Daniel Whitenack: I’m struck by the scenario that’s talked about in this article from Erik. He talks about a sort of mid-stage startup around 10 million… So that’s about the size that my wife’s business is, and looking at her marketing and sales/customer service department - if you think about that early stage, like you were talking about, it was basically her, she built up a ton of expertise and internal knowledge in terms of what was working and what was driving sales, and that basically boosted the company to mostly where it’s at. But then you start thinking “Okay, well it’s at a size where we’re hiring in marketing people, or people that are supposed to be driving sales.” Is it reasonable to assume that each of those people are going to have both the ownership over the business and the drive to build up that level of internal knowledge, and there’s gonna be appropriate knowledge transfer between all of these people coming in? It’s just not the case.

Like you say, you hit this wall where “Now how do we be creative, how do we try new things, and how do we make sure that we’re driving new sales, and growing?” It has to be data-driven at that point. But the culture wasn’t sort of set up that way organically. Not because they weren’t wanting to be that way, but because it just sort of organically grew into this department where they’re doing the things that, like you say, they know worked to some degree, and they felt like were still working… So I think now, in her company, they’re doing a lot of thinking about “How do they drive that data-driven culture in marketing?” And some of it is just the very simple stuff that even Erik talked about in his article, like “Do people understand how UTM codes and website traffic works?” There needs to be some knowledge sharing there, and then there needs to be common data gathering, like “Okay, we’ve got this stuff over here in Facebook Pixel, and this stuff over here in Google Analytics, and this stuff over here in Shopify, and this stuff over here in these random places…” No one can really coalesce around anything if all of that is fragmented out, so there needs to be data aggregation together, there needs to be a common way to look at it…

[28:20] And then, you know, building that culture - it’s also about people’s motivation. You have to think about “If I’m gonna show something to this marketing person, how are they motivated by that?” I mean, it could be like commissions, or something. If you make this much off of Facebook ads, then you get this commission or this incentive. Well, pretty quickly they’re gonna wanna know how much they’re making off of Facebook ads, and if they’re not setting up their UTMs right and they’re not using the common systems where data is coming in, then they’re not gonna be able to know

Jonas Downey: [32:12] So part of the thing is that when we set out, we wanted to design something that’s gonna be kind of premium. Because if you are a free email user, you use Gmail, and you’re fine with it, and you don’t care about email, and you don’t pay much attention to it. You’re probably not gonna go out and spend $100/year on a different email service. So the people we’re marketing to are gonna be people who aren’t that. You have to be at least somewhat intrigued by the idea of email not working for you in order to even care.

So I think our collection, our little ecosystem of features of HEY that make it premium are a few things. One is just the hey.com address. It’s just a cool email address to have. We’ve spent quite a good deal of money to acquire that domain…

Jerod Santo: Well, let’s talk about Plausible Analytics a little bit here, and set the stage… Because this post that you wrote was a brilliant piece of content marketing for Plausible Analytics, which is a service and a tool and an open source application that you two are working on. So the pitch is uninstall GA – and I would say, why do we still use Google Analytics? It’s because, well, what else are we gonna do? It’s free, it’s easy to set up. What else is out there?

So this was a nice piece of marketing, because it’s like “Here’s this great post all about it, and here’s some alternatives… By the way, Plausible Analytics is something that we’re building, which is an alternative to that.” So I would love to hear all about how that works, and some of the stuff Adam’s bringing up.

[12:10] Let’s pull Uku into the conversation, because you’ve been waiting in the wings here… Uku, when did you start building this, and was it because of the reasons that we’ve been discussing with regard to the status quo of tracking and analytics with Google and other large providers?

Chris Benson: They need good marketing and you’re providing that for a free service.

Anthony Eden: Yeah, so over the years I’ve had moments where I’ve really loved doing this, and other moments where I haven’t. Because operationally, it’s 24/7/365 operations, right? It never shuts down. We can’t shut down. And there were ways – I felt a lot better when we first started getting team members in, who could share that burden, to make sure that was happening. And then I spent a lot of time writing software, until I really wasn’t needed to write software anymore, because we had a team that was good enough to write better software than what I could do. And so then I started shifting focus. “Okay, I’m going to focus on customer service.” And then I was going to focus on marketing. So I’ve kind of evolved through all the different – I pretty much have been in every position at DNSimple, shockingly or not shockingly.

At the end of the day though, I’m a nerd for DNS. I don’t know what it is… I was in an ICANN meeting recently in Seattle, and I was loving it. It’s so nerdy. It’s policy stuff. I mean, it’s crazy.

Jim Haughwout: [01:02:04.03] Yeah, I’ve been on a few multi-hundred-million-dollar SAP implementations, and it’s a big investment, it’s a lot of change. What we’re trying to do with Backstage is make it easy, and start with one service, add the next… Stefan had mentioned in our product marketing team we’re trying to treat this as a true go-to-market. So when we got that feedback, we’re trying to make the templates the getting started pack, the documentation easier, so you can get it out of the box. You can even install it very quickly and get started, just like any open source.

As we’ve been doing that, as Stefan mentioned, our adoption has gone up. We’ve talked to over 200 companies now, we have 15 committed adopters on our list, that include some very interesting companies that have shared what they’re doing from that perspective. At this point, Stefan, how many external contributors do we now have?

Daniel Vassallo: Honestly, nothing very specific. I got to approach it like I’ve approach 2022, 2021 and 2020, with just an open mind, taking care of my downside. My main concern, my only goal, I would say, is to sustain my current lifestyle and to not go back to a nine to five job. I know it’s not the end of the world if it were to happen, but for me it would be a significant lifestyle downgrade.

