Faris Masad - repl.it

{/ TAB: SHOW NOTES /} This week we're joined by Faris Masad, a founding engineer at repl.it. We explore how repl.it is breaking down the barriers to coding by provide in browser coding environments. We also discuss the importance of building a community around your product and how creative repl.it's own users have been. Join our patreon for the full episode. Twitter GitHub repl.it {/ LINKS /} - https://www.inkandswitch.com/peritext/ - https://github.com/replit/evalbot Tooltips Want to hear use talk about our tooltips? Join our patreon! Andrew - https://github.com/binwiederhier/replbot - https://pothos-graphql.dev Justin - [Book] Four Thousand Weeks by Oliver Burkeman (share the fantastic summary from Matt Swain. - [Podcast] https://www.hackingyouradhd.com/ Faris - https://astexplorer.net/ - https://github.com/graphp/graphviz {/ Paste show notes /} {/ TAB: SECTIONS /} [00:01:11] What is repl.it? [00:09:58] How does it work? [00:19:01] How does it compare to competitors? [00:26:45] Community Creativity [00:33:22] Future of Development {/ TAB: TRANSCRIPT /} Episode 40 - Free Faris: we are looking to build economy within repl.it. I talked about the network and how people interface with each other's repls and all that. Having an economy built within that network would be really cool. Andrew: hey, before we get started, we'd like to announce that we've opened up a merchant store. Now you can rock a hoodie to show your support for the show. If you're a member of our Patrion, you get a 15% discount code. Now let's get onto the conversation with Faris. For a member. The full episode is only available to our members. Hello, welcome to the dev tools, FM podcast. This is a podcast about developer tools and the people who make them I'm Andrew, and this is my co-host. Justin: Everyone. Uh, today we're really excited to be joined by Faris Masad. Uh, Faris is one of the founding engineers at repl.it. Uh, and for those of you who might not know it is a service for quickly spinning up a repl environment to do, uh, sort of coding on the fly. Maybe I butchered that description. [00:01:11] What is repl.it? Justin: So Faris, uh, would you like to tell our audience a little bit more about yourself and then maybe after that talk a little bit about what repl.it is? Faris: Yeah. Um, I'm Faris I'm, as they said, I'm a founding engineer at repl.itI today I work mostly in the core product. So that's like the IDE environment, if you can call it an IDE. Um, and yeah, repl.it is. is an online environment for authoring software in a way that doesn't make you wanna shoot yourself. Uh if like, yeah, if you look at, um, the usual way people like software, you go online and you, you go online, you download a bunch of tools, like you download your code editor or, um, you know, maybe, maybe even you start like code spaces, but then like you, you have to figure out what to do, right? Like you have to figure out how, how do I even like, um, run my code? Like, what does, what does running entail? Um, what does, what do I need to get into a state that, uh, I can run code? I remember my first experiences with coding early on, I guess it was like visual basic at first. And that was pretty fun. Uh, it was, uh, it was a really good experience. You like you download, uh, I think it was visual. What was it? Uh, I think it was visual studio. I can't remember, but you, you, you download a Microsoft product and, uh, you have this UI that tells you what to do basically. And you know, you, you draw your buttons and you attach an event listener to them and it was like pretty cool, but then like I wanted to do real development. I think I, I wanted to try Python at the time, uh, cuz I saw Google was using it and uh, I downloaded Python or so I thought and I downloaded a bunch of editors cuz like, you know, there's different recommendations and um, I just didn't know what to do. I followed like 15 different tutorials. Uh, I set Python in my path, uh, like six different times and you know, I still, I still type Python main.py And it's like, it's not the thing that I'd expect and it, you know, um, and Amjad actually the reason uh, repl.it is here is because Amjad was like just sick of this. Uh, he was in university at the time and um, he basically noticed that every other lesson they spend it just setting up the environment like they don't code, they're like installing stuff and they're, uh, removing stuff and they're making sure like the previous student's environment isn't clashing with theirs. And I, I went through the same experience. I actually like quit university because of this, uh, dropped out. I was like, I'm not learning like how to build things, I guess. Like, it was useful for the theoretical