT, offline IMAP, not much, all that stuff, all of them are still available to users. All of them can easily, oh, I want to only sync these particular folders. So, oh, all I have to do is edit my offline IMAP RC file, and that's it. So it's supposed to be, it's not supposed to make, it's supposed to make installation easy, but it's supposed to make it even easier for you to get further into the configuration. So another thing I do that harms me, probably, is use Arch Linux. Now, I started using Arch Linux, actually, soon after I started using Linux, generally, and the nice thing about Arch is, actually, all the things I don't want to deal with are really easy to do. They're done automatically. But the nice thing is, because it forces you to have a minimal install, you don't have to deal with taking away all that software you don't like when you originally install an operating system. And that's what pissed me off about everything else, I guess. So I used Arch Linux, and I went through a period where I had to install it probably about 20 times in a week. That is, a hard drive would break, and I'd be like, ah, I should probably install it here, or oh, I have a couple computers, and I have this terrible, like I just have to have everything perfectly synced. I don't know if you guys are like that, but so I wanted everything to be pretty much running the same thing. And of course, I was still in Linux newbie, so it's like at that situation where, oh, I messed up something minor in my install, and I don't know how to fix it, so I'm just gonna reinstall the whole thing. So that's pretty typical. So anyway, what I ended up doing is, time and time again, I went to install Arch, I'd install all the programs I wanted via Pac-Man, all the ones I use on a daily basis. I'd have to activate all my services and stuff like that, user permissions, all this stuff, and I'd deploy my own .files, things like MupWizard, stuff like that. And this process too got sort of annoying, like it's something that I ended up doing time and time again, and so I figured I might as well just automate this. And what actually happens is, a lot of people on YouTube would be like, hey, I like how your computer looks. How do I make my computer look like mine, like yours? And I don't wanna tell people, oh, that's easy, just months and months of like strain and stress, and you'll get that way. But that's actually not true at all. It's actually very fun. It's lots of tinkering, lots of playing around. That's really what it is, but it takes a long time. But eventually, the instructionals that I made for YouTube turn into scripts, which turn into really cohesive systems. And that's when I made what I call LARBS, which is a ridiculous sounding name, so it has a ridiculous logo. It stands for Luke's ArchRice Bootstrapping Scripts. It's pronounceable, so I went with it. So the idea behind this is, this is really, it's not an Arch install system. It's sort of a way of deploying my configs. That is it, basically what it does is you take a fresh install of Arch. It installs the programs that I want. It reads from a CSV file. This is the program I want for this. If I've selected an option, so I'll install this. So it installs all the programs I use at a daily basis. It installs things like Network Manager, Pulse Audio, all the basic things. And then of course it uses my .files. And since I put this out, there are actually hundreds of my users who use this thing, which is actually sort of weird, it's weird. And it's like someone wearing your underwear when someone uses your .files. It's just very strange to me. But people do it, and not just that. The system is actually, you can take it and use it for your own .files. So you can replace it with your own stuff and run with it. So that's it. Where are we on time? Okay, not bad on time. So, oh, was I supposed to read that? How about you should I use the Ansible or something? What's that? No, yeah. Are you using another tool? Yeah, well, I always feel like I'll just do it my own way. That's my perspective. There are a lot of tools that do sort of similar things. I mean, it's the same thing like, oh, when I run my website, there are a lot of things that do 99% of what I want, but I'd rather it do 100%, so I'm just gonna write it myself. That's just my own thing. But yeah. So the goal of it, of course, is not just to make things easy. It does sort of do that, but the goal is to remove the barriers to entry, but leave the customized ability to me or whatever other user. So here is a system. Here is a working Arch Linux system that looks, well, it looks cool. It has all these programs. But the best thing is, here are the config files you can edit to do your own things on it. Or, and it also keeps, importantly, it keeps users aware that they're not just using my configuration, they're not using one cohesive system. This is a bunch of little parts to go together and you can change each and every one of these parts. And that's sort of the goal in it. And getting people to actually read the manual. That's nice. Which everyone, here's not a lot, but it's actually good advice. So anyway, the point is a decentralized and extensible system isn't actually hard, but it can be new to people. Like we're