Making DNSimple (Interview)

27:43And when they bid on an auction, some of those auctions got very expensive. For example, I think we were talking earlier about.app. .app, I believe, ultimately sold for$10 million. Google had to pay$10 million for that one TLD. The interesting part about it is the$10 million got split up against the other people who were in the auction. So they actually made money just by being in the auction. Who made money? The other ones who were submitting for it. I can't remember which other. For losing? They had money by losing? They made money by losing. Can I get into some of these auctions? I'm happy to lose.

28:18They had already put their$170 ,000 in to get all the way out of that phase, right? Well, they would have money to make money, I guess. Exactly. And so they had recovered some of those funds and moreover, they had earned money back from it. And there were a few companies that made quite a lot of money just doing that. and then they use that to fund the other TLDs that they bought. This round, that's not going to happen. ICANN is going to get all the proceeds. Oh, ICANN is going to get some of that money. So if you know ICANN, they actually have a huge budget. I think their budget for this year is around$550 million.

28:53$500 million? How many people is this and what do they do? I'm pretty sure they're at least over 1 ,000 people. So they run, not only run the... sort of the organization around the oversight of names, but it's also numbers. It's in their name, right? The Internet Corporation for Assignment of Names and Numbers. They also manage essentially IANA and all of these IP address assignments. And what they do is it's multi-stakeholder. So they really deal a lot with lots of different groups from government agencies, from organizations, from corporations, from registries, registrars, from security groups, like all these different parties and they have to constantly get them together and both virtually and in person and try to negotiate the operation of the internet.

29:41And so, yeah, I think that was an example of something that I nerded out on just this, these policy things. I don't know. It's just, it's a nerd. I can go to any one of these events, whether it's an event like that, where it's policy or go to an event where it's more technical and I just feel at home there. And so going back to the original questions after 15 years how can i still do this is because after 15 years it doesn't stop it hasn't stopped being interesting as technology and as just a foundational part of the internet i think that's really cool what has changed in that time because i know dns is kind of stuck in a certain degree where it is there are you know movements to make some sort of secure dns and you know adding things obviously adam said one earlier which is less encrypt came around yeah that was a huge change that was probably huge yeah it was but if we set that one aside what else is new or different in the world of dns and managing dns over the course of this business so i think if you ask anybody who operates dns in the dns space it's just the scale at which we have to operate now because the internet has just continued to grow and grow and grow and you have with the the addition of the iot devices that sort of started storming the internet 10 years ago or 15 years ago, and the growth of that, all of those things perform DNS lookups.

31:02They all have IP addresses. They all do these things that tax our systems. Plus, you have bad actors who have just gotten more and more power who know that it's a core part of the internet and they try to attack it. So I think the scale has continued to grow over time. On a technology point of view, I don't think there's been any major changes. There's been a lot. If you look at the RFCs for DNS, there's actually a website out there. I can't recall off the top of my head what its name is, but I'm sure I could find it. And they collect together all of the RFCs, the requests for comments that underpin how DNS works.

31:40And that site list, if you go down, it's continuously being updated because every year there's new RFCs for domains. Sometimes they fix old stuff. Sometimes they clarify it. Recently, you've had a new addition, which is called Service B HTTPS. And so this was an attempt to be able to do some routing things in the browser using DNS, like at the apex. So being able to say at the apex, we want to be able to terminate a name with not an A record, but with something that actually points to another name. We solved that problem years ago with the alias record, but this is actually a protocol standard way of doing it.

32:17So I think things like that are what we see in the world of DNS rather than big changes. Of course, you have people that have tried to do crypto names. That's been a thing that folks have tried. It's still being tried. By the way, I saw a new startup recently that got funded to try to do this again. It's really hard, though, to change something that's so foundational and that works really, really well. I mean, when you look at it, the DNS works fantastically well for what it needs to do. Until it doesn't, right? Until it breaks. yeah it's a scapegoat for all of our technical problems right it's always dns yeah it is so we so we hear i have a sticker somewhere around here that says that it's very cute i don't know where i put it yeah there's a fan fiction story floating around out there where the villain of the horror movie you're watching is dns the whole time that'd be a great that'd be a great twist ending you know It was DNS.

33:19And it was in the house. The whole time. DNS was forwarding inside the house. So a fairly unchanged business for the most part this 15 years. Would you say that's fairly true? I think so. Yeah. I think that from a business perspective, we have been able to stick to our core goals because the industry as a whole has pretty much stayed very similar. Yes, there's been slow improvements, but it hasn't been any big changes.