aspect. And, and I, I went in like, pursued that on my own, like followed the, um, followed some curriculums, but it takes a long time before you, you get to the part where you're building things. Justin: Yeah, it reminds me of my story. Uh, I tried to teach myself programming in high school. I was interested in it from like an abstract perspective though. I'd never had any like direct, I didn't have like a programming class in school or anything, but I mean, not knowing, you know, anything about programming or whatever. I just like went to the Barnes and noble, uh, and bought a book on, it was like C plus plus programming and tried to like do something with it. And like, I like failed miserably. And then when I got in college and I was like, got a class and they like show you had to, or they had the environment set up and I was like, yeah, open the editor and then start and like typed this in and run it. And I was like, oh wow, this is actually really like, so much easier than I thought. Faris: Yeah. Justin: but that, that barrier of entry is a, is a real thing that I think as experience programmers, it's really easy to forget how hard it was to start as a beginner because there's so much stuff. Faris: It's not only easy to forget. it's, uh, it it's, we like it. It's like, we, we love that shit. Like you talk to programmers, they can tell, like they can spend 10 hours talking to you about how they rice, their, uh, environment and all these things. And, and nothing's wrong with that. Like you gain productivity, but that, that should be like the default. Um, you, you shouldn't like, you shouldn't have to, uh, you know, spend your teenage years rising your VIM environment so that you, you know, when, when you you're in, you're in the, uh, industry, you're, you're productive. Right. Uh, because you, you can't, you can't go back and like rice, your environment after you're an industry. Like you, uh, it takes a lot of time unless like you have, you have the affinity to do that, I guess. Um, but yeah, we're used to it and we like it and, uh, it's like Stockholm syndrome. A lot of people get pissed at repl.it like professionals because of these things. Um, and you know, we're adding a lot of power, a lot of customizability on repl.it.. We can talk about that later, but, um, like the default is you, you, you go into the code editor and, and you hit the run button and, and, you know, you, it, like you require a PA a node package or, or Python or whatever it installs it for you when you hit run, like you don't have to go through the, uh, through the pain of, of, of interfacing with a package manager yet um, when a professional lands on repl.it, um, they're, they're used to doing that. Uh, they, uh, I've observed behavior where they jump to the shell rather than like, clicking that. Big green run button. Uh, they like NPM in it and like NPM install, whatever, and, and all that. Um, whereas, um, whereas you bring in someone who, who hasn't been conditioned to that world yet. And who's like, who would, uh, naturally be approaching things from first principles and their, their, their intuition is like to, to write a few lines of code or go learn how to write those few lines of code and hit that big green run button. It's like, it's an expected things thing for, for them. Um, and, and we do a lot of that. It's like subtle and you don't notice it. If you're new, if you're a seasoned programmer, it might get on your nerves a bit until like you are comfortable with, um, not having to do your, your usual rituals. yeah, there's all these subtle things. I, I remember reading this, uh, Reddit post, uh, is like, how can I, how can I make vs code, uh, save, like repl.it like there, uh, like, I don't know, maybe four years ago we got rid of the save. The idea of save. Like, you don't need to save, like, you know, just things are saved. Uh, you're just run. Uh, and, uh, and yeah, like kids that grew up on repl.it and like are, are going to industry are, are starting to learn like the professional way. Um, they, they just expect things to work and they don't. Andrew: Yeah. Uh, my, my fiance's sister is learning programming and a lot of her classes were done through repl.it and like, it was just, like you said, like she got straight into coding and like, didn't have to worry about environments. Didn't have to worry about like GCC, whatever, like those weird in incantations we all know. Uh, and then she was like, oh, I'm gonna like try to do this locally. Like, it's like the logical next steps, like, oh, this is gonna be easy. And then all of a sudden we're dealing with like, oh, C plus plus 2017, doesn't have the right blah, blah, blah, blah. And just like instantly just off the cliff