not necessarily used to it, but it's something you can overcome, not just with instructing people explicitly, but you can use wizards and stuff like this. And I think a lot of people should focus some of their, I mean, it's what I do. I like focusing my effort on sort of making this easier for people who are newer to Linux. So anyway, one example that I use a lot is Microsoft Word. So I, like a bunch of other people my age went to school and of course learned how to use Microsoft Word. Now everyone will say things like, well, let's put it this way. I think that people think of the sort of programs like Word or other what you see is what you get editors as being easy. And they think if I tell you instead, oh, write a document in Pandoc and then use, or excuse me, Markdown and then compile it with Pandoc. And that sounds complicated. It's actually extremely easy. It takes literally one command and it's very easy to do. But there's this sort of veneer of facility or easiness in the what you see is what you get editors. And I wanna say that that's really just a historical thing because back in the 80s or 90s or whenever it was, I guess it was the 90s, the government goes to Microsoft and says, hey, you know what, we wanna teach kids technology. What should we do? And Microsoft of course says, oh, how convenient. We just wrote this great program, Microsoft Word. We'll give it to you half price or whatever if you teach it to all your kids and the rest is history. But one example I use is what is actually a more productive use of time. Now, when I was in first grade, I learned Word. Everything I learned in first grade about Word wasn't true two years later when they changed it. And what I learned after that wasn't true after they moved everything around again. But if I had spent my first grade learning something like grep or awk or something like that, that sounds more complicated because we're used to that being, oh, that's the hard stuff. That's the programmer stuff or something like that. But conceptually it's not necessarily, like using grep to take, you know, organize a file or extract output from a file is something that's actually relatively easy. It can be something useful. And I think that if it's very easy in the same way, I'm not suggesting we all, you know, forcibly teach core utilities in school or something like that. I think it might be a good idea. I just don't think it's gonna happen. But I will say what I'm trying to say is it's something that's not necessarily difficult. Now, I think that Linux as an operating system is actually part of like a more general movement in different affairs. Now, first off, Linux as I said at the beginning is sort of like a decentralized environment. It's not just like one cohesive thing. It's a bunch of programs. There's not really one, there's the Linux kernel but no one really who actually has opinions about the Linux kernel. When you say Linux you mean everything out there. So really it is a weird sort of example of something that's highly decentralized that isn't necessarily winning but it's gonna be here a long time. And it's something that has sort of gone against the expectations of many people. Now, one of my favorite copy pastas that I get, some of you guys might have seen this. This one's actually from Adolf Hitler. But this is something you might see around. So are you saying that Linux can run on a computer without Windows underneath it at all? As in without a boot disk, without any drivers, without any services? That sounds preposterous to me. If that were true and I doubt it, then companies would be selling your computers without a Windows. This is clearly not happening so there must be an error in your calculations. I hope you realize that Windows is more than just office. It's a whole system that runs on the computer from start to finish. And that's a very difficult thing to achieve. A lot of people don't realize this. Microsoft's just spent $9 billion and many years to create Vista so it doesn't sound reasonable that a new alternative could just snap into existence overnight like that. Liada, yeah, you get the point. So you might have seen this as copy pastas. People post it all the time. But it actually illustrates something that people actually think. That is, people hear about Linux, they hear about this, oh, people write code for no particular reason because they're not making that much money on it, right? Like you need some kind of enormous company to build top-down this operating system. And it's sort of, people think of Linux as being a kind of miracle. But my argument more generally is that Linux isn't really a miracle. It's something that again falls out from these decentralized, the fact that it's so decentralized gradually building up into a larger structure. Now, as I said at the beginning, I think that Linux is just sort of, there are many examples of this. I think we're moving to a point where, I as an academic, we're getting to a point where people are starting to realize that the core of things isn't just planned systems. It's things that are sort of built bottom up. So traditionally, we've had this idea that, well, something is rational if we sit down in a corporate boardroom and we plan it and we