The Changelog: Software Development, Open Source - Flynn Updates (Interview)

Episode Date: December 20, 2013

Andrew talks with Jonathan Rudenberg and Jeff Lindsay about their hard work and updates on Flynn, their open source PaaS.

welcome back everyone this is the changelog and i'm your host adams takoviak we're a member supportive blog podcast and weekly email covering what's fresh and what's new in open source check out the blog at the changelog.com our past past shows at 5by5.tv slash changelog, and subscribe to the Changelog Weekly. It's our weekly email covering everything that hits our open source radar. You don't want to miss it. We ship it on Saturdays, and you can subscribe at thechangelog.com slash weekly. This show is hosted by myself, Adam Stachowiak, as well as our awesome, awesome host, Andrew Thorpe. And this is episode 115, and today's show is sponsored by our good friends at Digital Ocean and our good friends at TopTow.

And we'll tell you a bit more about TopTow here in just a bit, but we're huge fans of TopTow. They connect startups, businesses, and organizations and people like you to a growing network of elite engineers all around the world. So if you've been wanting to freelance, call up TopTel, tell them we sent you. They're going to love who you are. But huge fans of DigitalOcean. We want you to get hosted on DigitalOcean today. We're on DigitalOcean. We absolutely love it. Blazing fast SSD cloud servers. You can easily spin up a new droplet with root access within 55 seconds

with your choice of size, RAM, region, OS. It's really, really easy. They've got a simple and easy-to-use dashboard. And for our fans across the pond, they recently announced their brand-new data center in Amsterdam. Super awesome stuff going on there. And they're going to continue to invest heavily into their infrastructure worldwide as more and more data centers are added throughout the world.

But we want you to try DigitalOcean today for free using our promo code CHANGELOGSENTME. That is CHANGELOGSENTME, which will get you a $10 hosting credit or basically two months free if you're on their base plan. But head to digitalocean.com to get started. And now, on to the show. We're joined today by Jonathan Rundberg and Jeff Lindsay to talk about the Flynn Project. Welcome to the show, guys. It's great to be here.

Yeah. So we had you guys on the show previously, but things have changed quite a bit at Flynn. So for the listeners that maybe either haven't heard the show, shame on them, or are catching up, why don't you guys give us an introduction of who you are and what you guys are doing? Well, I'm mostly a freelance engineer, but I've been working on this project with Jonathan, uh, to basically build this sort of ideal platform or, uh, basically the system that I've, I've always wanted to work with at every company I've worked at. Um, and so that's sort of me and my involvement. Yeah. Um, I'm, uh, mainly working on Flynn right now. I also, uh, work on the tent protocol and a few other open source projects.

Basically, all of my work is open source at this point. And so Flynn tries to be essentially a platform for developers to deploy their work on. And that can be anything from a web application to stuff that anything runs on Linux, essentially. It runs in containers. So the idea is that it's the product that an operations team can provide to an engineering team. And I mean, that could be someone as, you know, just a single developer who maybe does a little bit of ops on the side, runs out of EPS all the way up to like a startup that, um, has a full like service oriented architecture with lots of different components

exactly just speaking purely about like the the growth of the project so we had you guys on the show it was about three or four months ago now where do you guys remember like where you were at then and kind of what's changed since august yeah so uh at that point we were basically at the end of our fundraising period and we um uh we hadn't really written much code yet. We'd done a lot of the architecture work, but there was very little code available, and we were still doing in the planning stages, essentially. So we're much further along now. We have prototype components of just about everything that we plan to build, And we have some releases coming up this month and next month, essentially like completing the work that was funded in August.

Right. So the Layer 0 release is set to, is that coming out this month or is the Layer 1 going to come out? Yeah. So Flynn has two layers. Layer 0 will be released probably around the end of December. And then Layer 1 is scheduled around the end of December. And then layer one is scheduled for the end of January. Gotcha. So what, what would you say in the last couple months has been like the, I don't know, not the biggest change, but just in, you know, you guys kind of went out there and you had a pretty successful fundraising experience. And so since then, now you guys obviously have a lot more, you know, news behind you. People know about Flynn and you guys are actually starting to put some code out there for people to kind of start to consume.