[58:10] So the attitude of doing whatever it takes to maintain this, I think it’s a helpful attitude for me. As I mentioned before, I almost feel incapable of being highly creative, highly productive, unless I feel the need to. So right now things are going well, I have sort of this routine, I’m doing these regular cohort courses, I have the Small Bets community, there’s a steady stream of people joining, things are going well; I don’t feel I need to do any major changes… I have some plans to improve things, working on some small improvements here and there, but nothing radical.

But again, I recognize that this thing might not last. No idea whether by the end of 2023 it’s still going to be working as it is today. It could be much worse, it could be even better, or anything in between. But I’m fine with that. I’m still with my eyes open, and if I run into some idea, some opportunity, I have lots of slack in the system right now, mostly in a state of wandering about, waiting to pounce on something… And to me, basically, after the year that started, I would have been incapable of predicting how it’s going to be ending. So if I were to get this question, at the end of 2021 I would have certainly not predicted how 2022 would have finished. And same thing for the previous year.

So I think I would be foolish to try to predict where 2023 will end, and again, for the type of personality that I have, I think I like it this way. Some people might say, “Oh my God, it would be a nightmare for me to not know what I would be doing, how much I would be making”, but I realized actually I think I need this… Because if I don’t have some variance, some unpredictability, some ups and downs, I lose almost the drive to actually do things. If things are too steady, I become incapable of working. So I recognize it, and when things are good, I take it easy; when things start to show some signs that they need some correction, I don’t need to convince myself that I need to work harder, or try other things. Suddenly, almost another part of my brain activates, and I become super-motivated. Again, I start seeing better, new opportunities, new ideas, maybe new, creative ways of marketing my existing products and services, and so on and so forth.

Dan Guido: I’m thinking about this like “I have a budget, and there’s only so many things that I can do.” First off, before I overwhelm the user – like, already I mentioned that there’s about 5-10 questions you need to answer when you deploy Algo VPN. I was always worried that was too much, and there’s been certain features that we’ve eliminated from Algo VPN after we realized people didn’t use them, and it made operating it too complicated.

But yeah, there are certain things, certain features that we can add to Algo VPN that would compromise the ideals, that would compromise the mission that it has, the values that it has. Installing OpenVPN is a good example. OpenVPN as client software has had tons of issues. If you wanna force your users to install client software, they have to patch that. I don’t wanna make people patch anything. This should be software you don’t need to patch. It should have enough defense and depth, and it should re-use existing components so that it’s much lighter-weight, which means I don’t have to think about it as much.

[32:10] We don’t wanna depend on TLS either, and OpenVPN does. TLS has been riddled with all kinds of security issues, a lot because you can negotiate those ciphers and negotiate the protocols that you’re using with TLS, which OpenVPN inherits. This is a total shoehorn in the first place.

TLS is not made to be a VPN. It operates at a totally different layer, it’s made to do something different. It was a poor fit to begin with, and we shouldn’t be using it here.

And then yeah, the marketing, really to differentiate ourselves from the VPN services. There are things that a VPN can do, and there are things that it can’t… And what it can’t is it can’t prevent you from getting arrested by the FBI, as we’ve seen many times. But that’s kind of how marketing works for these people. They say anything, there’s no repercussions about it, and a bit part of Algo VPN is I don’t wanna lie to anybody.

Guillermo Rauch: Everything that’s scary in a lot of cases ends up being liberating. I actually was listening to this great podcast of Gary Kasparov and Lex Fridman… And he lived a very interesting first-hand experience of what it’s like for AI to defeat what seemingly you had that was, first of all, unique in the entire world… Because if you’re rating as a chess player, but also you had what’s seemingly considered for a computer not ever be able to do. What he said in the podcast that I thought was really interesting is that “Well, that’s awesome that they beat us.” Okay, the computer is getting another 9 in the SLA of things that they can do better than humans. So now they’re a 99.9. And ten years pass and they get to 99,99. Here’s the deal - the remaining 0.001, which is human capability - now it just matters more. It’s not that because there’s fewer things in which you truly excel that your impact is diminished. If anything, it’s the opposite, because now you’re truly able to focus on your creativity. And I think ultimately, that’s what we’re seeing with frontend. Like, okay, now you don’t have to worry about the cloud at all. You might not even ever know what a server used to be or looked like. You might not even understand what the infrastructure is made up of. The other day I didn’t even know this anymore; nor maybe ever.

[01:13:59.05] But I’ll give you a really good example - S3, the CTO of Amazon said, is composed of 250 microservices. And I was like, “That’s insane.” There’s 250 independent services that are making up this greater service that I interface through two API calls, get and post. And patch maybe, sometimes. And it’s just astounding. I think that’s just ultimately gonna happen with – again, we’re continuing to like… Okay, now Vercel is helping at the other 9. Not only will you not know what file storage - how it actually works, now you’re not gonna know how a page is built and rendered in the cloud… But that’s awesome, because now your remaining 0.001% is how you express your brand online, how you engage with your customers, how you choose to communicate with them, what your values are as a person doing business, how you express your identity, and all these amazing things… And ultimately, your own product that you’re building, whether it’s hardware, whether it’s software… How you give that service, in what language it is, in what markets etc.

So I think it’s awesome, and I think we’re just gonna continue on that trajectory of more and more undifferentiated work being automated, and more power to you.

Kurt Mackey: Netlify - we’re actually pretty decent friends with them. Netlify is a company that if we’d reverse to started when I think you would have built Netlify on top of Fly. Fly I think is a good platform for building a Netlify type service. There’s definitely overlap, just because if you’re targeting developers, you end up doing 5%-10% the same things that everyone else is doing… But Netlify is a really good possible customer.

