Screaming in the Cloud - Merging Vision, Community, and Technology With Anil Dash

Episode Date: April 25, 2024

This episode features Anil Dash, VP of Developer Experience at Fastly, who returns to the podcast to share the integration of Glitch within Fastly post-acquisition. Anil shares how Glitch has continued flourishing under Fastly's umbrella, highlighting both platforms’ seamless acquisition and mutual growth. Anil shares the technical and cultural collaboration that has allowed Glitch to maintain its identity and mission while contributing to Fastly's broader goals. The episode highlights the power of community in tech, the importance of maintaining core values in mergers and acquisitions, and the advancements both Glitch and Fastly are making in the developer space.Show Highlights: (00:00) - Introduction. (01:59) - Glitch's role within Fastly's infrastructure and services.(02:16) - Comparison of AWS and Glitch’s approach to community building.(05:40) - Anil’s ongoing enthusiasm for Glitch beyond typical post-acquisition experiences.(08:53) - Fastly’s unique capabilities and impact on the internet. (14:35) - Fastly’s technical infrastructure and its performance advantages.(20:42) - WebAssembly’s implementation and significance at Fastly.(23:59) - Comparison of Glitch and Fastly’s developer engagement and pricing models(25:18) - Ethical responsibilities and building a healthy tech ecosystem are important.(27:50) - Importance of creating lasting and sustainable technologies.(30:24) - Anil discusses Fastly's work culture and its influence on employee innovation and engagement.(34:26) - Anil discusses Glitch's thriving post-acquisition integration into Fastly.(38:26) - The critical role of Fastly's infrastructure in supporting major open-source platforms and decentralized networks.(39:26) - Closing remarks and where to find more about Anil’s work.About Anil: Anil Dash is the vice president of developer experience at Fastly, where he leads the team behind Glitch, the friendly developer community where coders have collaborated to create and share millions of web apps. Anil advises startups and nonprofits, including Medium and the Lower East Side Girls Club. An accomplished writer and artist, Dash has contributed to Wired and The Atlantic and collaborated with Lin-Manuel Miranda on one of the most popular Spotify playlists in 2018.Links referenced: Anil’s Personal Website: https://www.anildash.com/Glitch’s website: https://preview.glitch.com/Sponsor:Panoptica Academy: https://panoptica.app/lastweekinaws

everything between you and your user should be controlled in software. Everything. There shouldn't be any part that's sort of this legacy janky hardware that you're wrestling with or whatever. Welcome to Screaming in the Cloud. I'm Corey Quinn, and I'm joined today by returning guest Anil Dash, now the VP of Developer Experience at Fastly. Anil, it's been a while. How are you? I'm great. Thank you so much for having me back. This episode's been sponsored by our friends at Panoptica, part of Cisco. This is one of those real rarities where it's a security product that you can get started with for free, but also scale to enterprise grade.

Take a look. In fact, if you sign up for an enterprise account, they'll even throw you one of the limited, heavily discounted AWS skill builder licenses they got because believe it or not, unlike so many companies out there, they do understand AWS. To learn more, please visit panoptica.app slash lastweekinaws.

That's panoptica.app slash lastweekinaws. That's panoptica.app slash lastweekinaws. When last we spoke, if memory serves, because time is a weird thing, especially post-pandemic, you were the CEO and founder of Glitch, which got acquired by Fastly. And you were super excited about it, and as you're generally supposed to be, and you assured me at the time it was absolutely not an acquihire, which is exactly what you say

when it's an acquihire. But two years in, the glitch.com still has its usual logo up in the corner. It doesn't have a by Fastly splattered underneath. It's its own going concern. It's disclosed down below the fold that, yeah, you're part of Fastly, but it's standing on its own two corporate feet and still is a viable brand that does awesome stuff. So I have to with the first principles and for people that aren't familiar with those or haven't heard, Glitch is a community where you can build a full stack web app in your browser in a minute. It's a great experience. It's a great community.

People are super supportive and helpful of each other. Imagine AWS's PartyRock if it wasn't terrified of its own users and what they might build and also understood community. You know what? Everybody, including the PartyRock team who builds tools that make it easier for people to create, they are our allies and our friends.

So we're always happy about that. Fair enough. But, you know, we're very proud. We're very proud of what Glitch has been. And, you know, that it's attracted millions of developers, earned their attention, had them build tens of millions of apps, I think is something we take deadly serious. You know, it's a very, although it's a fun community and very friendly and brightly colored, the responsibility of that is really, really serious. You know, it's a very, although it's a fun community and very friendly and brightly colored, the responsibility of that is really, really serious. And so we came into