of like, oh, well, programming's very, very hard, but, but repl.it Really changes that. [00:09:58] How does it work? Andrew: So, uh, how does it like actually work behind the scenes, cuz it's kind of this like magical experience of you open a tab in your browser. You go to repl.it, you choose a language and you start coding. But like none of that code is actually running in the browser. So what's really happening there? Faris: So, um, in the past, all that code used to run in the browser. I can talk about the history of repl.it a little bit before we get there in, um, uh, repl.it like is a, is a passion project for Amjad who is the CEO and founder of repl.it. Um, started in, in college as I mentioned. And, um, he just, uh, he saw that Google docs like is, is a thing. And he was like, wait, you can like write prose in your, in your browser. You can like do word processing and that's like pretty complex. And. uh, he thought like, what's the, uh, what does it take to run code? And, uh, and so, yeah, I, I remember he, his prototype was like pretty shitty, uh, is like a text box and a submit button. And, and it basically is a JavaScript eval. Right. And I think it was just before, like JS bin or whatever got released and all these like tools and like everyone was like trying to like, caught onto that idea almost at the same time. It happens a lot. It's very common with, with tools. Um, and yeah, everything was running in the browser. Um, first language Amjad added, I think after, uh, after JavaScript was scheme. And, um, I think he used a project called BWE scheme, which is like scheme in JavaScript. Um, he also did write his own lisp uh as well as, as, as one would. Yeah. yeah. Um, yeah. And then, uh, and then he continued to add languages until eventually, like, he was like, okay, Python, uh, what is it like he added, you know, these esoteric languages that transpiled to JavaScript like brain fuck and like a bunch of these. Um, and yeah, he, he got to Python and it was like, what now? And, uh, he came across this, uh, project by the LLVM, uh, people called mscriptem, which compiled, uh, C to, as ASM JS, um, the precursor to WASM, uh, or the predecessor, sorry. Um, and, uh, Yeah, everything was running in the browser for like four years, I think between like 2011 and early 2015. And, uh, late, maybe 2014 containers were like all the noise, like Docker. People are like, yeah, it's like, this is amazing, like sandboxing and reproducibility and like, you know, the deterministic environments and all that. And Amjad is like, wait, this is great. Like I can, and it's like really lightweights, like C groups. It's not like a VM where it uses a lot of resources and basically create emulates everything. Uh, it's a very thin layer. And so he put that in, um, in a server and the code actually, like I looked at it not long ago. The initial code is like very simple, like, a web socket connection. You have, uh, you know, something that spawns a Docker container and then just gives you a session to that. And, and there's like a simple JSON API to like run, run things and stuff. Uh, it was pretty cool. Uh, and, uh, maybe six, seven months later, um, uh, Kubernetes got released. So it was pretty cool. Like he built Kubernetes before Kubernetes. Uh and, uh, and yeah. So what happens today is basically the same thing. You, uh, open a project, we call it a repl, uh, you know, people from the lisp community and all and, and stuff like that. Like, they don't like the fact that we call it re cuz it's like a project. It can be anything. Um, so yeah, you do that. It, we, you grab a token, uh, from our main servers or like, um, you know, CRUD app, and then you take that token and, and hit the, our evaluation API with the, with the web socket. And, um, uh, there's a lot of internal stuff that goes on with like distributed systems and all that, to know where you should go and, and, you know, if it's a multiplayer session, then you're like attaching to an existing container. and, We have a protocol, uh, between the, client and the, the container. and basically it has all these like primitives for, for talking to the back end. We have like, Uh, services or like file system service. We have a service for like, uh, you know, auto complete and language intelligence and all that. We have a service for, uh, the debugger and, and running and, you know, and packaging. All of this runs as like pid one. So like it's the init process that talks to the client, which is pretty cool. Uh, and yeah, that's it, I guess. Uh, there's uh, uh, the containers shared the protocol is like multiplier first. So, you know, there has like support for loop back and, and, uh, and broadcast and things like that. Um, yeah, I think