ZEIT is really fascinating. They made some genius code. I feel like ZEIT is a better Heroku… And I should say a more modern Heroku, because it’s actually still really difficult to make something better than – Heroku is still really good. The developer experience just works; it’s all kind of still amazing. I think if Salesforce hadn’t bought them, it could be a really interesting alternative to, and asuming it had stayed alive and gotten bigger, Amazon, Google Cloud and Azure part of the world.
The people we get most often asked about and compared to is Cloudflare, because they have service workers, and Cloudflare service workers look a lot like Fly apps. I think it’s a natural API to implement for the types of stuff we’re doing.

The big difference we have with Cloudflare, and investors asked this one a few times, is AWS is like a fully programmable data center; you don’t ever call a salesperson to turn on a server or something else for you at AWS. Previously, there were less programmable data centers, places like Savvis, places like Verizon, AT&T, all these guys, Equinix sold space and power to people. Cloudflare I feel like is a slightly older generation of a similar problem we have. Whereas we have an API for creating a WebP image out of any other image to serve it up to the right kind of browser, they have a button you check to turn on the WebP. They’ve put these service workers in the middle, so you can script some stuff and there’s a fairly decent set of things you can solve with Cloudflare that you could potentially solve with Fly too, or vice-versa… But we’re like developer-first all the time. There’s not part of anything we do that we want to hide from an API effectively, instead of this being a smaller bit of a much larger stack.

[44:18] So you can see that because we’re open source. We also have local testing, so you can write and run and test apps locally, which means it works on like a continuous deployment process… Which is an interesting thing, because one of the people we initially thought we were competing with was Fastly, because Fastly is a great company that builds really cool stuff and is kind of the pinnacle of what a CDN should do… And one of the complaints we kept hearing from customers of Fastly - and there weren’t many complaints; everyone’s generally pretty happy with them - was “I’m terrified that I’m gonna change something and break my whole site. I’m a developer, I’m used to this whole development process, I’m used to writing tests, I’m used to having some level of comfort when I make changes”, that they didn’t give them.

Compared to those guys, I think we’re much more like the application-level developer-first, developer-all-the-way thing. I think more broadly we’re probably competing with anyone who wants you to write new code on top of them, like Firebase or any of those people. But it’s such a broad thing… The people we’re talking to customer-wise aren’t gonna compare us to Firebase usually. It’s just like – we’re kind of all trying to go the same direction.

Techno Tim: Yeah, odd. Yeah, yeah. Because it has the potential to go, I think, either way. Either people are burned out and don’t want anything to do with it – well, there are people who don’t want anything to do with it. There are people who are burned out about hearing it. And then there are people who might want to hear about it and want to do it. So it’s always – like every video, it’s a crapshoot talking about those things… But there are people, yeah. Absolutely. I mean, that video for me is performing pretty well, and people have asked, and there are other creators talking about this, too… Especially the DeepSeek thing really got people interested in self-hosting AI, because they wanted to figure out and play with this DeepSeek thing. And the easiest way to do that without sending all your data wherever it would be online was to host it yourself.

So oddly enough, I think that DeepSeek helped people realize that they can self-host LLMs, probably more so than Ollama’s ever done. You know what I mean? Because it did two things. One, it let people know that it’s a possibility, and it also let people know that “Hey, it’s private.” And those are two things that I don’t think a lot of people knew were possible. And nothing against Ollama. It’s fantastic, it’s great. But it’d be hard to market them, or that product or service, in the way that DeepSeek I think marketed it. All the DeepSeek hype really was marketing for other stuff too, which drove so many people getting interested in this.

But yeah, I think that’s what people are going to do. I honestly still feel like there’s an opportunity for a company to put it in a bow. And there’s going to be that opportunity, I think, for a long time, and do a whole product. But in the meantime, we can do this ourself. I just hope that companies continue to allow us to plug in our own models.

And one of the things that I talked about in that video too was that – I kind of hinted at it, but I use Mac Whisper for my subtitles and stuff, to get them transcribed right - because I care about them - and I think a lot of people should do; that’s my pitch for accessibility in general. You should care about them, because there are a lot of people who can’t hear, or have hard of hearing, or just don’t speak English, and they want to be able to read English.

So anyways, Mac Whisper is an app. It’s freemium, you can get it, you can download it, run it on your Mac… It does great with the small models. But in there, I saw – I paid for it. I paid for it myself, so I use it. And no relationship to them whatsoever, but I kind of talked about them in the video, where I said “Hey, it would be great if a lot of companies let you plug in your own endpoint, or plug in your own models.” And I kind of showed their screenshot of them only allowing ChatGPT… Well, sure enough, shortly after, now it allows Ollama. I probably had nothing to do with it, it was probably on the roadmap…

[01:04:14.16] But that’s what I want to see. I want to see, instead of application developers thinking “Hey, how can I hook into ChatGPT so everybody can now use ChatGPT?” Well, think about Ollama and other things, too. Think about how you can offer that option. And I think that’s a huge differentiator, too. If someone developing software right now says “Yyeah, we can plug into Ollama”, that’s a differentiator. Every software company right now is saying “Yeah, we can hook into ChatGPT. Just put in your API key. Here you go.” But I also think that they should be including Ollama, for sure.