Fastly almost two years ago, you know, with a really strong kind of North Star about like that responsibility of helping people create the open internet together is really important. It's something that's animated, you know, the last 20 plus years of my career. And I think many people on the team feel that way too. And so, you know, we came into Fastly saying, how do we take that? And this is the thing everybody says when they put out the press release, how do we combine these strengths of these two, you know, these two different teams? But I think it's been really real. So, you know, today, just to go out myself personally, I lead at Fastly, both our developer experience team and our edge computing platform, which we call Fastly Compute. And so this is basically our portfolio of the services and tools

that we make directly for developers. And Glitch sits really, in a lot of ways, at the center of that. It's kind of the warm, beating heart of, like, where's the place you connect? Where's the place you show off what you made? And where do we also show our values about how much we support the open web and people being able to create together? And that's a real anchor point, you know,

sitting alongside the work we do on the standard things you do for developer experience. Here's the tools and the SDKs and the APIs and all that kind of nice stuff that's there. And then even sort of the buy-in, especially as Fastly's had a lot of new leaders come in in the time that we've been here and they've all kind of come on board and been like, oh, this is really cool. We have something sort of special here. So I think that's been really nice and it is the ideal of how these things work. And there's always a learning curve. There's always an adjustment period, but I think it's been really incredible.

And the biggest testament I can say to it is we're in the midst now of one of the biggest upgrades in Glitch's history. We've got a site up at preview.glitch.com that we are building in public in collaboration with the community around modernizing Glitch because it's seven years old. There's technical debt and legacy stuff, but also reflecting just the resurgence in interest there's been in the indie web, the open web, the Fediverse, all these different areas of the human-created web. Glitch has, I think, been at the heart of a lot of those communities. I think even Apple put out the Vision Pro last year and the WebXR community doing open source VRXR stuff. There's hundreds of thousands of people creating that stuff on Glitch.

And so they've been, same thing, a shot in the arm with saying, okay, Apple's on board now too. And there's pointing this stuff. So us being able to say, okay, we're going to support those kinds of platforms with an open platform itself that you can create this stuff in and get started in a few minutes. That that's been very inspiring to me, very motivating. Even all these years later, I've been working on glitch for more than seven years and it still is as fun and it's exciting to me. And also I got a day job that I get to work on that and work with that team. And so many other people across Fastly, hundreds of other people at the company have embraced it.

People use it in their everyday work across the company. They use it in teaching people how to use the platform. I think it's the ideal. It's the ideal of how you get these things to come together. You have still that spark in your eye when you talk about it, as opposed to most post-acquisition founders who have this dead look. I'm like, oh, I'm here for the long haul, they say, right before they leave a suspiciously round number of months post-acquisition when they hit a cliff or whatnot. But you're still building. You're still doing exciting things. And let's be clear, your entire career has been about community and egalitarian access to things and decentralization. You remain on the board of the EFF. You're an advisor to

Medium. You have done a lot of stuff that demonstrates this is not just a convenient talking point this quarter. If so, it's been a convenient talking point for decades for you now, and I think it's more salient now than ever before. Yeah, I appreciate that. That's very kind. I mean, it is something that's been a North Star. I mean, I think about even one of the salient examples. I was, you know, on the board of Stack Overflow pretty much from its inception as an independent company

until its exit a couple of years ago. And, you know, it's been, obviously there's been a lot of tumult and then sort of the rise of AI world. Advancements in copying and pasting via computer instead of by keyboard. Exactly. But it had a very, you know,

and it was, it had flaws in terms of like you know people could be hostile and all this kind of things not dismissing any of that but it had a very egalitarian bent which was democratizing access to the knowledge of being able to create technology and and in any other industry that would have been this sort of exclusive priesthood and instead to have this driver be everybody should be able to see this stuff for free and we should have it under open licenses. Now that means you're vulnerable to exploitation. But, you know, it was a really, it was something I still find really inspiring and actually Stack Overflow shares co-founders with Glitch. And, you know, I think that's something where that's an old fashioned idea that's, I think, come back around

the idea of building community together that way and sharing publishing in the open that way like there's a lot of collaboration that i think is it goes in cycles now i see having been in this industry long enough things come in and out of fashion centralization versus decentralization open versus proprietary like these things are these waves that come in and out from the shore and it feels like we're we're riding the wave towards the good place again in a lot of ways where people are seeing that you can give people agency control authority over both their individual creative processes as coders as developers as product designers as as architects whatever their role is and then also the platforms will respond in kind where they're sort of saying okay let's give you more more power because we see

that's going to be table stakes to to being in the next wave. You've been there two years now. So I'm curious to know if you figured out what exactly is it Fastly does exactly, because they've done a very good job at flying beneath the radar. And then a few years ago, and I promise I'm not being crappy when I say this, before your time, they took a roughly 40 minute outage one morning Pacific time. And what was amazing was how many sites broke, including, and this surprised me, Amazon.com for a few minutes. That's when I did some digging, which it turns out you can do,

because who someone's CDN is, is real easy to figure out if you know how to read a DNS result. They use CloudFront, which is because they own it, and Fastly, presumably, because they want something that might understand