So since then, in that time, what would you say has been the biggest struggle with maintaining this open source project, but also now you have demands and expectations because people have funded this thing? Well, I mean, putting code out there, I'm actually more personally, I am not so afraid of putting code out. I do a lot of open source projects, although the attention that this one has gotten is kind of interesting because I'll build a component that's sort of like, it's a placeholder component. I'm talking about shelf, which is one, it's just like a little HTTP file service. It's like the simplest thing. And we put it up as a placeholder because it kind of serves the job that we need right now. And all of a sudden people are submitting patches to it and there isn't even really a read me. Um, and so just the number of eyeballs, um, has made the experience a lot more enjoyable.

Um, yeah. Um, I also publish a lot of open source code. Um, so I'm used to people looking at stuff and I'm not really afraid to push things, but, uh, just the amount of attention is really interesting. There's a, it's, it's really nice to push something and see that people are already using it the next day, just kind of experimenting, figuring out what's going on. And unfortunately our, our docs are rather lacking in the readme department right now, but people are struggle to get things set up. So the one thing that I do want to point out before I get too far is that we actually have a demo of Flynn available that you can essentially boot up all the prototype

components in Vagrant and just test out. You can do like a sample, get push, deploy, scale up, scale down, look at the logs, et cetera, very much like Heroku. And so we plan to obviously develop all these components further and we are doing that currently. But if you're interested in just like experimenting with like the very alpha Flynn, it's available right now. We released that last month. So in like a kind of just layman's speak, what would you guys say is like the roadmap for like a production-ready version of Flynn that people can actually start to use? Yeah. You know.

So at the end of this month, we're releasing Layer 0 Flynn. Layer 0 is essentially everything you need to schedule and run containers on a cluster of nodes. So this is much closer to something like Mesos than it is to Heroku. It's very low level. And then next month, we are going to release Layer 1 in a working configuration that is close to Heroku. It'll probably be very bare bones compared to Heroku, but this will be suitable for deploying probably just internal apps that you have at your company

or if you're just a hobbyist, experimentation, et cetera. And then over the coming months, we will be working on improving the stability and fixing bugs and adding features, et cetera, to bring it to something hopefully that in mid-2014 will be usable to serve production traffic. Yeah, that's crazy. So you guys are, I mean, it's so young. And I remember when we brought you on the show last time,

I believe the only actual repository on GitHub you had was FlynnSpec. Is that right? That's absolutely true. There wasn't a single line of code written. Yeah. And now I just look at the repositories you guys have and it's crazy. The amount of growth has been shocking and it's only been a few months. So, I mean, how has the other work that you guys have been doing with Tent and with other things struggled or has it at all? So I still do work on Tent. I've been focusing on Flynn because I really, really needed to get Tent to the next level because we're actually going to be deploying our Tent server on Flynn and releasing that

in a nice packaged version. I don't know whether you've ever tried to deploy like an open source Rails app or anything like that. Deploying complex services currently, even open source ones, tends to be very complicated. So I'm excited for what Flynn can do for deploying open source stuff as well so that you could, for instance, run a very robust tent server on your own server using Flynn and the management and operations of it would be very simple compared to what you'd be used to deploying just about any other open source application.

So that's why I've been focusing on Flynn lately. And but tent is definitely bouncing around in the back of my mind. To be honest, it's been great to have a break from like thinking about tent full time and just kind of switching to something else because you, you just think of different things when you uh when you have something bouncing around in the back of your mind instead of thinking about it full time right what have you guys thought about doing this in go i mean go is still pretty young and and you know it's a i don't know it's it's

it's definitely a powerful language and it's got a lot of acceptance but have you guys kind of ran into any gotchas that have hit you at this point? The only thing is dependency management. That's the only thing I'm aware of that is a problem at all. And there's some solutions in the works. GoDep looks like the best option right now. But that's basically the only thing that's bitten us, I think. Jeff? Yeah, no, I've really loved it.