Preston Doster: Yeah, yeah. So we started out really marketing to engineers. So people who wanted to build apps to send, say, SMS. And what we’ve grown into is this CPaaS, or customer service platform, really, where it’s all about – so like when you order from Uber Eats, or from DoorDash, or whoever, and you get that text that says “Hey, your dasher is waiting for your order.” If it sends it to you by SMS, there’s a potential that’s going through Twilio. Because they would use our API to abstract all that complexity. 2FA… Yeah, exactly.

Shay Banon: So that’s totally illegal usage of trademarks. That’s totally illegal. You’re allowed to say “I’m a shop that services Toyota cars”, you’re not allowed to say “I’m a Toyota service shop.” That’s the difference. And as long as the name is Supabase – they’re totally fine to say it’s an open source version of Firebase. That doesn’t create confusion in the market between Supabase and Firebase. And that is totally allowed. It’s fine. I actually encourage that type of marketing for any open source project that there is. Go for it.

Steven Wu: There’s kind of a couple things. So basically, we’re cloud-native. A lot of these older generation stuff, like Splunk, like Datadog, like stuff based around Logstash, etc. a lot of that tends to be – it’s very important to these companies, because of a function of the times that they have good on-prem solutions, and they kind of have these applications that are designed to be run on physical computers, and so on. So if you’re running a Splunk cluster, you have like a – maybe not a physical computer, but even on the cloud option it’s like a VPC somewhere, and that’s like a node, and that handles all of the capacities of the system, which includes concerns like storage, it concerns compute… And one of the big things is that for the most part your usage of those resources are not uniform over time, and they are not uniform in any given instant. For instance, you need storage all the time, but you pretty much don’t query most of the time. So if you kind of think about what the average query use case looks like, you spend like five minutes typing stuff into the search box every day, and then the other 23 hours and 55 minutes you’re not, right? So when you have physical computers that have some constant resource constraints, and some very spiky resource constraints, you have a lot of like weird utilization issues… And this leads to a lot of basically costs.

So in our system basically each of our things, each of the pieces of that system, each of those concerns are farmed out to the cloud service that is specialized towards handling that. So all of the index files go in object storage in S3, we run ECS clusters, or whatever… Well, not whatever; that’s what we do. I actually know that. For the actual indexing. And then we run basically serverless compute for doing the query, and each of those things are chosen specifically because that particular operation has a particular set of properties in terms of the volume, the spikiness in variance, the availability requirements and so on, that are very, very different. And I can kind of talk about any of those in detail; I don’t want to get lost too much in the specifics, but…

I think we’re kind of seeing a world now where in the current generation of these log search platforms people are more and more – you’ll see the term separating compute from storage more and more in people’s marketing material, because I think we are realizing now that even very, very large companies like Fortune 50 companies are moving all their stuff to the cloud, that you can now in fact build things with the assumption that pretty much even the biggest companies have enough cloud stuff for you to work with… Whereas I think even like five years ago you couldn’t. If you wanted to work with a big financial institution, you had to have some literal on-prem distribution where you gave them a computer. So I think it’s there’s a sea change that has kind of enabled this to be possible kind of in like the business landscape, and just in terms of people moving from on-prem to cloud generally.

Joseph Jacks: …physics-intense type of company. When you’re talking about the world of bits, raising tons of venture capital, in my opinion, is a scam. I’ve actually said that publicly. And I think that if you were to just siphon the size of venture capital and focus it on the bits companies - not the ones that are doing hard atoms type stuff, but just bits - it should be 95 percent smaller of an asset class, a cottage industry, a service industry, very much focused on the absolute best possible founders. And you do not need a lot of capital to make a huge, huge, huge, huge impact. That’s my learning. Can you make money with open source companies? Absolutely. Will there be more of them in all these different categories? Yes. Will there be 10 or 20 false positives in each possible category? Probably. Perhaps because maybe we overhyped the category. Because I’ve been doing a conference, and blogging, and tweeting about it, and there’s more funding going into these companies now, and all the VCs love it now too, and all this. That’s overfunding. That’s inefficient. That’s bad. It’s actually bad. In fact, open source companies, if they’re, again, focused on a pure bits business and not shipping a hardware product or something, they can be the most capital-efficient companies you can imagine.

Because if you have a huge open source community that is like – this is kind of a bad way of saying it, it’s kind of distasteful or whatever, but it’s true… Like, effectively subsidizing massive engineering costs that don’t have to show up on the balance sheet of the company, and also subsidizing go to market, marketing, lead gen, customer acquisition… Because the moment you flip the switch on a product that solves a customer problem, you have a huge amount of that user base that experiences the pain of that problem saying “Look, sign me up. I love your product, and I’ve been using your open source thing for the last five or six years. It solves a fundamental problem in our business, and we’ve been looking for the thing that you are selling for the last five years. We’ve just discovered it. Please sell it to us. Take our money.” That is an extremely efficient sales process, that rivals any other type of sales process you see with proprietary technology companies, that have to spend a ton of money on advertising, on sales, go to market, and all this kind of stuff.

And there’s a lot of other aspects to it too, that I didn’t even talk about. But it is very hard, and it’s very nuanced as well. One thing I think that is still very unsolved, frustratingly, and painful, is that I continue to see people who are not philosophically aligned, or remotely as interested in sort of the essence of this stuff as I am, who are very successful VCs, take the exact same playbook that we have, and then when the companies get to a point of 50, 100 million in revenue, a few hundred million in revenue, they really start to heavily encourage the founders to change the license of the core technology. I find that to be unnecessary, duplicitous, harmful, two-faced, and despicable, to say the least. I could probably talk a lot longer.