what a user experience looks like when it's good. Not to besmirch CloudFront, but anyone who's used it more than 10 seconds in knows exactly what I'm talking about. So I'm curious to know, have you figured out what Fastly does? And if so, could you please share with the rest of the class? I will say it took me about two years to figure it out. And this is having known most of the founders before the company was even started. And my impression, I think like a lot of folks was, oh, that's at CDN. They're nice people. I heard it's really good. And in 2024, no one wants to be a CDN. They want to be something that's perceived to have higher up the stack value. Yeah. And also, I think it was also,

and it's that one you use if you've made one of the biggest sites in the history of the internet, right? Like that's who uses it, not the everyday developer. And, you know, I've come to realize it is much broader than that. And I think the first thing I didn't get until I was talking really to Tyler McMullen, who's our co-founder and like deeply technically he's been CTO. He's had a bunch of roles. I mean, like founders all stupid, but you know, he's kind of like one of those brilliant architect types. And the thing he kind of revealed to me along with Arthur Bergman, we always wanted to build this sort of instant global brain, this instant global computer that you could program and

you could do whatever you wanted to all around the world instantly. And the approach they took was to build a killer app for that kind of platform so that they would get to do it. And the analogy I came up with sort of talking to them was a little bit like, you have to build the iPod before you can do the iPhone, you know, and then someday the iPod becomes the music app on the iPhone.

It's still a great music app, but it's just an app. And I think that's what CDN had been for Fastly in its first many years, was like the first killer app of this instant global thing. But what I didn't get was, so instant global programmable, like these three things. Instant, first of all, and this I genuinely didn't understand

until being inside the company. Instant here is like real. It's literally the blink of an eye, literally 150 milliseconds. Like that is the expectation. When you say something is instant and fast, that's what it means. And that's unique.

And the reason it's unique is because they built the whole stack down to the metal. And I was like, who does that? Why would you do that? And they're like, this is why we built a company. That sounds like the before picture, some sort of cloud migration narrative that, of course, bears little, if any, resemblance to reality. It sounds like this, like, don't do this. It's like, this is a cautionary tale. Except it works, and it works well, this, don't do this. It's like, this is the cautionary tale, except it works and it works well.

Exactly. And that was it. And I think they were like, look, if you tried to get funded to build a platform now, you're going to like build everything all the way down to the metal. People like, you are crazy. We're not going to let you do that. And, but they, I think they were in the right period of time

and the right technical skills and all, you know, all the sort of like the stars aligned that they got to do this. And so they built the stack all the way down to the metal and were religiously uncompromising, like relentlessly strict about, we never compromise on it being instant and we never compromise on it being global. And by global, it's this like, how do you make it regionless, right? How do you make it, like you can, if you need, if you want to know where you're at, you can

know that, but you don't have to know. Because I think about, and again, nothing but respect for the, you know, well, if you want to know where you're at, you can know that, but you don't have to know because I think about, you know, again, nothing but respect for the, you know, the, the teams that run these platforms, but like any day that I have to hear about us East one, I'm having a bad day. Exactly. Things always live there, but you don't have to think about the fact that they're there. Right. And so, and so this idea that you could give people tools that were instant, that were global, and then the sort of philosophical bent of everything between you and your user should be controlled in software. Everything.

There shouldn't be any part that's sort of this legacy janky hardware that you're wrestling with or whatever. And they held the line on those sort of three principles for like through thick and thin, like 13 years in, and they're sort of doing this thing. And I came to understand this

because I started late last year running the compute platform. And I came to understand this because I started late last year running the compute platform. And I was like, okay, there's a lot of compute platforms out there. There's a lot of nice tools. Like it is what it is. And I looked at it more closely. I'm like, no, you guys actually made

something completely different. It's completely different, like architecturally, technically, all these things. And the thing that got it to me that made it really clear, I'm going to nerd out. I feel like this is an audience where I can do that. I read a white paper on bimodal multicast and I'm going to forget the guy's name who did it,

but it came out in 1999. And we're 25 years later, and what bimodal multicast lets you do is fast, reliable, predictable. So it basically was one of those classic choose two kind of things. And it was, can you deliver messages reliably around the world at that 150 milliseconds instant speed? And this was a white paper on how you could do it. And nobody ever built it before. And so you go back and I'm like, you madman, you made this thing. You made it happen.

You did the thing. And they're like, yeah. And I was like, you didn't tell anybody. They're like, well, you know, we don't want to brag. And I was like, and I went and I found this blog post from an intern at Fastly in 2014.

This is a true story. 10 years ago, an intern wrote a blog post, like we implemented bimodal multicast. It's pretty cool. It lets you instantly send configuration or updates or data around the global network of points of presence that we have. And as a result, we're able to reliably do anything in software at that speed.