I feel like I've also become a better programmer using it. That's awesome. Yeah, you have to learn a little bit more about lower level stuff and stuff that you might have taken for granted if you work more in higher level stacks, so that's cool. Absolutely. So let's talk a little bit, I mean, the last time we had you on the show, it was Flynn spec, that's all there was. You guys now have a working demo and you're close to your layer zero release. Let's talk a little bit about, now that we have the ability, what Flynn is and kind of take a deep dive into Flynn and who it might work for and what it would look like for a production application. So what would you say is the elevator pitch for Flynn?

Okay. So, and I said this before, Flynn is the product that operations provides to engineering. So what that means is that instead of doing like one-off deployments of applications where you have to write chef or puppet scripts just to deploy a single application, you as an operations team, you are managing this platform that engineering can then use to deploy and manage the applications themselves. It's very self-serve. And you also get the consistency of configuration. So if you have a bunch of applications that all use MySQL, they're all using MySQL in the exact same way. We have these things called service appliances, which Jeff can explain, that make this very simple.

Yeah. So while service appliances are kind of neat, every component of Flynn is basically almost a standalone appliance. And we think of these as basically software as a service in a box. So each of the components has its own API and is actually made to work in a cluster. So, for example, when we talk about things like a MySQL appliance or any other kind of database appliance that has kind of a master-slave clustering setup, the ideal is to basically spin up a couple of these appliances, and they would, via our service discovery infrastructure, be able to self-organize into a master-slave and provide a lot of the administrative functions that you would normally do by logging into the machine via an API so that you could then write your own code or systems to then manage them in a consistent way as opposed to using tools that assume that you SSH into hosts.

Because very much Flynn is about abstracting away the host. It kind of pushes you to think more about services and service-oriented architecture. And for the most part, hosts are pretty kind of homogenous. I mean, you still, the operators still have to know, you know, which hosts, they still actually can manage the hosts, but in terms of what goes where, that's mostly taken care of by the systems that you would be building with Flynn. Yeah. The only decision you're going to really make is just like, you're going to pick a few classes of nodes that you have deployed.

So maybe you have like high memory nodes for doing caching and high disk nodes for doing like databases, that kind of stuff. But other than that, you're going to be just managing everything through Flynn. And you don't think about the host, you just think about what you're running on your cluster. So you say, you know, I need a highly available MySQL deployment, I need a highly available Postgres deployment. And then this application needs to connect to that. And it's just going to keep on working. Even if one of those nodes goes down, you can scale up, scale down without really thinking about where things are running unless you really need to.

And you, yeah. No, keep going. Go ahead. Go ahead. Okay. You always have the ability to kind of drop to a lower level and say, hey, I actually know exactly where this job should be running. For instance, for data locality, if you're doing lots of MapReduce jobs, you'd want the jobs to run near the data on the same nodes possibly.

Right. You talked a little bit about how deploying open source applications is a pain. How does Flynn, you know, I don't know, when you're talking about deploying, let's say you're talking about deploying a Rails application and you have all the, you know, the kind of de facto standards for deployment. Whether it's Heroku or you're rolling your own thing with Capistrano or, you know, those things. How does Flynn help to alleviate the deployment issue that you kind of talked about? So in the case of a Rails app, most Rails apps these days are going to work on Heroku. So it's going to be very similar.

Sounds like we had a... Yeah. Sorry. Completely forgot to silence my phone. I'm sorry. No, that's okay. We'll edit this out. Probably not, but it'll be funny in the show. Okay. Sil's balanced. So the question was, how does Flynn help to alleviate the deployment issue that you mentioned before?

Yes. So Rails applications, typically on Heroku, you're just going to get pushed them. You're going to say, I need a Postgres add-on. I need Memcache, whatever. That's great, except when you don't want to run your application on Heroku, you're just going to get pushed them. You're going to say, I need a Postgres add-on. I need Memcache, whatever. That's great, except when you don't want to run your application on Heroku. When you want to run it on your own infrastructure, perhaps you have a VPN that you need to have this service behind. Or you have your own hardware. Or, you know, there's a variety of situations where you'd have this.