Adam Stacoviak: This Microsoft era you mentioned that was pre-ZIRP, and maybe even after it - like, I think one of the reasons maybe there’s churn is that it’s a hard, even to this date, kind of hard to find what exactly is Dev Rel. And when you have a challenge defining what it is, you have a challenge in defining what it should do, what the function should do, which maybe is the reason why people flunk out of the job, or move along or churn, because maybe even in the Microsoft era that you referenced, where they hired lots, it was challenging to define “What exactly are you trying to do?” Because if the job primarily is focused on a function that sits between the company, which usually is a tech product of sorts - a SaaS, a dev tool, a dev service etc, maybe even an open source company, a [unintelligible 00:21:26.09] company, or an open source project - they can have Dev Rel as well - if the function is to nurture the relationship between the company, potentially the purchase of a product, and the developer community, there’s a lot of ambiguity in there in terms of what you could do to be successful. And it might be challenging even as a manager to manage Dev Rels. Like, what do you really do here? What can you do here? What is success in your role? And when you have lack of clarity as an individual, it’s kind of hard to maintain what a good friend of mine, coach Michael Burt says “the prey drive”. You have to have a reason. It’s your because goals. “Because I’m a Dev Rel, these are the things I do in my role.” Or “Because I want to do these things within the developer community, I have clarity.” When you have ambiguity in your role, or a lack of clarity, it’s kind of hard to kind of wake up every day and be motivated and get something done. Or when you’re tweeting or doing social media, or doing these things that aren’t really seen by peers, or adjacent peers, like engineers, or marketing, or sales… It’s like, that person is Dev Rel and they’re just tweeting. Like, is that work?

So when you feel like you’re not clear, it’s kind of hard to kind of just get up every day and just do you do well, unless you’re a self-motivated person. So I guess all that to say, how much of this churn is because the management or definition of what the role is, and what success is of the role, is well defined, so they can be successful?

Indu Alagarsamy: Yeah… I was thinking, like you said, Chris, there’s trade-offs. And also how you design these things. You can use the orchestrator, you can use the choreography, you can do both. So let’s say that in the process manager – so you’ve shipped off this thing, right? So the process manager [unintelligible 00:36:14.14] events, and you’re good. Now, let’s say business has a new requirement that says “Hey, we want to track our customers who are interested in this product, or customers in this area, that we ship to.” Something, an interesting requirement that comes in. Now, do you go and modify your existing process manager? You could. You could go and add logic after the thing got shipped, go do this other thing. Or in your choreography model, you have an event that says “This got shipped.” So you could have a listener that listens to that event and does that extra bit of logic, or keeping track of customers, or something that was interesting for marketing. And that is completely autonomous. And I think this is where, for me, the power of choreography comes in, because now you have these – you can write your services that sort of align with your business needs. So in one model you would go and change your process manager, making changes. In the other, you just introduced a new service that just consumes this event, does its thing; you don’t have to touch your process manager… So I think there’s a balance, right? There’s a balance in - yeah, your business requirements, your needs, and where does choreography fit better, where does process manager fit better… And just because you have process manager doesn’t mean you can’t use choreography. So it’s all about trade-offs. Yeah.

Abdel Sghiouar: Well, I think that there are a few things that have been mentioned by the two of you. So let’s start from the last thing you said. I think debating Cloud Functions as a service versus Kubernetes is the wrong debate, because Kubernetes can do more than just functions, right? If you really think about it, it has APIs for functions, jobs, apps, static, stateless and stateful… It has pretty much an API for any sort of deployment that you would need… Which technically developers are in either of these three categories. You’re either writing a backend, an API, something like that, or a frontend, you’re writing a job, with pipeline and orchestration, or you’re doing functions… Which, in my head, functions are just a special type of backends, right?

So Kubernetes can do a lot of things. So the question then becomes, for you, Joe, for example, if you tomorrow need to go from writing Go functions that deploys on AWS Lambda, to glue a bunch of services together to write jobs, are you going to be able to use lambda for that? I would probably assume the answer is no, and you have to pick up another tool. And then we could argue there is also a learning curve for that next tool you have to learn. This is one thing.

The second thing, which I think is so important… And I’m not tooting on our own horn, but there is a person in my - he’s actually my director, Richard, and he wrote a very interesting article debating the… There is this movement called Shift Left, which essentially is what DevOps is all about, like making more developers more empowered, more responsible etc, etc. And his argument is “Instead of focusing on shifting left when you are doing modernization, you should focus on shifting down.” Essentially, shifting down means giving that complexity to the cloud provider; letting somebody else deal with all the complexity… Which, because we’re on the topic of Kubernetes - I mean, using a managed Kubernetes is pretty easy. Like, deploying your own Kubernetes is the problem. But using a managed version of Kubernetes is pretty easy.

And you said something very interesting, Joe, that is getting out of your marketing Manchester to work with international customers you realized a lot of people use Kubernetes. I think that that’s the unfortunate consequences of the DevOps and Shift Left movement. Because what’s happening right now in our industry is that we’re just expecting developers to know more. 10-15 years ago you’d go to work, you’d write code, you’d ship a binary and you’d go home. That’s it. That’s a Java developer. Probably you do some tests, or not… No one expected you to know how to SSH into a server. No one expected you to know how to set up an SSL certificate. That was not even an expectation.

[16:23] But now, I think you’re all aware there was this massive thing called full-stack engineering, where we suddenly expect engineers to do everything: frontend, backend, security, storage, databases… Like, when was the last time you heard of a DBA? It was maybe like 15 years ago. They probably still exist, but we don’t hear about them that much.