The first killer app for that is updating the CDN. So if you want to purge content or you want to change your config, it happens at the blink of an eye, but there's no reason it's tied to that application. It can be used for anything. And you have this this like and that's like the last mention like we let one intern write a blog post 10 years ago and they're like i hope they got it and then just never revisit it and so i come in you know like look at the compute platform and i'm like i think this is the killer app for the whole damn thing because you know we have a firewall everybody okay everybody has got a laugh too right but only this laugh is getting those signals at the speed of blink of an eye and can preemptively address the security concern instead of you having to

work after the fact they add some regular expressions foundational things you build in early on architecturally shape what can happen down the road you can do like wizards from the future if you get it right but often it's just a weird confluence of things lead to that outcome like few people can see two decades ahead in this industry. And I would say this with them sitting in the room. They're just brilliant engineers, the founders, and pathologically humble. And there's this kind of geek thing where you want it to be like, well, if it's sufficiently technologically advanced, everybody will just be drawn to it.

And I was like, no, they'll be like, I don't get it. I don't know. You need to have somebody like me who's moderately a dumbass to be able to be like, I don't get it. I don't know. Like, you need to have somebody like me who's like moderately a dumbass to be able to be like, let me translate this into like what civilians can understand or somebody who hasn't like read the white paper or whatever. And, and, and the one that jumped out to me is like, okay, everybody's got a data store. We have a key value store, like a KV store. Yeah, I got one.

Everybody's got one. Okay. But could you populate it in the blink of an eye globally, automatically? And is that different? And all of a sudden, everybody who's tried that out has been like, oh my God, I got something I want to use this for. You just immediately think of what's the thing I'd make.

I mean, you could always change it, have the API return success and then count on eventual consistency on the backend, but it turns out that doesn't get you very far. Why did we build the whole stack? We could have just been lying on the returns. That's what Gen AI does, to my understanding. He just makes it up, but it sounds

so confident when it does it. It's glorious. We call it a data hallucination. Can we charge for that? But I mean, jokes aside, I think that was the thing for me. I can go down the list and it's like every category where like, take the good thing, make it instant. Take the good thing, make it instant. what a joy that is

to go and be like, okay have a superpower and it's like you know so the classic thing i think that that people who have been customers are using the platform have been like this um if you want to make if something is slow and inefficient sprinkle some fastly on it right that's what you do and so thinking of it the same way for us where we're like are there categories where i've been frustrated like i've been that person sitting at the whiteboard in a product meeting or sitting on a zoom call like this part i've circled on our architecture is slow as hell and it's killing us and it's making users unhappy and it's causing all these other problems and if you're like i can just make that part fast like that's a real superpower so like that's that's the like not very secret

mission right it's like get people and the great thing about it for me too is it matches this ethos about that open web piece that we're talking about, which is I'm like, I'm not trying to get anybody to lift and shift and go through some mass migration, misery, blah, blah, blah, blah, blah. Like this is why we work with all the big platforms. Even the trillion dollar companies,

when they want something that's instant, they use Fastly. And the reason why is it works with doing. Like that's, you know, it's a valuable thing and they don't have to leave their platform or migrate off or do whatever. They can integrate it into their workflow, into their stack, the things they're doing.

That to me is like a really, it's a fun place to be because I don't have to go, I don't have to go like pointing, oh, their stuff sucks to anybody. I'm like, we have a thing that is unique in the world because nobody else was crazy enough to build it. And then I get to sprinkle that unique, you know, hot sauce or, or, or, or, you know, special seasoning or whatever you want to call it onto all these things. And then tell everybody else, if you want to integrate that

into what you're doing, go for it and I'll help you do it. That's my job. Allow me to do a bit of a fact check on this, because what you've done recently is fascinating to me as well. You've hired Leslie Carr as a director of engineering, and she has a systems and network background, better, like me, if I were actually good at those things historically. the promises one time too many. And she's excited about what Fastly is doing. And getting someone with that background excited about something is a heavier lift than, frankly, getting rando VCs hyped up about something. This takes, you have to do the work. And she's excited and frankly, terrific hire. I look forward to her getting spun up and seeing what wonderful things she's able to build over there. But yeah. I mean, as we heard this, she's just joined. So I'm just so excited to get the chance to collaborate with somebody of that caliber. There's so many people like that here. You're lucky to have her.

Let's be very clear on that. And the only way to get that is to earn it. Yeah. And I think that's the thing I would say is like, I definitely, like I'm not at that technical caliber, but I definitely am one of those people that has a little bit of cynicism or skepticism about the industry. Everybody says they could do the thing and everybody says they're a pioneer. But I look at, one of the things is like the proof is in the pudding. Why would,

like I said, the trillion dollar tech companies that have infinite resources and incredible engineering talent, incredible technical skills, all that stuff, infrastructure, why would they choose to work with this smaller player that we are? And it's like, it has to be something there that is sort of special and unique. And, you know, I was going to say,

it's not just our charm, but I think it's nice that they like to do business with us too. Like, I think that's, that is a, a not small part of it, but it really is like, you have to have something that is sort of special and unique out there.

and then somebody like Leslie, you know, she's working on the technical work around this compute platform. And, and I think the exemplar of that is the pioneering work on WebAssembly that so many people at Fastly have led that community and really been, you know, helping to find it and shape it. And, you know, WebAssembly is one of those things where, like, the geeks who geek out about it are, like, kind of all in. And everybody else is like, okay, let me know what is real. Or, you know, I've heard you say this is exciting for a long time.