say, oh, this is what this is gonna do in the operating system. This is what this is gonna do, maybe in some socio-political affair or something like that. And this is an idea that people have because it appeals to your sense of logic and it makes sense to people. But the reality is, and the reality that I think Linux demonstrates very strongly, it's really a very small, the reality is the world is a much a very small and decentralized systems that sort of play into each other. So this has actually sort of become a meme decentralization. As I said, it sounds like a TED talk at the beginning. So that's one of the reasons I feel weird about it. But this, actually, this guy's sort of a meme on my, yes. Just really quickly. Yeah. You know everybody's here. We're on the park or whoever, back in the day. We don't have a collaboration board. Right. And we wanted to hear about it. Right. But it was probably created by people that were, experiment or whatever you wanna call it. Yeah. You know, we're not necessarily a monetary thing. Right, right, yeah. And that's my point. Actually, perfect slide. So this, I don't know, you guys know Nassim Taleb. He's a meme on my channel. Yeah, he's a great guy. So one of the things that actually he talks about is how scientific discovery is, isn't an issue of sitting down and planning, we're gonna discover this tomorrow. It's an issue of just sort of messing around, you know, if you're a chemist, sometimes you just play around with chemicals and you tinker a little and you find something great. And I totally recommend this guy, by the way, if you should read Anti-Fragile. Actually, I'll go ahead and spoil the book for you because it's relevant to the talk. So one of the concepts that Nassim Taleb, and he's part of this more general, I guess intellectual movement to decentralization, one of the concepts he coins is what's called anti-fragility. That is, you know, something is fragile if you expose it to sort of stress and it breaks. Whereas something is anti-fragile if you expose it to stress or something dangerous and it actually gets better. And there's a sense in which Linux is exactly like that. In the sense that like, it's nice to have a bunch of little programs and it's nice, one of the best things about Linux is that it's buggy. There are programs that mess up, bugs happen, they happen all the time because there's so many moving parts. The great thing about bugs is you can, especially in an environment where everything is decentralized is, you can deal with them at a very local level. They cause local damage, the world doesn't collapse, your entire operating system doesn't collapse or something like that where you have a bug. And when someone complains about Linux being buggy, I think that's a good thing because effectively what it shows is all the problems in coding that are happening in Linux are visible and we can deal with them. Whereas if we deal with like a operating system that is closed-source, they're not necessarily visible to us. That's something that, it doesn't work that way. It's more dangerous because when the problems actually show up, they're catastrophic. Oh yeah, that's sort of just what I said. So anyway, so this is something that defines software, it defines a lot of social affairs and it defines how we, even how we model things scientifically. So again, in my field, this is Jerry Fodor. I don't know if you're familiar with him. He wrote the modularity of, Miles knows. He wrote the modularity of mine. Actually, who knows Jerry Fodor? Oh, not as many people as I thought. I thought he was like a household name. Okay. Yeah, I guess, okay. Your house. Yeah, my house. We, all of us know about. Okay, so Jerry Fodor is sort of a philosopher. He's actually recently late, he died recently. That's what I'm trying to say. So yeah, one of his ideas is effectively the mind is modular. That is, it's not one big unit. It's actually a bunch of subunits and stuff like that. But one of the nice things about this is, this is sort of a decentralized idea, but it's now come under fire by people who are even more extremely decentralized. That is nowadays, you know, you have the rise of what's called PDP or connectionism that's parallel distributed processing. And the idea behind it is, you know, this is like neural net kind of stuff. The idea that has to do with it in sort of cognitive science is that basically we don't so much even have modules, but the level of, you know, the level relevant in even neurology is lower. That is really what you have is small neuronal elements that interact with each other in very complex ways. And the architecture of the mind is so decentralized and distributed, you don't even have modules. You have something even more, even more diffuse. And one of the reasons, for example, if you knock your head against a wall or something and you don't immediately die is because your brain, in a sense, it isn't necessarily antifragile, but there's a lot of redundancy in it. If you lose part of your brain having a stroke or having severe damage, it can recover and stuff like that. And this is actually, you know, is the field more generally of cognitive science is sort of moving in this direction. Even though, you know, all my