And I think this is the core of probably the – I wouldn’t call it the problem, but the debate… It’s like, why do you as a developer feel you need to learn all of Kubernetes? So the core of the problem is, if you think it’s complex because you have to learn it, then the question is “Why do you have to learn it?” Are you in the right setup? And what I mean by the right setup, I mean the company, the organization, the team, the way you are structured, that is trying to utilize your time to the maximum, instead of having you waste time doing things you don’t need to do.

I’m going a little bit philosophical here, but I think that the whole question then is, if you don’t need to learn Kubernetes, then just don’t. It’s not really that big of a deal. And if you feel the need to learn it, then probably the question is, “Are you organized in the right way, the most efficient way, or not?” I don’t know if what I’m saying makes sense…

Michael Grinich: So I’ve been driving forward sales, and doing that – I’ve been doing sales. And I’m not like an expert sales, a VP of sales person… But I’ve sort of figured out how to do it for our type of business, and for what we needed. But I don’t get to continue to do it to the point of mastery. I hand it over to someone when it’s just kind of put together in a way. And I don’t write code and build stuff anymore, at least on the production service. We hire a lot of other people to do that, and build stuff.

And so you have to love the process, I think, of like constantly handing over your Legos, giving away your Legos to other people… And the job for me is actually somewhat the WorkOS his product, what we’re going to build there, but actually for me it’s building the company. It’s kind of like building the machine that builds the machine. How should we think about sales and marketing? What should our strategy there be? What does it mean to build a design culture that’s really technical? That’s a hard thing for us; we’re designing software, not just kind of consumer products. There’s so many questions and levels to it that the job feels really different every couple of months, honestly.

And then I think the other counterpoint here is I just really like building tools and platforms for people. There’s a certain kind of satisfaction from creating something and giving it to someone else, and then it enables them to create another thing, that you could have never thought of, or you would have never built yourself. It’s a very high-leverage thing to do.

My favorite product, I think of all time, is actually AWS. Not because I think it’s particularly well built or well designed, but the impact that AWS has had, the amount of leverage that it’s given people is astounding. There’s businesses that exist because of AWS, that would have never been built or never been even started in the first place. And so we’re doing that a little bit with WorkOS; we’re able to kind of give people that leverage to expand their business, and grow. But I don’t know, I kind of like working on infrastructure things that are sort of in the sidelines.

I guess, kind of backstage… Sometimes I use the analogy of like - WorkOS, and the people we hire at the company, and all that, we’re not really the people on stage, with the spotlight on them, singing the Prima Donna. We’re kind of like the people that run the soundboard; that are supporting the opening act, or supporting the main act… But without this, you can’t put on that kind of a show. But it’s not necessarily the thing that’s drawing the most attention. And we find that so really satisfying to do, that type of work, enabling other people.

Tim Stiles: But if you’re writing something truly novel, that no one else has done before, and you’re the expert, you really have no fear of anyone copying you, because… You have to market yourself, that’s the big thing. You can’t be the quiet guy in the corner. You can’t be like “Oh, I wrote this thing”, and then you let some super marketing guy just come in and be like “Ah, this is my thing now. I made it” etc. You can’t let them do that; you have to have some hold there. But I don’t see much of a difference between a company whose product is open source and whose it’s not; it’s still the same thing for the most part. But you do get the benefit as an open source company, as people can believe that your software is good, because they can literally see it. The proof is in the pudding there. They can look at it.

And then with a lot of biotech software, they’re like “Just believe me that it’s good. Please, give me $100,000 to start, before we even show you the code. Just believe me.” And so it’s a big difference there. And then you get into – you know, there’s obvious code quality, then you have people that are submitting pull requests and little bug fixes, little documentation fixes… It’s not like you have a real engineering force behind that, but it does add up. And it does give you also some marketing power. Now people are like “Hey, have you heard about –” Word of mouth can happen there a little bit. I show up to meetups here in the Bay Area, and people are like, “Oh, yeah, you’re the guy who wrote Polymerase.” I was like, “Nice to meet you.”

So there is some real value to it, especially if you’re not the traditional purveyor of truth, if you’re not the PhD who’s gone to Harvard, who’s the fancy guys, fancy guys, grad students… It’s a great way to show that you’re the best, and you’ve got the right stuff.

So I feel like there’s gonna be more companies, hopefully, like GitLab, Red Hat, MongoDB… And I’m hoping companies like Gitpod, and NextFlow, sort of get to this IPO stage too, where people can see the engineer, and they can see the service, they can see the value in it, and they don’t care that it’s open source, because they know that it would cost them 60k to have a dev do it internally, or it could cost them 5k to 10k to hire the guys who wrote it. And I think that’s just sort of - I hate to use VC terminology. I’ve been spoiled by the city… That’s “the moat”. You are the moat. You’re the creator. You’re the person who made this. You are the artist who brought it to life, and there’s some real value behind that.

John Spurlock: There are certain agreed upon ways… So there’s a few things. I could talk about this for a long time, but let me try to keep it short. Basically, you need to identify the things that are common, like what the app is, right? So you have to know, is this a request coming in from an app that a user is using, versus Chrome, or a bot, or something like that? So that’s one dimension, is like what the app is; you can get that from the user agent. The harder thing is IP address, right? So there are ranges where you know to expect some interesting data coming in from them, for various reasons. Some are VPNs, some are corporate networks, where you see thousands of requests coming in from the same IP, with different user agents, so they’re legitimate… And then you see the bad actors, you see the bots… Any open system is open to all sorts of automated traffic that you have to identify. And they often use Overcast as their user agent, right? So you have to do some application of filtering at the IP level as well.