You could deploy those applications in exactly the same way because Flin is compatible with Heroku applications. But perhaps you have an application that consists of multiple services that are more complex to orchestrate. Flynn gives us all the hooks we need to deploy that. And the exact way we're going to do that has not been completely sorted out yet. But the idea is that you have a manifest that describes exactly what services are required and you can, um, run those, uh, on Flynn very easily and deploy those.

And running, running and deploying those backing services, uh, is pretty much the exact same, uh, as you would deploy the application. So it actually, you know, the difference, I think when you, when you're working with a Rails application, at first you're worried about deploying the Rails application, but you also have to worry about deploying the database server, and then later you have to worry about deploying a caching server, and then later you have to worry about deploying some sort of background worker type system. And very often, a couple people talk about this quite a bit. A lot of web application development starts simple, but eventually turns into this very service-oriented type of architecture where you have a number of, it's more complex than just this sort of three-tier architecture. And so, especially larger organizations, either as you get bigger, existing large organizations have a fairly service-oriented architecture, and they get to a certain size where even their application is broken down into smaller pieces or services. And so that's the sort of thing that Flynn really helps support because you can manage and deploy all these things pretty much the same way you would a Heroku application. So it's interesting because when you talk about service-oriented architecture,

it sounds to me like Flynn is a great solution for just the more complex. I mean, Heroku is great, right? Heroku has enabled a lot of things that are, you know, it makes a pretty, what could potentially be a difficult task, as long as you're running a relatively generic setup, pretty simple. And it sounds like Flynn kind of is going to take that to the next level, which is going to make the more complicated tasks relatively simple. I mean, would that be accurate to say? That's absolutely accurate. And I should point out that we wouldn't have Flynn without Turoku, because because Heroku kind of demonstrated a lot of the concepts that were just kind of taking a bit further in Flynn.

So Flynn absolutely, it starts with like something like a Heroku app. And probably a lot of people will just use that functionality of just like Git pushing your applications. It's very natural right now to do that because we're familiar with Heroku, et cetera. But then you can kind of dive in and say, I actually have a more complicated application. For instance, I could build a tool that does continuous integration and then continuous deployment. So I run my tests using Flynn. And then the output of those tests is a build artifact that literally is a container image. And then I am going to perhaps manually or automatically

deploy that image, the exact same image that came out of the tests, and deploy that to production on Flint. Yeah, so just to be clear, this is built on top of Docker, right? And Docker is one of the things that has kind of helped to enable some of these, what's the best way to put it?

These, like, box development or you're deploying the – I think Docker's really pioneered this model of, like, a high-level container. That's what container. I couldn't think of the word. Thank you. Because containers have been around for a long time or things like containers, but Docker really introduced sort of this, um, you know, slightly more VM like model for containers, um, that I think, uh, is actually a bunch of people are kind of pushing that to be an open standard.

So there will be Docker competitors and stuff. Um, but, uh, but yeah, right now it's all based on a lot of the amazing work that the, the Docker guys have done. So one of the questions I kind of have for you is like Docker in the last couple of months has changed a bunch too. I mean, and it's continued to evolve and grow and get and become a full-time project and essentially like take over the company. And so have you guys had any issues with that or is it, has it pretty much been in lockstep with what you guys are doing? They're taking Docker in various directions, directions adding features mostly focusing on like single host use cases whereas we're very much focused on the the multi-host use cases of containers um so a lot of their uh their features are not currently being used by flint yeah but i

i think um i mean they've gotten a lot of attention. And so they've also gotten a lot of contributors and a lot of people pulling it in a lot of different directions, which is great to kind of get all those ideas out. But I think, you know, at the core, there's a really simple idea in there. Um, so, but the other thing is a lot of people are still trying to catch up to thinking in terms of containers. Um, and so that's, um, you know, kind of, uh, it's, it's great for us because we're based on this concept of, of containers. Um, and, uh, but it's sort of like, we're still, Docker's doing an amazing job. It's basically prepping everybody for Flynn.