33:59Well, friends, I'm here with a good old friend of mine, Terrence Lee, cloud native architect at Heroku. So Terrence, the next gen of Heroku called Fur is coming soon. What can you say about the next generation for Heroku? Fur represents the next decade of Heroku. You know, Cedar lasted for 14 years and more. Still going. And Heroku has this history of using trees to represent ushering in new technology stacks and foundations for the platform. And so like Cedar before, which we've had for over a decade, we're thinking about fur in the same way. So if you're familiar with fur trees at all, Douglas Furs, they're known for their stability and resilience.

34:39And that's what you want for the foundation of a platform that you're going to trust your business on top of. We've used stacks to kind of usher in this new technology and what that means for FUR is we're re-platforming on top of open standards. A lot has changed over the last decade. Things like container images and OCI and Kubernetes and CloudNative, all these things have happened in this space and instead of being on our own island, we're embracing those technologies and standards that we help popularize and pulling them into our technology stack. And so that means you as a customer don't have to kind of pick or choose.

35:11So as an example, On Cedar today, we produce a proprietary tarball called Slugs. That's how you run your apps. That's how we pack to them. On Fur, we're just going to use OCI images, right? So that means that tools like Docker are part of this ecosystem that you get to use. So with our cloud-nated build packs, you can build your app locally with a tool called Pack and then run it inside Docker. And that's the same kind of basic technology stack we're going to be running in Fur. So you can run them in your platform as well. So we're providing this access to tools and things that people, developers are already using and extensibility on the platform that you haven't had before.

35:44But this, you know, sounds like a lot of change, right? And so what isn't changing? And what isn't changing is the Foroku you know and love. That's about focusing on apps and on infrastructure and focusing on developer productivity. And so you're still going to have that get push Foroku main experience. You're still going to be able to connect your applications and pipelines up to GitHub, have that Foroku flow. We're still about abstracting out the infrastructure from underneath you and allowing you as an app to help her to focus on developer productivity. Well, the next generation of Heroku is coming soon.

36:12I hope you're excited because I know a lot of us, me included, have a massive love and place in our heart for Heroku. And this next generation of Heroku sounds very promising. To learn more, go to heroku.com slash changelogpodcast and get excited about what's to come for Heroku. Once again, heroku.com slash changelogpodcast.

36:39Do you try to grow a business like that? I mean, obviously you do, right? But you don't seem like you're, and maybe I'm wrong, and this is a wrong assumption, but because you want to stick to engineers and that flow, and you resist it seeming like a registrar, which to me seems like free marketing. It's like, okay, a lot of traffic coming here to just do domain lookups and they'll probably buy from us and then buy the services. It feels like a great marketing arm you've never leveraged, really. So how do you how do you look at growth when it comes to a company that's sort of just built on DNS and provides a core service and it keeps it simple?

37:16Have you grown a ton? How do you grow? What are your thoughts on that? So we've never, aside from the very early days, we've never grown a ton. My CFO likes to call us DN reliable because the growth rate is just so kind of steady. Right. This is to say that's exactly how that's fine, because that's exactly what I wanted in a business. when I first started D &S Simple. I very, again, I made a conscious choice. I said, I don't want to take other people's money. This is not a business that's going to be a rocket ship that if we throw dollars into it, it's going to return tens of thousands of dollars.

37:54It wasn't the goal, right? For each dollar we put in. It was meant to facilitate the life that I wanted to have. I have four kids. The last one is almost ready to graduate high school. And they were kids throughout the entirety of The Unsimple. And the vast majority of time, I was able to be there with them and for them. We were able to travel together. We were able to enjoy our lives together. They had all kinds of opportunities to do things because it wasn't a business where I had to go, oh, I'm going to spend 80, 90 hours because I'm trying to help somebody else become rich. I was already rich in time.

38:33And that was an essential part of the construction of the business. We have grown, though, and we continue to grow. So the way we grow, for example, right now is that engineers are still our champions, but we found ways to work ourselves into some enterprise level contracts. And so we do that because there's a lot of enterprises that are still on legacy software or legacy registrars that are really suffering. And so we go into them and we say, we can help you. We can help you onboard. We can help move your domains off these other providers. We'll do whatever it takes to get you on board. And once you're with us, you'll have a stable system and you can have a consolidated system.