That used to be easier, because you could kind of identify users versus servers, and just kind of block all servers. But now it’s harder, because a lot of people use VPNs, and a lot of people use Amazon Web Services for listener apps. So a lot of traffic goes through Amazon that is actually representing a listener hitting Play on an episode. So it has to be sophisticated.

There are shared – obviously, Amazon publishes their ranges, there’s some public server lists, and I plan on incorporating as much public info into the calculation as I can… But as I’m sure you know, there’s like day-to-day things that you have to identify. Like, you kind of rate limit some things, and then certain ones pop up and you have to add exceptions for those… That will just be part of the operation of the site. But that’s part of why sponsorship is needed, because there will be some aspect to this maintenance-wise going forward.

There are certain – and I’ve talked to certain people that are doing something similar. So they are doing their own for-pay analytic service. They have the same problems, they need to identify the app, so there’s common ways of identifying user agents… And I’ve been having good conversations on sharing a lot of the even IP ranges as well, because that’s something that – it’s not super-high-value-add, especially for identifying VPNs, Tor, that sort of thing.

So I could see – there’s some existing server lists, but I could see kind of getting together and putting those in the public as well, specifically for podcasting user agents, because that traffic actually looks a little bit different than regular web traffic.

So Cloudflare actually has really good bot detection, but it’s really horrible for podcasts, because their bots look different. It’s something that will have to be sort of domain-specific, but I do have – it’s not impossible, right? And it’s something that you do the best you can. We haven’t even talked about – I guess we have a little bit, the downloads themselves are not listens. So it’s already sort of like – you already take a percentage and just say… You just pick a number and say – so you really do the best you can and you identify certain anomalies as they come up… But anyone in accurate digital marketing has to deal with this.

Adam Stacoviak: We didn’t ask if you’re a listener of this show before we brought you on this show… But if you are not a listener of this show, I would suggest you go back and listen to Adam Jacob talk about this, because… I asked you that question thinking “What is he going to say?” Because I kind of know what your answer might be, but I’m kind of hoping that it is in light of what Adam Jacob said… Which is, essentially, they’re your marketing funnel, right? Why get upset when AWS offers your thing as a service, because they’ve just blessed you, essentially, as worthy. Worthy to use, worthy to try, and let the tech, the usefulness of the tech and the community behind it and the people behind it be the support mechanism to say “This is worth keep using.” Versus “What is Typesense? Who are they?” when AWS hasn’t chosen you yet. You’re just a in a sea of obscurity, essentially, of search land. And if they blessed you in that way, then it’s like, wow. That’s a better go-to market strategy, potentially, than the free tier of Algolia.

Adam Stacoviak: Yeah. The maintenance of that is scary, because as you had said before, back to the hero analogy, as you had said, you have service owners inside of FireHydrant, so you’re sort of identifying these owner stats, as you mentioned. In an ad-hoc scenario, you’re not doing that. So you’re allowing the heroes to be heroes, right? Which is okay. It’s not terrible, but that means that there’s a bus factor there. If that person leaves, then all that manual process and the knowledge of how to do those manual processes leaves with them. And so in a high churn environment, which is engineering, specifically… So if it begins in engineering, but then permeates to the rest of the business world - marketing, legal, whomever else - then that knowledge of how to deal with these things, if it’s ad-hoc, that ad-hocness also leaves. And someone has to relearn, by – how do you learn? By doing. Having more incidents, right? So it’s kind of scary, honestly, in a lot of ways.

You mentioned Honeycomb before, and Christine Yen has been on the show before, and I would call her friend… And one thing we talked about recently, specifically to their most recent fundraise, was observability for everyone. And this sort of like “for everyone” after it seems like maybe teams currently use observability as the smoke signal potentially. Observably is almost the ad-hoc version of incident management. Because observability is like “Did my CPU spike in production? What’s happening in production? Alert me. Tell me. Let me go there and ask questions”, all that good stuff. How do you see, I guess, since you mentioned Honeycomb - do you use Honeycomb? Is that part of FireHydrant’s tooling? Do you use that behind the scenes? That kind of stuff. And then two, how does observability play into the bigger picture of incident management, and then more importantly, reliability?

Johnny Boursiquot: If you’ve never done a thing before… For example, a few years ago I was asked to look into this whole serverless thing. “Is this going to work for our workloads?” and things like that. And I was given a particular problem to solve, and then it says, “Well, can we do it without EC2 Instances?” I was like, “Well, I don’t know, it depends. How long is the workload?” Because at the time, I think there was a five-minute time limit, two Lambdas or 15 minutes, whatever that was, whatever the initial limit was. And I was like, “Well, if your workload is going to take longer than that time, then no, you shouldn’t use serverless; you shouldn’t use Lambdas and things.” And they’re like, “Well, what if we break it apart? What if we do smaller things, a bunch of small things”, like the best practice at the time? What if we make a bunch of different Lambdas, or invoke the same thing, put it in the pipeline and invoke the same Lambda multiple times, and you can scale it infinitely, according to the AWS marketing? Just do that instead.

I’m like, “Well, now we need to orchestrate”, and at the time AWS had Step Functions, which I think pretty closely followed Lambdas. Or at least the use of Step Functions for Lambdas basically became a thing. And then I’m like, “Well, orchestration - yeah, rather than to have one Lambda call another, and then one function waiting on another and timing out the whole thing, and now you can use step functions, and you can have ASL, Amazon State Language, whatever it is, you can write some of that YAML or whatever, and basically orchestrate all your things” - these were things that we had never done before. So in that case, yeah, I’m definitely doing a proof of concept, because I don’t know what I don’t know; I need to understand the characteristics of the platform, what it offers. I need to understand what all the knobs are right for the service that I’m supposed to be using. So there’s a lot there that I need to do.