Yeah. Yeah. Yeah. Um, it's, it's great to see people thinking about the different ways that you can deploy applications because for a long time it was, was you know standardized on things like capistrano and fabric which are great and they've they've gotten us a really long way but there's there's lots of new and exciting ideas that we can have about how to do this yeah how much of your like day is spent not day but how much of your you know time is spent like defending hey how does this compare to heroku what is this as compared to heroku or docker you know how does this relate like how much do you have to clarify

that when you're explaining flint to somebody um not much i mean to be honest it's uh mostly comparing to perhaps other open source passes um cloud foundry yeah cloud foundry um mostly uh most people get it like it it's at a very simple level, it's Heroku that you can run on your own hardware. We don't compete with Heroku, really. The people that want to use Heroku aren't people that want to use Flynn because you need to have some operations support

in order to run Flynn. You don't need that with Heroku. That's the feature that Heroku provides is operations. Yeah. And it's kind of nice when you're, we were talking about earlier, if you, when you start with a simple application, you can run it on Heroku. You get to a certain point where people kind of need to need slightly more than Heroku. And really their only choice after that is to redo everything. Or sometimes

they'll be afraid to use Heroku because they know that at some point they'll outgrow it and have to go basically back to, you know, host-oriented architecture cloud, you know, like EC2 or something. And all that infrastructure that Heroku provided is gone. And so hopefully Flynn will kind of fill in the gap between EC2, um, and a Heroku, but it kind of, in a way I, I, it's sort of like, um, in terms of, uh, in terms of operations, but in, in a way Flynn is more than Heroku. I almost think of it like Heroku plus plus in terms of the software, because it gives you so much more power, um, because it's basically Heroku given to you inside out.

You have control over how it works. It's really interesting that you say that because I wonder if the future for Flynn, and maybe you can kind of speak to this, is you want to encourage people, hey, start out with Heroku. It's painless. It doesn't take much configuration. You don't have to know much about operations at all. And then as you start to grow and your demands grow, Flynn kind of steps in as the next evolution for the apps that are on Heroku.

How would you feel about that kind of like a flow for people? I mean that's great. You can also start out with Flynn. Because a lot of people I I think, are a little hesitant. If it's something really simple application, pretty standard, Heroku a lot of the time just does the trick. But there are so many times when people don't think that Heroku will do the trick. And I think in those cases, people might want to start with Flynn. Um, and, and it, you know, even if your app is really simple, um, hopefully Flynn is, um, so extremely simple to set up that the, uh, you know, kind of overall, you know, it is a kind of a complex system, um, but overall it should be

easy to set up and that it's worth it because you're going to grow with it as opposed to trying to, and it sort of has a path for you, right? As you, as your system gets larger, as opposed to you kind of doing your own thing and having to relearn everything and try out a lot of different approaches to, to scaling your application in terms of both complexity as well as handling load and stuff. So I think with Flynn, we're kind of baking in all these kinds of best practices and ideas from, you know, Google and Twitter and basically every, every, everything we've learned from, from all that stuff and putting it into a package for you. Right. So you talk about getting started with Flynn. And one of the things you mentioned was you're behind on the docs thing.

So how will somebody get started with Flynn? I mean, just plain and simple. I mean, if you're using Amazon Web Services, you're probably just going to boot an AMI that contains Flynn. In the case of running it on your own hardware, you're going to start with a base OS. And currently you would just install Docker and then probably run one or more Docker containers, which would then talk to Docker and get everything else set up. It'll be very simple. But as far as like education on Flynn, and so obviously like getting up and running with Flynn is going to be pretty painless, but as far as like how to take Flynn to the next level and configuration and all that stuff, where will that exist for you guys? I think so that, that spec that we had last time we talked is actually

kind of want to turn that into like a guide. Um, that's sort of, you know, we would have our normal sort of getting started and here's the basic usage stuff. But, um, I kind of envision a guide that is sort of, um, uh, a really great sort of user's manual, both for operators as well as engineers, um, to sort of understand how the system works and how they can make it work for them. Um, and so that's, you know, I guess kind of, kind of the vision. Yeah. Um, obviously we'll have documentation on the website, um, at varying levels of detail, um, depending on what you're interested in. Right. So you guys were hosted a meetup, uh, I guess last month. Is that right? Yes. What was it? Yeah. We had a meetup in San Francisco.