39:13And then we built other software. So one of the things we launched a year and a half, two years ago was we call our domain control plane. And the idea was that if you have some domains that you're managing at say Route 53 or in Azure, you can actually connect from DNSimple to those services and see those domains and actually edit them and have bidirectional synchronization between DNSimple, Route 53, as your core DNS on-prem. And that's cool. Like that's something we think that some of these folks that are running in legacy systems or the reality of running enterprises that you don't have just one place where you put everything.

39:49You were talking about that earlier. Like how did we end up with four or five different registrars? Well, that's just business, baby. You are gonna get spread out because as your company grows, you're going to have people off here doing something different than people off here and their needs are going to be different. So, yeah, that's how we've kind of facilitated the growth over the years, but we've never strived for massive, fast, rapid growth because it just wasn't in what I wanted to do. It wasn't, I had done a little bit of that in the previous startups. I worked for Living Social as well for a little while.

40:21So, I saw the good and the bad of trying to do that. Is that still a thing, LivingSocial? They do still exist. Both them and Groupon do still exist. Yes. That's true. Groupon was the incumbent and then LivingSocial came along and copied and pasted. Challenged. Yeah. The great thing about LivingSocial was it was actually there were a lot of really good engineers who went to work there in the heyday. And they were writing stuff in Clojure and in Ruby and I swear anything that would come like they were doing a lot of really, really interesting things. So there was this really great tech scene in space inside of living social by itself.

41:00Hundreds of engineers doing some really interesting things and they were willing to try pretty much anything and throw it at the wall and see if it stuck. So it was great fun from business. I'm not sure I thought it was great because you got to actually make a profit at some point, but Hey, there is that. Is there any cool tech inside Dan simple? I mean, I'm, I like our name servers. I'm a big Erlang fan and our name servers are the core of it's an open source Erlang name server that I wrote based and using, of course, other libraries from other folks that are much better Erlangers than I am. And so to me, that's I really love that we have our own.

41:40We basically have an umbrella app that sits on top of it. It adds some proprietary functionality, but the majority of it's serviced by the open source stuff. and I find that to be really interesting. I'm just - And that's stood the test of time? Like that's been in there since the start or when did you - 2000 and I think 14, 13, 14, somewhere around there. So it's been there, yeah, for quite a long time. It's not perfect. So there's some things we're trying to work on now. The downside having a cool tech like Erlang is very few people know how to use it and write it. And so while I can love it, one of the lessons learned, of course, Versus that I may be alone in this boat for a while.

42:20So I spent a lot of time writing Erlang code where it was just me and didn't have anybody bounce ideas off. Now I have other folks who can pick up that and run with it a little bit. That's probably how you sleep well at night, right? Because of its ability to stay online. Yeah. Tell us more about the architecture, why Erlang's cool. Go ahead and nerd out. Okay. Well, so, I mean, fundamentally DNS is cool because we can run it as an Anycast service. And this allows us to run a lot of redundant servers that are basically broadcasting the same IP address. And so that's part of what gives DNS its redundancy as a whole.

42:55And then inside of Erlang, you have Erlang was essentially built on this idea that processes. So it has built in process management. It really does want to take over the entire machine. And those processes will die fast. So the whole goal was let something die fast and then start another process if it needs to be recovered, rather than trying to keep something alive indefinitely. And so that helps a lot with its ability to recover. It's basically fail fast recovery mode, rather than trying to just keep things going indefinitely. um what i really liked about it in the dns side is that is it's it has a particular syntax for um for basically converting binary data into internal data structures they they call it the i think it's the bit string syntax and it's just so elegant like you can literally write out a packet and have it deconstructed into variables just right there in one line and i always thought that was such a beautiful, including like the number of bits that's going to come out in this particular section.

44:02I just always thought that's such an elegant and beautiful way to handle packets when you're dealing at the network level. And it makes a lot of sense. I mean, Erlang was created for telecom. It was created for telecom switches. So they needed to be able to pack and unpack things pretty quickly. Have you looked into Elixir? I have. Yeah, we actually, so I did a port of our Erlang name server from Erlang to Elixir just for fun. And we had no intention really, we weren't sure if we were going to use it, but I wanted to see if it could be done. So I did a port of that and then made that open source as well.