So this is kind of an example that goes that’s sort of at perhaps the infrastructure level - to some code level, obviously, because the programming language that you use is going to have an impact on your serverless stuff, no doubt. So if I’m working with a Java project, because of the JVM, and I know that needs to be booted up, whatever it is, I’m going to take that into consideration. If you’re telling me to build something with Java and Lambda, as opposed to if you tell me to build something with Go and Lambda, then I’m not going to suffer the cost of a JVM. So again, all these trade-offs that you’re looking at.

[48:13] So again, the more you don’t know about the kind of problem you solving, the more I think you need a proof of concept. And that goes – this was an example at that sort of infra level. But if somebody says, “Hey, decode this data stream.” Like, I’ve never seen that before. It’s not like anything I’ve ever seen, so I might need to do a proof of concept to sort of understand what I’m doing. So I think it’s going to depend. To your point, Ian, I think it’s just something that we don’t naturally gravitate to at first, because - maybe it’s hubris, I don’t know; maybe we think we can just start working and we’ll figure it out, and whatever we come up with, “Ah, ship it to production.”

Gerhard Lazu: I 100% agree with that. 100%. Honestly. If this was a bull’s eye, you’ve hit it. So I can see how shipping, getting it out there, whatever “out there” means, by the way - whether it’s a product, whether it’s a service, whether it’s a utility, it doesn’t really matter… Getting it in front of customers, hopefully paying - that’s what we all want; nobody wants to work for free, because things aren’t free… So typically what happens is you get all types of layers between the engineers, the ones that are getting it out there, the actual code, the value, hopefully, the business value… It’s not just code; there is some value being delivered. The customer is using it, and then the customer is feeding back what is not working, or what could be improved, or stuff like that.

So how can teams and companies resolve this tension between having the different layers, like sales, marketing, product management, or just product, engineering, and the customer? Because the bigger you are, the more complicated these interactions become, the slower the information flows, and the more difficult it is to do anything. It can take many months, many years, and it’s just a function of the number of people involved. Too many actors, too much chatter… Everything moves slowly.

Autonomous teams - I can see how that would help, but what does an autonomous team mean? Does it mean like a group of ten people, one of them being a salesperson, a manager, a product person, a designer? What does that mean? What does an autonomous team mean to you?

Mitch Wainer: Well, the guides and the tutorials were extremely important for us in terms of first and foremost helping the developer community… And as a result, we were able to generate 3-5 million unique visitors a month to the website… Which - again, when you go back to that developer population number that I threw out, 30 million, 3-5 million a month, you’re hitting a lot of people. You’re quickly saturating the market.

You mentioned Etel’s name earlier when you were speaking with her - she ran the community tutorials, which in many ways was our best channel. Now, I believe the best form of marketing is to give first, and to not try to hard-sell people on buying your service or buying your product. It requires a certain level of thoughtfulness, and a kind of give-first approach. And that is actually the best form of content marketing, if you even wanna call it content marketing… But just helping people.

So the tutorials that were written, to your point earlier of like “Hey, I’m not a sysadmin”, I was able to kind of pick this up pretty easily… Etel had this saying – we had a standard for writing tutorials. The goal and the objective was to write it as simple as possible, so that a drunk four-year-old can understand it…

Lars Wikman: [08:07] Concurrency and parallelism at the same time, for essentially no extra effort. It makes you do concurrency and parallelism correctly and reasonably, without tripping you into sort of mutable state and the dangers of concurrency and parallelism. So that’s one.

Then there’s the whole resiliency thing, which is built on sort of the same idea, or some of the same ideas, where there will be things that happen to your application that are unexpected, that you can’t really catch with just catching an exception. Maybe the disk was full, maybe the service you were talking to was down… There’s always something to make it blow up. And it has been described as the “Let it crash philosophy”, but it’s not always the most – it’s not the best marketing. It makes managers very, very nervous… But the idea that it’s okay if certain components fail, and an important thing is to have a recovery strategy. This actually sort of feeds into the Kubernetes thing, which has a similar approach, but on a diffrent scale. And this sets me apart from a lot of functional programmers… Some functional programmer enthusiasts really like their types. I’m very glad that Erlang and Elixir are dynamic.

Paul Dix: So Amazon’s claiming they have the moral high ground, and it’s not true. They’re just doing what’s best for their customers and their shareholders. They’re trying to optimize shareholder value. And obviously, all their customers are saying “Host Elastic for us.” And then Elastic is trying to say like “Oh, we’re doing this to protect ourselves from Amazon, because they’re stealing from us”, and the truth is there are tons of hosting companies that have been hosting Elastic for a long time. And if you look at where Elastic makes its money, it’s probably mostly from log search. How many log search companies are built on top of Elastic, and they just open that up?

So Elastic is just upset because Amazon outcompeted them on the hosting front… Whereas other hosting providers, like Compose, and [unintelligible 01:19:28.08] and stuff like that, didn’t really make a dent in Elastic’s top line. So their claim, like “Oh, we have to do this” - like, no, you don’t. You could have kept the code Apache v2, and you could have written more and more code in your service offering that’s closed source, and kept the two separate.

This is actually one of the things that I agree with Amazon about, where they said the reason they created the open distro was because Elastic was polluting the open source repo with code under different licenses; under the Elastic license, and stuff like that. And what they’ve done now is they’ve gone all-in on that strategy. So basically, they want all the benefits of being an open source company, free marketing, free adoption, getting other people to talk about it, use it, whatever, but they don’t wanna pay the price. The price of being really open source is that you’re giving software away for free, yes, but –