And, you know, I bring that same skepticism to it. And what I see is sort of two things i think the first is there's a class of bugs endemic to software over the last 50 years 75 years around memory safety and similar concerns that we have known mathematically are addressable but practically have not felt we could actually use them in the real world like address them in the real world. And I think a lot of the people around the entire open web assembly community were like, once and for all, we're going to kind of go in and we're going to sort of solve this.

I don't want to come back here and solve this a second time. Let's get it right this time. Yeah, exactly. Well, let's be honest, I don't want to come back here and solve this an eighth time because there have been attempts at this sort of thing all throughout history. It's like the energy of like parents being like, kids don't make me turn this car around. You know, like they're very like, I'm going to finally solve this thing.

And that's been amazing to watch. And then I think there's this other part, which is I never want to push tech for tech's sake. Right. Like, is the format good? I'm sure it is. But for me, I'm like, one of the greatest things that I can do with working with this team is give the WebAssembly world a killer app, which is what I think this set of instant services is,

which is like, we can't give you unfettered access to the raw power of an instant global network without putting boundaries around it because you'd knock off the internet offline, as it turns out. It turns out that when you apply jerks to anything, they can turn it terrible.

It's like, this is part of the problem too with Gen AI as an easy example. It's great for basically mediocre content that spits out about a bunch of things. But as soon as you encounter add jerks to it, it's oh great. You've built a harassment cannon and that becomes a problem. And so,

so, so we think deeply about like the responsibility to the ecosystem. And a big part of that is putting the right guardrails around it, putting the right boundaries around it, putting the right, you know, containment around it. And, and, And WebAssembly gives us that. So we're able to say, like, we will, in exchange for using this environment that has these, you know, safeties built in, that has these precautions built in, WebAssembly will free you from having to think about which programming language you're using.

It'll free you from all these other concerns. And in our case, we can be like, and we can give you access to the special sauce that we've been doing internally of this instant engine that before we had to hold really close to the vest because if we gave people unfettered access, they could cause all kinds of problems. And now we can say like, it feels unfettered,

but it's bound by the security model of, you know, WebAssembly in that case was born in the browser where you had to have this like really, really smart security model and web assembly in that case was born in the browser where you had to have this really, really smart security model and sandboxing model. And so they're taking that revelatory moment that so many of us had in HTML, gosh, I can make this once and it'll run in everybody's browser, billions of people around the world, it's incredible, and bring that same thing to the whole rest of the stack. And so I think that's just an, just a, it's an exciting thing to get a front row seat.

Few things are better for your career and your company than achieving more expertise in the cloud. Security improves, compensation goes up, employee retention skyrockets.

Panoptica, a cloud security platform from Cisco, has created an academy of free courses just for you. Head on over to academy.panoptica.app to get

started. One thing that I don't fully understand, and it's why I admit I was wrong, I was convinced this was an acqui-hire regardless of anything anyone could tell me, based upon one clear differentiation. One of the first things I do when I see a company's website is I ignore most of what it says and I make a beeline for the pricing page. Because unlike anything else that it says elsewhere, this is really going to determine, is it for me? If there's not like a call us for pricing enterprise tier, then it's designed for small individual use case. And that is Glitch at the moment. There's a free tier and there's a pro tier for $8 a month.

There is no call here for special pricing option. Conversely, all of the tiers that Fastly have have the same pricing breakdown at the end, which is request a pricing quote, which tells me I'm speaking to an enterprise sales person and I will probably not be going to space today as part of the challenge. But also, you're not going to charge me $40 a month if I have to go through that process to wind up talking to someone. It's going to be fairly expensive. You're doing developer experience, but that feels like a top-down, not developer approach on the Fastly side. And I viewed these two things as being inherently incompatible, where one would have to shift. Clearly, I'm wrong.

What am I missing? You're not wrong. We need to do a better job of connecting those things together. And I think that's going to happen. But I think the fact that we've had both winning in their lane has meant we can take our time and get that right. And we didn't have to sort of rush to like pace the like, you know, sign up here, over here. And we want to have a really designed and intentional experience. And I think that's been, you know, what's there.

But I look at, I mean, to be straight with you, I think the number one thing people have asked me is saying, well, you do developer experience with Fastly. It's like, when can you get free accounts? Like when are we going to get, you know, when are we going to get like a developer account for Fastly and try this stuff out? So like, I'm not unaware that that's a demand and it's a really good, healthy sign.