44:33I've written a few little web doohickeys, nothing that ever saw the light of day in production in Elixir as well. We had some other team members who were big fans of it. I actually prefer the Erlang syntax. I like the purely functional nature of Erlang and I like the prologue syntax of it a bit more. That's me personally. But at the same time, Elixir has a lot of strengths. I was actually there at the... when jose first announced elixir to the world at the emerging languages conference at strange loop this was back in i can't even remember when it was but i sat there watching him talk about elixir i was like man this is going to be really cool and it turned out to be pretty cool a lot of people really like it it's very cool yeah yeah it's funny because i looked into erling probably way back when you were starting DN Simple and took a few tutorials and I just couldn't get the syntax to stick into my brain.

45:30And I liked, for instance, the pattern matching that you're describing. A lot of the properties of it just seemed really smart. And I couldn't quite get over the hump. And then when Elixir came out, it actually, because I was a Rubyist before that and I knew Jose and everything, it all kind of connected. And so I've been writing Elixir for a long time and enjoying some of the advantages of the Beam with a language that for me is just more syntactically aligns with the way I think. But I think once you get past that, if you've written some Erlang for long enough, like it's like the matrix, you know, you don't even see the ones and zeros anymore.

46:02It took me three tries before I could actually get Erlang. And in between the second and the third try, I learned Clojure. And learning Lisp or a Lisp version was extremely helpful in getting me past that hump of what is a functional? how do I functional language you know and and that was really hard like that hurt my brain to get there but when I finally got there I was like this is actually really cool forget objects objects are dumb let's use functions everywhere I get you I get you I've been functional for a long time but you know I've been thinking in objects lately so I'm just kind of I go back and forth as the pendulum swings you know too i tease but i'm totally back and forth all the time how do you how do you choose obviously you make a choice but you know is there an angst when you make the choice you're like man i kind of miss objects or man i kind of miss functions me personally do you want to do you want to answer this chair i don't know i mean i can tell you how i choose but you're the guest anthony you tell him first i'll go next what usually when i select a language, I select the thing that is the easiest thing for me to use right now.

47:13Right. So it's purely of satisfaction of completing whatever it is I need to complete. And quite frankly, I almost always reach for Ruby first. And maybe that's because I know Ruby really well. I don't, maybe it's just because I find the Ruby syntax easy to drop back into after having not programmed for a while. But that's what I typically jump to. And I do often really feel the pain when i go in i'm like oh man like this i just it's boring to me now because i've written so much of it i guess the way i got around that though was just stopping to write code for a while and then then every time i write code i'm like oh this is so fun until i get about five minutes in and it's actually becomes hard i'm like oh this is hard yeah i used to sit in code for six to eight hours a day and now when i stream together a couple of hours i'm exhausted you know like man how do these kids do it nowadays?

48:04Of course, the correct answer, Adam, is you don't choose. You just vibe code now. You don't even have to choose. Don't even look at the code. Just let the LLM figure it out for you. I tried that the other day for the first time. I was pretty bearish on basically using generative AI for writing software. And the other day I was like, okay, I'm just on a day off, like on a weekend, I was like, I'm just going to get the editor up and try to use just, can it even run? a route like a rails project i started a new fresh rails project i said run this i want to run this rails project just run the test all right 30 minutes later i still had not run the test but it had installed a ton of software on my computer which is nice i i vibed my way into i don't know using up a bunch of space on my computer to have software because it just couldn't figure out what was wrong and i was it was hilarious i think because i have a lot of old software built up on my computer for doing things with like older versions of ruby and older versions of rails it just drove that thing crazy so at one point i just like i said to the lm i said you and me buddy it's been fun but we're done for today not today i'll see jerry he's kind we're done buddy i'm sure he didn't use those exact words he did he pulled himself he was he was scared of the future lm getting him we're vibing now but we're not vibing in the future that's right you bet you better be nice to our robot overlords that's right i was trying to get the new so i've been testing everything as as you do and the latest google what's it called gemini no yeah but it's like the latest one i don't know i'm trying to pull up zed right here it's called uh 2.5 something on something gemini 2.4 2.5 pro experimental just got crowned the king of coding lm just temporarily until the next release you know until next week and so i'm like all right gemini let's see what you can do and it's it's right in this function that i'm just i'm just like this function sucks dude i have to teach you everything this function sucks dude i'm sorry i'm sorry but i I don't have the patience.