that's like, that's like the core of. Justin: We, uh, we did an episode recently with, uh, a guy that I attended recurse center with his name's Paul Butler and Paul, and one of his friends, uh, started a, a startup called drifting in space where they're trying to build like a part of like what this infrastructure is. Is that like, oh, I want a long lived session, a persistent, long lived session on an instance that, you know, I can sort of come back to. Um, but it it's, it's interesting to hear about like how much of this infrastructure that you really, you know, had to sort of build up from scratch. Um, so that's, that's kind of a, yeah, it's fun, fun insight. Faris: maybe if like we were start today, we would use Kubernetes, but maybe, but there's all these like, guarantees that we have. And, um, like to make sure we don't have like duplicate containers and, and like routing and all these things. I, I dunno if I think gitpod uses, um, uh, Kubernetes. Justin: There's an interesting parallel here. So talking about like custom infrastructure versus Kubernetes, that I feel like aligns really nicely with the conversation of, of repl.it that, that idea that programming could be simpler, you know, that it could be simpler to approach, uh, and the sort of like scaled complexity that we see, especially in infrastructure, in the infrastructure tooling world. So like Kubernetes aims to make a lot of problems, a lot simpler and give you some vast capabilities that like would be pretty hard to build yourself manually, but at the same time it's hard. It is, it is rough. It's a lot of complexity that your entire organization has to eat, and there's a lot of concepts to understand and everything, and, and it's not, you know, it's not the repl.it Of the infrastructure world, right? It, it, it, it takes a lot of, a lot of context and, and a lot of time just sitting with it to really understand the edge cases and how to configure it and all that other stuff. Faris: Yeah. It might not like have the order of magnitude it, like I, or approach to repl.It of like abstracting things away. But it, it is in a way, like an improvement over, you know, running your own, um, VM orchestration infrastructure Justin: mm. Yeah. Faris: yeah. Or your own like scaling or, or sandboxing and, and reproducibility, uh, type things. Justin: I almost think of like serverless as being closer to like repl.it, when you're thinking about the infrastructure world, it's like, I just wanna focus on the thing that I'm doing and, and not all this like auxiliary stuff that like has to happen to Faris: But, but serverless also has like all these caveats, Justin: Oh, sure. Absolutely. Faris: no state, like you don't don't use memory, like things like that. [00:19:01] How does it compare to competitors? Andrew: Yeah. So you just mentioned, uh, gitpod so how does repl.it, uh, compare to services like gitpod and code spaces? Like what differentiates it? Faris: We always like to say this, but like putting an IDE in the, in the browser is, um, have you heard of like the term horseless carriage? Yeah. So like, it's, uh, you're basically taking the same thing and you're putting in the browser. Now there's like a lot of advantages to that, you know, portability and you know, so maybe reproducability is better in, um, in all these things. you know, There's presets environments, I guess. And, um, you know, if, if you, uh, if you have a nice, like internal dev tools team, they'll write your, uh, I don't know, what is it like a container config file for you for code spaces and all that. And, um, I, I think they're, they're focused on the status quo and, we are trying to leap frog. Um, so we're like every time we add something, it took us like a couple years before we were like, okay, let's add a file tree. Uh, it seems silly, but we're like, you know, that there must be something better for like project organization. And now there there's like these. I forgot what I saw it, but there's this editor in like, uh, programming language, um, that basically tells you there's no files, you know, it's just a, this massive program document, you know, you organize it whatever way you want. Justin: Kind of makes me think of there's this language called unison unison, web, where it's like, the code is actually all stored in a database and you just have a little scratch pad that you can just like run some code. And then you, like, when you quote unquote commit it, it like stores it in the database. And like you can query for all your code or whatever. It's like an Faris: That's pretty cool. Justin: abstraction of, of how to model. Faris: yeah. There's, uh, I guess dark lang Justin: Mm. Faris: might be in the same, uh, field and I guess