There's a lot of examples we can point at. The ecosystem has a lot of players that rush to provide a quote unquote free account that was not sustainable. And whether that was on the licensing around an open source product or a overly generous limit in a paid SaaS product or whatever it was,

it's what I talked to developers and it's towards the top of their list of stuff that has caused unexpected stress or churn for them is we got jerked around by this vendor in this way. And to me, that's like a core trust thing is, is we have to, you know, make sure everything we're doing is something that is sustainable, defensible in the long run. You screw me over that that's on you. If you screw me over a second time, then how it's my fault for in some level,

letting you be in a position to do that. Yeah on you. Yeah. And also just like I've been a developer a long time, you know, decades now. And I want to build stuff that is going to last and that is going to help other people make stuff that lasts. And so, you know, we look at that. I wouldn't open the floodgates to everybody coming in and trying out the entire fast experience until that entire developer experience is up to snuff and you know that is something that obviously that's what I do first thing in the morning when I wake up is I think about that and it's the last thing before I go home you know from work at the end of the day I think when we get there we're gonna let everybody know you know that's gonna be one of the things where we sort of say, we've got something. Because to me, the bar at least is Glitch.

The fact that you can come into Glitch, no account created, build a full stack app in your browser, not even having logged in, and share it with your friends and make something amazing, I'm really proud of that. It's that simple in a minute. We have millions of people who built their first ever or the first one they've done in years website or app on Glitch.

And I built a small thing on Glitch years ago. And it's awesome. I love the experience. Like so much of what I do starts off as those weird small projects. Like right now, my screenshot domain, where I automatically hit a key combination drag and it winds up there and copied into my clipboard. Great.

The domain I use for that is shitposting.picture. I feel like if I had submitted a contact us for access thing with that domain, it would have so immediately been spam binned that I never would have heard from anyone at that point. But that experience building that has led me to use similar technologies on that same platform

in more serious commercial endeavors that are actually revenue-bearing and not just I need a thing. Exactly. I think we all have a site where we kick the tires on the weekend project or we tinker with something or I want to try out this new API or whatever it is. And so I think we need to get to that level of simplicity. And also I want the whole ecosystem to be there. I think that's really it. How do we get everything that people want to do in not just the Fastly world, but like this whole cohort of developers now that are building, you know, all these new exciting areas to have what they need.

And, you know, and we'll get there, I think, sooner than later. That's obviously the sort of top priority for me. wish more i wish more vendors more platforms talked about their responsibility for predictability and continuity and trust for the developers and the companies that rely on them you know i've been through that pain where oh we don't have that tier anymore and you gotta switch this thing and now it's going to cost 100 times what it used to cost or 10 times what it used to cost or your limits are going to be way lower and um it's not just like a frustration. Like it's this like mandate that comes out of nowhere. We're like, oh, our roadmap is blown up

because now we got to go chase this other thing that was foisted on us by a vendor. It's clearly worked for Fastly this long. So I'm not suggesting you're wrong. I'm saying that it doesn't put itself on the line of the way I adopt technologies, but that's okay. I've accepted a long time ago that I am atypical in a number of ways. developer is going to go to and say like i got something i can build here um but i think glitch is the proof point that we do know how to do that and so i look at like how do we put those two bits of knowledge together and um and sort of catalyze the the evolution there um you know i think we'll

have more to say about it soon i i'm you know i'm mindful too of like i don't want to over promise like i want it to be the kind of thing where where I want the developer world to tell you when we got it right. Yeah, I think that's fair. You don't get a second bite at the apple. And right now what you're doing is saying no to a certain subset of people coming by and looking at it. And I think that's the right answer because when you suddenly start saying yes,

people will forgive the no answer, but they're not going to be nearly as sympathetic with, well, I tried it two years ago and it was just absolutely terrible, so let me try again and see if it hurts as much the second time. People don't do that. And that's it. I think we respect people's time, we respect their energy and attention,

and it's easy to just sort of slap a thing, oh, try this. Changing the webpage is not the hard part. Having the thing be up to our standards is the hard part and up to the standards of the developer community. But, you know,

I think we're headed that way. I'm excited about it. I think, I don't think it'd be too long before we'll have more specifics to sort of share there, but I'm, I'm mindful of like not getting ahead of, like I said,

I like to under promise over and over. And to be fair, you do have a, on the contact us request pricing page or the question, is this a business trial account or is it a developer account? Like, okay, there's at least that progress.

But people are still scared of basically every terrible experience they've had by filling out a form and suddenly they're getting email from half a creation trying to sell them something that they don't understand.

Yeah. Yeah, no, and I mean, that's one of the things too where like, it's been a joy for me to come into, I'd never worked with like a publicly traded company or company of this scale before to come in. And I'm our, um, our lead for our New York city office.

And we have, you know, a couple dozen people here. Great, great energy. But a lot of them are literally the people who follow up when someone fills out that form. So I get to hear the phone calls. I get to talk to them over lunch about like, oh, what's the emails that you sent? I get to see this sort of follow-up.

And they're generally like fairly young, like early career people. Sometimes this is their, one of their earlier jobs in the tech industry. And they're so thoughtful about it and earnest about it. And like the culture of like, well, we want to help these people. You know, like we want to be really helpful.