So what was that experience like for you guys? It was really interesting. I really wasn't expecting so many people to show up. We had, I think, 40 or 50 people show up. And then to have most of them, pretty much all of them, get it. Yeah, a lot of people got it. It was good. You guys were in the Twilio offices.

Yeah, Twilio. I'd been there a couple of times since they moved, but they have a pretty nice setup. Yeah, really nice space for meetups. It's really cool to see. I mean, I'm just kind of trying to not harp on it, but it's cool to me that we had you guys on the show at such a young part in Flynn's future. And obviously, we were talking about fundraising at that point and trying to help you guys get off the ground. And it's really cool to me to see a project like this just take off. And from the ground floor, it's two guys sitting – working on this thing. Well, two guys in the open source community working on this thing.

And just to see it take off like you guys have been able to do this. So, you know, I just want to get into that. What has the experience given, you know, in terms of Flynn and, and the future of Flynn and what has the experience been like for other people who, you know, potentially could get into something that would take off like this? What, what, what should they expect? What should be something to look out for? And, you know, the thing that I hear a lot of is demands from the community. And you start off with just this idea, it's just you and your friend. And, you know, you start off and it becomes a bigger deal because people have invested money. And I'm curious to know, and I want to kind of tie this into the fundraising goals you have for the project in the coming year. But I'm curious to know, was there any added pressure once that happened? Maybe the approach of our fundraising kind of helped. You know, we talked about how we

focus more on companies that were willing to work with us and understood how it worked. I think if we took a more traditional, you know, Kickstarter-esque, anybody can put in a little bit of money, we would get a lot more people kind of feeling more entitled that we have to listen to their opinion. I mean we listen to everything and just go into the IRC channel. But most of – I think – I haven't felt a whole lot of pressure. It's mostly been personal pressure of like, cause there is this very clear vision for what it is and it seems like everybody understands it. So it's more about kind of meeting for me, my own, my own expectations. Yeah. It's a, it's a very positive thing to have a bunch of people that, that get it contribute

money. And so far there has been very little, if any, uh, pressure from sponsors about specific features, um, or demands. Uh, it's been really great. We've been, we've been very, uh, upfront about what we're building and people seem to get that and are on board with it. Um, and now it's just a matter of delivering that, which we're, we're rapidly working towards. I think we're a little lucky just in the choice of the project because it's very much a problem that it's infrastructure, right? And everybody has to deal with these problems. And we are using a lot of existing landmarks, things like Heroku and existing systems like Mesos or just schedulers in general, and a lot of these kinds of best practices. So I think building a system that's based on those

means a lot of people already understand. We're not really introducing completely new crazy stuff. We're really just aggregating a lot of existing best ideas into a single system. And that makes it easier for people to understand and get behind. What's the team look like now? Is it still just the two of you or would you? There's actually like three of us. Yeah. We also have Daniel who also works on the tent protocol and he's doing kind of community management type stuff, but he doesn't work full time on it. Gotcha. So you have a fundraising

goal for 2014. Why don't you talk about that a little bit? Yeah. So our goal for, uh, 2014 is $350,000. Um, and that is, uh, entirely going towards development that will pay for myself and Jeff. And we're also looking to bring on, bring on at least one more person to work on the project full-time. And we're looking for monthly contributions from companies that are interested in working with us to deploy Flynn into production, essentially. And so we've just kind of started on this goal, and we expect to hear more about to hear more, uh, about it in January. So you have your layer zero release coming up pretty soon. Layer one release is coming a few months after that. And then the fundraising goal will go towards what layer two or, you know, what would be the next step? Well, I, uh, we've taken a really kind of broad