we're, uh, if I had to put us somewhere on some spectrum of like known tools, I guess we're somewhere between like code spaces and dark Lang. uh, as a, as a programming environment. Um, so yeah, I think we are trying to leapfrog, but we're also backwards compatible, um, and like dark Lang, you have to use dark Lang, but on repl.it, like you can use anything. I'd say that's the main thing. Um, and in terms of like programming, but there's this aspect of, of repl.it and that's why we really believe in it more than anything is like the network. And, uh, we have a pretty vibrant community. Um, it's, it's really fun to see kids like, you know, chatting about code and, and sharing their, like, You know, small programs and, and you can go to like, you know, their profiles, uh, you know, when they're like, uh, 10 or whatever, like they're building like tic tact to, and like coin flip and blah, blah, blah. And then, you know, uh, you know, five years later or, or two, three years later, actually like it's, it's, uh, it's amazing how fast these kids progress. And I, I think partly because of repl.it, uh, um, it, yeah, you'd see them like writing all these like complex programs, writing their own programming language and yeah, it, it gets pretty intense pretty quickly. Um, and they sh they share this code with each other. It's like, uh, it's like open source, but, but, uh, but fun and... not fun. I guess the better word is like, uh, welcoming and. You know, it's, it's really cool. I don't know how to describe it. Uh it's yeah. It's like, uh, hanging out with your friends online truly. Andrew: yeah. It's, it's truly amazing how low the barrier to entry has gotten in programming. Cuz like back when my dad was learning to program, it was like, well, you could go buy a book. You could learn in isolation. You could maybe run some programs. Maybe they work. And like, even back when I was learning, like code academy had just come out and like the barrier to learning, like web development had started to lower, but like me trying to go and learn like C plus plus or any of these like integrated on my machine things still super hard. And it's really cool to see that repl.it is, is changing that. Faris: Yeah. I, I also, uh, like got started again in, in code academy. Uh, and I did work there for a summer as a QA. It was like my first like tech job, I guess I worked in it, but I don't think that counts. Um, and, uh, yeah, it is, uh, it's pretty cool using code academy, uh, is amazing. And I think, I I'd like if, if, uh, if repl.it had a guided experience sort of like good academy in, like, we we're really doing it with our onboarding and it's getting better soon. Uh, but I I'd like this to be like a trove of like, uh, lessons and, and courses. but, but more focused on like building rather than learning what a variable is. Justin: Yeah, there's this really magical moment when you're starting out and you have a tool like repl.it or something where you're enabled, but you still don't really know the shape of the environment that you're in, where the limitations of what you can build. Uh, just. You don't know. Right. You don't know, you don't know about all the difficulties. You don't have all these like frictions to like, oh no, that's way too much work. I'm not even gonna try to do that. You know, it's just like, you're like, oh, I wonder how you'd build a programming language. And you just like, start like hacking with stuff. And like, you know, the, the sort of ideas of like what constitutes a good programming language and, and sort of like what their shape should be and, or, or even the, the sort of like long term level effort to like build a mature programming language is not something that's in your head. So you have this like almost permission and freedom to explore topics that are like, you know, one from the outside might consider way out of your area of depth. But the only thing that it really takes is, you know, a really fast feedback loop and, and, and to feel like you have the control over your environment, that you kind of understand what you're your, your sandbox, if you will, like the area that, which you're playing. And, and I think that that promotes a lot of really powerful learning where, you know, as we were saying earlier, a lot of, a lot of other people who were like getting into programming just got frustrated. Like I can't, this thing's not in my path. I don't know what that means. Like, I don't know how to, I don't know how to move from here. You know? Like, Faris: yeah, It's, uh, it's, it's, it's overwhelming is the right word for, um, for getting into like a, a non sandbox. [00:26:45] Community Creativity Andrew: Let's focus more now, uh, on like what, what things people