And we want to be like that part, honestly has been the kind of surprise, like one of the joys for me is like, that's not something that i'd had that kind of visibility into before and to see the the earnestness and sincerity and motivation and and you ask like for me i asked them like why'd you come here they're like the the values they're like the tech they're like tech is amazing but the people are really good and that's i guess the reputation that precedes them because a lot of them like i said this is their first job in tech or earlier in their career and they'll be like yeah i heard the people over here are good that's, I guess, the reputation that precedes them. Because a lot of them, like I said, this is their first job in tech or earlier in their career. And they'll be like, yeah, I heard the people over here are good. That's a profound thing. That's a rare.

It is. And you have the right people. You do interesting stuff. I find you in client environments all the time and people rave about you. So it's where I don't see you is in the hobbyist projects that I tend to start with for an awful lot of things. You're not going to be a dependency hooked in by some random GitHub project. That's right. And those folks are close to my heart. So, you know, you can imagine that's obviously something we're paying attention to. And you are the VP of developer experience.

I don't get the sense you were looking for a sinecure where, well, we don't have to worry about that because we're enterprise only. And that's the end of it. Like, I'm looking for someplace to rest and vest. You don't strike me as someone who holds still very well. No, no. I mean, I think that was it. It's like, there's, there's always a trepidation, you know, when a company gets acquired, like, do they mean it? Is it real? In our particular case, they went through, um, a CEO transition on the day when our deal was announced, um,

that we were going to become part of Fastly Selected. Like, uh, welcome aboard, by the way, we're going to have a new boss in about five minutes, you know, it was a really like kind of a little bit of a white knuckle experience. But we've had incredible leadership coming in. You know, the founders are all still here, which is unbelievable. And then to have like, you know, a new CEO come in with like a shot of the energy that you get from that and new product leadership, new marketing leadership, all those kinds of things. Like those are things that, you know, like I knew that I'd, like I said, I've been on the board of big companies before, but like seeing it from the inside, I thought was a really interesting perspective to see how much culture

can change and accelerate with just a handful of people sort of pushing on those things. And so, you know, that was, it started from, I mean, I can't even count the number of people I know who made something great, got acquired and watched their thing get slowly, you know, death by a thousand paper cuts into being something that they hated or resented. And most of them take their little nest egg that they got from that and go and try to build the product again. You know, like that is the most common scenario. And I'm like, you know, one, I'm too old for that. I'm not trying to build this again.

But also like Glitch is special to me and meaningful to me. The community is meaningful to me. There's people, like I've made friends and acquaintances and people that we've hired have come out. They can be like all those kinds of things. So I wasn't going to just be like, eh, we'll see how it goes and we'll cash out. And like if it crashes and burns burns it's somebody else's problem that was not ever the goal

you don't have ultimate control over that it's a publicly traded company it's like all these different things but I think everybody's realized the value of it and so we've gotten to do the ideal I think it's also a testament to the community there's no two ways about it people see the power and the force of like people who are actually engaged.

And like the cool thing is like glitches well past the like buzzy stage. Like we're not the like hot new thing. You know, it's seven years old and then people, oh, that's still around, huh? Like people, you know, I go to industry events. Are you still doing that? Like, yeah, yeah, yeah. It's, it's, you know, it's, it's, it's, it's real. You do have a chance to buck the tide

of when a company gets acquired by a larger one. Usually the response from its users and customers is, ah, crap. Because how many times has it ended well for those people versus how many times has it been disastrous? And like I said, the proof is in. You can go to preview.glitch.com. You can see the new site, Bugs and All, that's being built.

We're literally like, here's what we just did this week. Like, it feels like startup mode again. And we have support to do that. And I think that's a really special thing. The other thing I'd say that was a real early validator that we got things right was Fastly It All has been a big supporter of open source. And then right when we came in with the Glitch team, we were able to revitalize and relaunch a program we now call Fast Forward, which is, you know, supporting all the open source ecosystem, the open web, the good internet kind of ecosystem.

And, you know, Fastly has long been that platform that delivers, you know, the content for the open source ecosystem. So whether that is, you know, Python packages or Ruby gems or Rust crates or, you know, like Perl, like everything, right. That's all there. Or like Kubernetes, when you download, you know, from Kubernetes, you were, you know, facilities helping deliver that download. There's a lot that we're doing sort of enabling, but then also being able to go out and seek out places to invest in providing services, support community to, and the Fediverse was really the manifestation of this. I think that we really nailed, was like, as Mastodon took off,

you know, when you have a, they're, you know, decentralized, open source social network that had its huge growth moments as a lot of the big platforms have been sputtering. And in a decentralized federated network, success is technically fairly hard to distinguish from Adidas right like they're

like you squint at it and they're pretty similar looking and so like i said you know if something is slower and efficient or you're being ddos what do you do you sprinkle some fastly on it so they kind of felt the same way they're like what are we going to do um how do we one solve the problem for our users and two feel good about who we're doing business with. And we were able to help them out. And, you know, as we record this, I'm sitting in our boardroom at the New York office of Fastly, and we had, you know, Eugene and the team that builds Mastodon sitting here a couple weeks ago,