advisors and stuff like that don't like it, but it's sort of because they're old and they're, you know, they call everything, they don't like behaviorism, but it's, you know, whatever. Skinner didn't do anything wrong. So neural nets, so the other thing, actually this is relevant to my work, so this is in my mind because we went over it in a seminar recently. But, you know, one of the things about a lot of scientific modeling, again, I'm a formal linguist, is that there are a lot of domains and cognition that aren't even like modelable without highly decentralized networks. So something, one of the things I work in is how the prosody of language actually affects things like word order or the syntactic categories of language or something like that. This actually is an article that argues that really, I would go into it more, but I don't know how much time we're gonna have, that really the nature of language is such that you need some decentralized, because so many of the modules of the brain are looking at each other, they're recursively defined, you can't necessarily model something without a kind of neural net or a highly decentralized network. That's not a neural net, that's a neural net. So anyway, so anyway, the idea behind this is really the brain obeys the Unix philosophy, not just the brain, but a lot of things functioning in society in more or less obey the Unix philosophy in the sense that they do one thing, they do them well. And a lot of the aggregate features we observe in the world, they're not designed, they come sort of their emergent properties of the smaller features. I was like, oh, I already talked about this. Anyway, that's a syntax. Oh, well, one other aspect I should talk about. So this is a syntax tree and linguistics, we do those all the time, they're all fake, they're not really real. But one of the things that people notice about the structure of language is, even though verbs, if you think of verbs as being like functions, some verbs seem to have multiple arguments, stuff like that. But one of the findings of linguistics generally is that languages are usually very decentralized, they're basically binary in their structure. Like if you draw a sentence out, they're basically as decentralized as possible with as many nodes. And this is actually, we're coming full circle because as I mentioned, the beginning of the talk, there is during the 20th century with Herbert Simon and a lot of other people who worked on cybernetics, there was the idea that of the watchmakers, they're watchmakers paradox, yeah, the watchmakers parable where decentralized systems as you construct them are actually a lot more efficient, they can survive stressors better, they can survive, there are more future proof, that's another TED talk word. But in general, that's the idea behind it. And this is, I only put this book up because Jeff Samson, who's an interesting guy, noted this as language as well. So anyway, I think we're wrapping up at this point. So my point generally is people in free software, people in Linux, they shouldn't shy away from the minimal, the things that have bugs, the things that are hard to understand for introductory users. And I think people should focus more on making things accessible. That is making things accessible to new users, and not by like throwing more bloat on top to make sense of it, to give sort of a familiar environment to it, giving people the tools they need to navigate the sort of Linux world. So my TED talk takeaways are overcomplicate design, that should be a D at the end, overcomplicated design, okay, that's not even a sentence, forget about it. Overcomplicated design is bad and small is beautiful. That is, if you can't break it, you can't improve it. That is, if you can't find bugs in something, you can't improve it. And it's great to have bugs that are at a small level and that are at a larger level. And make the learning curve easier, make it give people tutorials, give people wizards, make the life easy for them. But don't get rid of learning curves because that is just sort of paving over the problem. It's exporting it for later on. So that I think the next slide is conclude, oh yeah, that's it. So that's my contact information. And I guess that's about it. If you guys want to do questions or something. Yeah? Well, I have probably realized that you can't prefer it to Linux as like this whole system. Right. And I mean, I would say that like Richard Salman. Oh yeah. You know, I disagree with you completely. Well, I think everyone- So he knows like really what you're referring to, like all the GNU stuff? Well, it's even bigger than that. It's even, well yeah, exactly. That's why I tried to hedge my bet. And it's not even GNU Linux, it's greater than that. When I'm talking about it, I'm talking about like all these programs that, I mean, the entire environment of development. I think everyone understands if I use, if I say Linux, most people sort of understand, ah, yeah. But yeah, I'm not talking about the kernel and I'm not talking about even GNU Linux. I'm talking about something more general. So, yeah. Yes, it's a little bit like, you know, it's like the concept of computers in general. Like, you know, there's this thing where, in