have built using repl.it. So do you have any, like go to projects that you can point to where you're like, wow, that it's crazy. That they were able to build that in repl.it. Faris: Um, yeah, there's, there's a few, um, off the top of my head, like, um, let's see. So here's the, here's the catch? Um, most, most things built on repl.it, um, are in this like hobbyist state. So they're not quite like a production thing. Um, but they're like really fun tools. I think the, the most fun things are like, you know, just like games, kids build, um, multiplayer games. There's this fun thing that people do with repl.it, uh, uh, you know, there's Minecraft speed running and the way you get better at that is by, uh, you know, how Minecraft is like, uh, is gen generated. Uh, and there's like a random seed. And so they use repl.it to find seeds that make their, uh, speed run faster. And so like, they run through all these like iterations to find like a glitch in the seeding algorithm so that they can speed run pretty quick. And I think there's like a couple pretty in like impressive results that came out of that. Um, Justin: That's wild. Faris: yeah, it's pretty cool. Uh, what else? Uh, the first, the first thing that blew my mind was. This chat app that, uh, a kid built, I think he was, um, 12 or 13. Um, and what he built was a terminal app. Uh, repl.It now has hosting like you open a port and you get a URL and all that instantly, uh, before that, uh, we only supported terminal apps and, you know, you could share them and publish them to the community still and all that. Uh, we actually had a website that's called like repl run and then like your repl and it's like a, just a webpage with a terminal. Uh, it's pretty cool. Um, so yeah, it was a, it is a terminal chat app and the way it worked is that you, uh, take their, take their terminal app. You fork it and you run it on your own. And then. Everyone talks to everyone, uh, through like a request like a P2P request thing through the repls. And, um, yeah, it is is, is a crazy like chat app. Uh, uh, that's just like a, in the terminal distributed. Pretty cool. Justin: That's awesome or 13. That's huge. That's like, Faris: yeah. That, that same kid built like a load balancer on repl.it Justin: wow. Faris: yeah, Andrew: definitely goes back to the theme that we've explored on this podcast. A lot of constraints, breed, creativity, and these, these kids being constrained by repls found ways to be very, very creative. Faris: yeah, yeah. A hundred percent. Um, yeah. Repls are, are cool. A lot of kids just like build languages in in repls. Yeah, there's all these like utility apps that like, especially like kids are, we're try, we're starting to see PE like kids selling these things. Um, so yeah, like a kid would build like a recipe thing and they'd go like sell it to the, to the neighborhood, like, uh, moms and, and, and dads, like, I don't know. Andrew: So like, like how, how would that work or the, the, is it like a terminal app or like, like with, has a GUI. Faris: now we support like hosting. And so like there's all these like front end, uh, uh, web apps basically, uh, when we first released web apps, um, actually that kid works with us right now. When we first, uh, released, uh, web apps, one of the kids wrote a, like a scheduling web app for his school. And, you know, he was trying to sell it to different districts and different schools. And Justin: Wow. Faris: that, was pretty cool. Yeah. His name's Shane he's. He is pretty cool. Uh, he, he's a support engineer with us right now, Andrew: yeah, that that's awesome. Repl.it Is not only taking down the barriers to entry of like getting into coding, but it's also kind of taking the barrier of entry to like distributing your code and like hosting it too. That's that's super cool. Faris: Yeah. Um, I'm working on a project to bring that more, um, closer to what production feels like without, you know, having to go through the pain of like understanding production and. And things like that. So it's just like, basically like publish this version under, you know, I pick this URL, uh, because now it's, uh, like your development environment is the same as your production environment. Um, a lot of kids get around that by like forking the repl and, you know, that's their production and then they, they basically just push things to it. Um, but if it's like in one place it's, it's better. I think hosting is like, uh, is a, like a distribution channel, um, uh, is, is one of many distribution channels that we eventually want to build. Like, you should be able to build mobile apps on your phone, uh, build them and then like push them to the app store, things like that. Justin: Yeah, that would