talking with Hannah Aubrey on our team who leads the Fast Forward program. And we're just like, what do we need to do to help you thrive and succeed? And, you know, they were like, I'm paraphrasing, but almost exactly, it was like we couldn't have done it without Fastly. And that's like an incredible feeling to be like, we can bring to bear the resources of this organization in a way that is both vital to the open source world, but also we can, if we're testing out new configurations on the platform, we can be like, hey, Kubernetes, do you want to kick the tires on this and see if it works for you? Because if it works for you, it'd work for everybody else

because your scale is so big. So we're getting learnings out of that. It's not, there's a fair exchange. And also like unapologetically, I'm like, yeah, if you're using any of those open source platforms and it gives you warm fuzzies, that fastly is helping you to come be a customer and tell your boss that we should be using the platform.

Like, I think it's a fair trade. Like, I think it's a really reasonable, like I said, that sustainability piece is there. It's not like just charity work. I love that exchange being understandable and legible. There's not some secret skill degree. And it's not like, oh, we just have the purest of hearts. Like our motivation is good, but it is tied to like business value.

And so that's how we know it's sustainable all these years in that we can sort of support these programs and they don't have to worry about the rock getting pulled out from under them and that's the part where i see like to me you know i look at something like the xe vulnerabilities and exploits and it's like that the vulnerability there was human fatigue was that the utter exhaustion of being a maintainer a thankless maintainer of an open source project and like we can't solve all of those and and you know we're just one company in an industry full of tit like, we can't solve all of those. And, you know, we're just one company in an industry full of titans, but we can meaningfully change things

for a handful or a pretty large cohort of open source projects for whom infrastructure that would be unimaginable to them or impossible for them is something that we can put our energy behind providing and lighten their load you know that to me is like the the fact that i was you know the part of a team that was able to get that going

so quickly after we came into this organization is like the latent desire was already there you can't make a company be something it doesn't want to be you can't make a culture be something it doesn't want to be but if the desire was there but they were just lacking like who's gonna lead or execute or contribute or collaborate on it um then you to be. But if the desire was there, but they were just lacking, like who's going to lead or execute or contribute or collaborate on it, then you can make it happen. And so I think that's why we've seen that success in Fast Forward, the open source program.

I think we've seen why we see that success with Glitch getting to flourish. Why are we going to get to do many of the other things that we've got coming is we're going with the way the stream wanted to flow, right? We're not paddling upstream. I really want to thank you for taking the time to talk with me about this. For what it's worth, you've turned me into a believer of sorts. Not that that was much in

doubt. I've been a fan of what you do for a long time. If people want to learn more about what you're up to these days, where's the best place for them to find you? So what I would say is everybody here that still is motivated by that belief in the open web and the good internet swing by glitch.com build something get back that feeling that the first time you did view source in your browser or you made that thing light up it feels good to do it and remind yourself of what you love about the internet you know with my business hat on absolutely if you've got any part of your site or your app that is too slow, not secure enough, not performant enough, whatever it is, you know, come to fastly.com. We'll hook you up.

We'll help you out. And you know, you got some part of your thing that's not, you don't love, that's not working right. Me personally, you know, I try to practice what I preach. Anil-.com is my site. I've been blogging there next month, 25 years. You know, you will find bugs in it and flaws and stuff. And they're like, I hacked together the CMS and like all that janky personal website stuff. Still doing it. I love it. It's a hobby. And thinking of it that way, like, you know, I cook meals at home and they aren't all winners,

but they feel good when I get them done. You know, I feel the same way about like, we should have a lot more locally grown, organically raised farm to table technology and websites made, you know, with love, with an old family recipe shared with people that you love to sit around the table with. Like there's no reason we can't have a lot of home cooked internet sitting alongside our fast food internet. And so, yeah, just really encouraging people to spend their time doing that, reconnect to what they loved about tech in the first place you know and if you didn't have that relationship where you're just like well i came to the industry because it's a good business opportunity

seek out the people for whom it is a you know a motivation for them that is a cause for them that has meaning to them and and ask them why because that is a thing that is has animated my work it's been able to keep me motivated for decades and doing this stuff because I get to be surrounded by people who think the internet is something special and it's something that we can make together. And yeah, I'm still every day motivated by that. Well, thank you. We'll put a link to that in the show notes to be sure.

Thank you so much for taking the time to speak with me today. I appreciate it. Thanks for having me. Anil Dash, VP of Developer Experience at Fastly. I'm cloud economist Corey Quinn, and this is Screaming in the Cloud. If you've enjoyed this podcast, please leave a five-star review on your podcast platform of choice. Whereas if you hated this podcast, please leave a five-star review on your podcast platform of choice, along with an

angry comment in which you wind up talking about how excited you are for your company's getting acquired with the same dead flat tone that you would expect from someone being held at gunpoint slightly off screen.