education, it's a broad statement. You know, people say, well, certain people may not have access to stuff and they don't have money, they don't have the best computers, they don't have iPads and stuff like this. But, you know, a lot of people in this room, they know that, you know, like you said, there's a lot of powerful utilities and things like reds and things like that that have been here for a long time and are gonna continue to be here. And, you know, just because you can, you know, work a program, a tool or product per se, doesn't necessarily make you smarter, especially when they change it later. Yeah. So, you know, there's apps and there's kids that can use apps and things like that, but I see value in teaching people, you know, for stuff, you know. Right. That makes sense. Yeah, yeah, it does make sense. In fact, one of the things about Linux, so I don't necessarily think Linux is gonna become, you know, the number one OS anytime soon, but I think there's a sense in which we can say that Linux will outlast both Microsoft and Apple. And the reason I say that is because both of those are just centralized authorities that are tied to particular companies and who knows what happens in the next, maybe next month, maybe next 100 years, but those two are eventually gonna pass away. But something, a distributed system like Linux or something like that, or, you know, whatever you wanna call it, the GNU slash Linux general environment or something like that, that's something that's gonna be around. So for that exact reason why you say it, that's something that's going to be, I mean, it's valuable learning stuff like that. Yeah, right, right, yeah. So, yes, right, right, right, right. Right, no, that's actually entirely correct. I mean, so take the most extreme like FSF positions on everything. I don't necessarily agree with them, but I like that they're there. I like that there are people who will throw a fit whenever they have to use something proprietary. That's a good thing. We might think that's weird, but that's something that's a good thing for everyone because if most of us are going to be sort of ambivalent to if they use free software or proprietary software, if we're ambivalent to those, obviously the powers that be or the corporations that have a motive in it sort of want to nudge us every little bit to proprietary software. So, yeah, I think that's a good point. Yeah. This is broken. Yeah. And everybody was mad about it. Nobody, you know, really there was only one version of Windows and it was broken. Right. In Linux, like if Ubuntu puts out a bad version, it doesn't break Linux. Exactly. And that's the big thing people miss about that. I remember back in the day, people were calling for, we should have one desktop. We have one desktop, everybody come to us. Well, that has worked for everything. Yeah. An old laptop, you went from open box and something else when you run something else, and Linux is good the way it is. Right. Yeah. I just, I'm going to repeat that for the microphone because that's a good point. Yeah. So his point was, you know, it's a lot of people have said that, you know, oh, we should just have one distribution, one, you know, desktop environment or something like that. But yeah, it's having all your eggs in one basket. Right. It's not very conducive for the long term when things mess up in them. The community came out, out of time, or my life then. Right. It was moved over to Dora. Right. Right. It's nice having that optionality. Right. It's more like a state of mind kind of thing. Right. But that's all over the conversation. Yeah, well, the way I put it, my subscribers always make fun of me for saying this on YouTube. But yeah, Linux is sort of a social construct. It's not really, there's not really one essence of Linux. There are just many different working parts and you put them together and it looks like a Linux. So that's what it is. Is that the kernel? Yeah, well, the technical definition, you know. Any other years from now or something? Yeah, you never know. But the important, like even when Linus is gone and the kernel, you know, who knows what happens to the kernel. The system in general still survives. Right. It's easy enough to, there are other kernels out there and yeah. We have Linux distributions run on Windows and that is everything except the kernel and we still call it Linux. Right, right. Yeah, and I also feel, yeah. And I think, I mean, one other example is like, so Android is Linux. But is it really? I mean, it has the kernel, but it doesn't really feel like the same thing. You know, the term is... Is it the kernel one? Well, okay, yeah. I mean, don't get spun. Well, yeah, but I'm just saying, in the sense of Linux that I've been using it here, it doesn't really, yes, I know that it's technically Linux, but it feels wrong. Yeah. Right. Yeah, exactly. Linux is very unique. Yeah. Yeah, I think we just need a word for that. I don't... The word is unique. Okay. Yeah, well, I should say the, I guess the title of the talk should be, you know, the Unix environment is the Wild West or something like that. Yeah. All right. Any other questions or? Just rumblings? All right, I guess that's it. Well, it was nice talking to you guys. Yeah.