strokes, um, approach to this. So, you know, just because we have these releases out doesn't mean that they're, you know, really ready for production use. And so it's really about filling in a lot of the gaps and building out some of the extra stuff because there's just tons of nice-to-haves. And the system is so open-ended that it's sort of like, at some point, you can't really tell the difference between this system and your system, right? Because, in a way, you start building your system in this. And so there's a lot of things that we can get into

that are kind of designed for, you know, the architecture of Flynn, things like, I mean, log aggregation is coming up pretty soon, but things like metrics and stuff like that. Auto-scaling. A lot of people are interested in scaling up, down on various public and private clouds. And so, yeah, the funding we raised initially covered six months of development,

which will end at the end of january and culminate in our uh kind of initial like i guess beta release um and then uh what we're raising money for is to cover development in 2014 because we want to fix bugs and make new features and make cool new stuff that runs on flynn that just makes everyone's lives easier so after that what are you guys thinking about for as far as, you know, I don't know if sustainability is the right word, but after the, you know, 2014, I guess my question is would you guys consider just continuing fundraising efforts or is there some plan to monetize Flynn in the future? Currently no plans to monetize.

Things change rapidly. I expect that late in 2014, Flynn will be quite stable and that there will be a strong community around it developing add-ons, etc. So I don't foresee needing to fundraise past 2014, but who knows? It's interesting. I mean, it's definitely a different take on how to do open source. Well. I mean, it's a different, definitely a different take on how to, how to do open source. Well, I mean, sort of a different, right. It's fundraising and developing open source, but it's a really neat way to, you know, how to, how to do that. So you've had, you have some competitors that, you know, you mentioned Cloud Foundry and things, how many of these kind of got this started around the same time as you? And how would you, you know,

talk about the experience of kind of entering a new space? You know, I say new in quotes because this isn't new, but it's a potentially new solution. That's different way of thinking about it. So what has it been like with your competitors or is there any, any relationship there to, to kind of speak of? Um, so Cloud Foundry has been around for a while. A few of the others have been around for a while. Uh, there's at least, uh, one other one that kind of is started around the time that we did, uh, as far as publicly goes, but they've been working on it for a while named Deus. And they actually share some components that Jeff built for Flynn. Um, so it's, everyone is kind of working towards what they, their model of what they think life should be like, deploying applications, et cetera.

And so it's very healthy I think. Yeah. We're going to pause the show for just a minute and give a shout-out to our awesome sponsor, TopTow. They've been sponsoring the show for a little bit and we've had a chance to tell you about some really awesome stuff they're doing. I've been working with Brendan, their co-founder and C cto and i mentioned that you know i wasn't quite sure what to expect from them and i was but i was excited about what they're doing they're helping developers who want to freelance with some really awesome companies find ways to do that and it's their mission these guys are the real deal they're engineers themselves from top to

bottom they're not technical recruiters trying to pimp developers. So if that's what you think, then you've got them completely pegged wrong. They're a network of elite engineers all around the world who work with some really awesome clients. And for those of you out there who are freelancing or would like to freelance, you've got to check out TopTie. You can work on special projects with companies like Airbnb, Artsy, IDEO, and many others. You can work remotely.

You can go to Andrew's favorite place, which is on a beach, or anywhere in the world. No office is required. And to get started, head to toptop.com slash developer, click Join the Best, and because they want to work with only the best senior engineers out there, they've got a well-thought-out four-stage screening process that begins with a personal phone call via Skype

to kind of get to know who you are and introduce you to who they are and what their mission is and see if you're a fit. And from end to end, the screening process includes an English-speaking test, a timed algorithm test, technical interviews with core TopTile engineers, as well as a test project. And once you've made it past the screening process, the sky is the limit. And if you think you have what it takes, head to top.com slash developer right now to get started, tell them the change law sent you and enjoy. Now back to the show. So this is a unique way. This is the first show we've done where we want to bring a previous guest back on the show to talk about progress. You guys have

been tremendously successful up to this point with this project and and we hope that it continues to go into the future um is there anything specifically that you guys want to talk about on the show and i want to kind of get give you guys a part a moment before we wrap the thing up um anything you specifically wanted to hit on about Flynn for the listeners to hear? I don't think so. I think we, we covered, um, everything. Um, I mean, not everything, but, uh, everything that we want to talk about here. Yeah.