be pretty cool. Uh, I mean, this all sounds like pretty hard problems to solve. I mean, you know, on the, on the face of it, it's like, oh yeah, this wouldn't be too bad, but then you, you know, we dig into any of these things. [00:33:22] Future of Development Justin: So, uh, we've talked a lot about like, uh, sort of the history of how repl.it has developed and, and sort of, you know, your role in it. And just some of the, the projects that have been contained there so far. But when you think about the future of repl.it, what, what do you think that entails in as much as you can share? And is there anything that you're all working on currently that, that you can share that you're excited about? Faris: uh, yeah, as I mentioned that, that like, uh, collaborative coding and like a, a better even like async collaborative experience, I think is a, is a big thing for me personally, something that I'm, I'm passionate about and like, you know, I, I enjoy working on, um, but like even even more. Uh, I think transformative is, um, is the fact that we are looking to build an economy within repl.it. Um, uh, you know, I talked about the network and you know, how people interface with each other's repls and all that. Um, I think, yeah, having an economy built within that network would be really cool. So like, I can provide like a, a payment API for you and, and, and charge you right. As, as me as a, as a repl, like creator, and I can create an environment that is, um, you know, that, that enables like yeah. Multiplayer features. Right. And like, you could use that as an API and then like, you know, uh, we have this, uh, uh, , it's not, maybe it's maybe crypto, maybe not like a thing where you, you know, exchange value based on that. Like you may by API call or, or whatever. We have a bunch of prototypes in that space that, that are really exciting. Um, yeah. And, uh, and just like one click, everything, uh, one click payment, one click ads, if you want, like, it should be like one click databases, uh, should be really easy to do a lot of those, I think. Yeah. Just like naturally like SDKs, um, and, um, in maybe not lock lock in fashion too. Uh, I think it's totally possible to do it. Um, yeah, so we're, we're really like, we're gonna release some stuff freely soon. Um, uh, I don't know how much I'm supposed to share, but, uh, . I, I hope it's okay. Uh, yeah, but we, like, I think the first thing that you might see come out from repl.it is um, in the economy aspect is making a post, like sort of like a freelance thing, you know, I want X and Y can someone build it for me and, you know, they just give you a repl and, and you pay them. Uh, I think that that's, that's like very easy to build. Uh, there's like an operation, no cost for it. That's very high, but it'll be worth it. Justin: Yeah. And this goes back to something you said earlier, you said that y'all, y'all tend to look at what people are building and, and what they're hacking on and, and use that to inform sort of the features that you build. And you said that there's a kid who built a schedule who like wanted to sell it to different school districts. So it was like, there's an obvious signal here that, like, this is probably, uh, an interesting area of exploration. I'm sure there's a lot of enterprising folks out there who will be able to take, uh, pretty good advantage of this. I think the interesting challenge will be in the sort of like moderation space because you kind of get into the, like the, the roadblocks sort of area of like development and, and kids and monetization and, and, you know, that can get Faris: Yeah. It gets pretty Justin: hard to navigate. But from the broader ecosystem perspective, like being able to spit up a monetized app really quickly with like very little infrastructure is, you know, that's. That, that could be huge. That could really change the game for a lot of people. So excited to see what y'all do. Faris: yeah, I, I hope it works out. Uh, I think, yeah, it's, it's gonna change the software world if it works. Andrew: Okay. So last question, before we move on to tool tips, this is a kind of like a general question that we ask everybody that comes on the show. Like, what is the future of the space they work in, but what do you think the future of development is? I think I kind of know your answer now. Uh, but do you think like tools like repl.it will become the standard and a local host will die? Well, that's it for this week's free portion of the episode. If you want to listen to the full conversation with Faris. Head over to our patreon and become a member. For just $5 a month, you can help support the show and help us get to our goal of delivering content weekly. Thanks for listening. ​