So let me ask you one. I think we might not have, uh, uh, mentioned this actually. How can, let's say somebody is listening and they want to contribute as a company or, you know, whether it's recurring or not, how can they contribute to Flynn? Um, so you just go to flynn.io. Um, there is a sponsorship form, which you can fill out, um, and drop your credit card number into, and, uh, that'll set up a, uh, one time or a recurring sponsorship. You can also email us. The email address is on the site and we're happy to work out alternative arrangements. Yeah. Actually, if anybody's interested in using it as well, it would be really great if you got in touch with us because, you know, at a certain point we're going to probably want

to start working with some companies, whether they're sponsors or not, just in trying to get them set up with Flynn and getting their feedback and stuff. So we typically ask the same questions at the end of every show, but being that we've already asked you guys, we will ask you kind of just one question now to see where it's changed. And the question is, what can the open source community do to contribute to Flynn? And obviously start using Flynn is a big point now, but just in general, what would be something you guys would like to see

the open source community kind of pitch in with? Support, I guess, and not just financial, but I mean's something that you, you think that you would, uh, love to see your, whatever organization you're working at, um, have, uh, you know, start, you know, maybe talking about it internally about, you know, this great possibility, um, and, uh, and we'll see what happens, but I think it would be great to just have, um, you know, even more, uh, love and support because it kind of makes, uh, our job easier in terms of fundraising and continuing to exist in that regard. Yeah. And, um, in the meantime, um, absolutely play with our prototype components. Uh, we have it all packaged up into a VM and hang out in the IRC. Um, feel

free to, you know, raise issues on any of the repos, send us email. If you're, you have any questions at all. What's, what's the IRC channel that people can come to? It's a Flynn on Freenode. And how will people email you? Is the, you said the email address is on the website, which is at Flynn, F L Y N N.io. That's correct. Awesome. Well, I guess we had nothing else, nothing left to talk about because Flynn is something that is ever growing and, uh, it's still, you know, you launched or you launched your, I don't, you said probably the best way to say it is you're like super alpha version of the product. Um, it's, it's going to be cool again, you know, and I. And I said this last time, but I mean it again this time.

You've got a big couple of months ahead of you, and I'm excited to see what the next few months bring. And hopefully you guys don't burn out. I think one of the things that I read on your blog was everybody has been working on Flynn. Everybody, meaning the three of you have been working on Flynn much more than you expected, which is a good and a bad thing, right? And the good is that you've, I mean, it's crazy. It's only been a few months and it's already to this point, but the fear is burnout. So, you know, hopefully the next few months will be pivotal and you guys will be very successful with it,

but not get burnt out on it so you can continue to deliver great things with Flynn. Thanks. We hope so too. Yeah. Well, once again, I wanted to say thanks to Jonathan and Jeff for joining us on today's show. I also wanted to give another shout out to our sponsors, DigitalOcean and TopTal for supporting the show. You can head to digitalocean.com to set up your cloud server today and make sure you use our promo code changelogsentme. That's changelogsentme to get a $10 hosting credit. If you want to freelance with companies like Airbnb, Artsy, or IDEO, you can head to toptal.com slash developer and click join the best to see if you have what it takes to join TopTal's

network of elite engineers. Again, the URL is toptal.com slash developer. And that's it for this week. Thanks again to Jonathan and Jeff for joining us. And also thanks to the listeners for tuning in and for your support. If you haven't yet, subscribe to the ChangeLog Weekly. It's our weekly email where we share everything that hits our open source radar. You can subscribe at thechangelog.com slash weekly.

So for now, guys, let's say goodbye. Bye. We'll see you next time.