50:19I'm not here to code review. I'm still, it's a give and a take. Some days I'm very happy. Other days I'm like, you're really bad at this. I know it's not going to be long before it's much better than I am. But for now, I still hold the high ground. So you choose. How do you choose, Jared? Did you answer that question? I mostly was joking about vibe coding. How do you choose? Yeah, you try to pick the right tool for whatever job you're doing. So I agree with Anthony on that. That being said, sometimes the right tool is the one you know the best. even if it's not exactly the right tool for the particular use.

50:51And so I will just pick things that are general purpose. You know, I like languages that are multi-paradigm. I like being able to do functional and object-oriented. And then I can kind of flow in and out based on what I'm up to. But I'm not going to lie. Also, it's just like personal style trends. You know, you do functional for long enough and you get sick of just passing these bags of data around. And you're like, I wish this object was smarter than it is. Just a struct effectively. and so you start going a little more object oriented and throughout my career i've changed weight the way i think code should look the way i write it you know i haven't changed from tabs to spaces but i'm sure eventually i will i could and change back if i could see him more often i probably would i just don't have any sort of consistency even as a human being so uh it's not logical adam it just feels i decide based on how i feel and that's about as true of an answer I can give you.

51:44That's good. I don't know if it's good or not, but that's how it works. Well, you can live in honest. Well, so I think Anthony to pull back one layer, Jared will have to live in a world soon where he manages a Phoenix application, which is Elixir, and then a Rails application, which is Ruby. And so that's obviously going to be different worlds, but similar worlds where you have objects and you've got functions. and you know I suppose you have to be left brain and right brain to do that because that's what's required. Context switching between languages is a skill. There's no doubt about it and it's something you have to practice independent from writing software in general.

52:28We run three key languages like our key core languages at the N-Simple are Ruby, Go and Erlang and the three of them communicate with each other every day so Ruby is our the Rails front end. Go is kind of the glue that does messaging and shifts data from our core to our edges. And then Erlang that's running on the edges to do the name service. And so we have to be able to run that entire stack. And if you really want to know how to get data from the customer's hands all the way out to the name service, the edge, you have to know all three of those languages. So it forces the team to, one, operate all three of those languages.

53:11which operate very differently. And two, actually be able to get in and understand and then make changes to those languages, which also forces the team to ensure that each one of them along the way has the appropriate guardrails through tests, so on and so forth, to make sure that they don't shoot themselves in the foot. And frankly, we haven't always been great at putting in the test cases needed to ensure we don't shoot ourselves in the foot. So we pointed gun and pulled trigger and it happens. And then we have to go play cleanup. But it's definitely a skill. Like context switching is absolutely a skill.

53:41How does that work out then with you having that many, I guess those three core languages, you said key languages? Yeah, we could say essentially languages that have enough use at the end simple that an engineer needs to know them really if they're going to run the whole stack. The answer is most of the engineers spend most of their time in Ruby. Like the vast majority of changes, the Ruby code, what the customers see changes a lot faster. so it has the highest rate of change, then the other two languages have a much slower rate of change. And so the reality is they don't get constant practice.

54:15And then now what we're trying to do is get it so that they can ensure that locally they can stand up the whole stack so that they can get practice with each one of those things and make changes with each one of those things. If an engineer needs to do it every once in a while, most good engineers will be comfortable switching languages. um they'll they'll they'll they'll be able to pick it up it's it won't be instantaneous they're not going to write the most idiomatic code but they will be able to pick it up and do what they need at the most basic level with it and especially if there's good tests that they can run and ensure that whatever they're doing doesn't break things horribly what do you all look like operationally in terms of observability and deployment, ops, et cetera?

55:00Is it complicated? Is it simple? Do you sleep at night? Oh, I sleep very well at night. Who gets pinged? Are you on pager duty? So that's a lot of questions. Let's go. Let's start. So I gave up the right to be on call when I stopped writing production code a couple years back. And so that's good from my point of view, at least. It doesn't mean that I don't pay attention. Anytime that there's an incident, and we have very well-documented incident handling policies. We have one person from the engineering team and that rotates around the whole engineering team that's always on call. And then that person, if an incident starts occurring, they can, they basically, we have instructions, fire up a, we have an incident channel in Slack, fire up a Google Meet.

55:43Get whoever's online right now, triage the incident. If you can't triage it, then let it ping up everybody and all engineers who are available or who are woken up can come in and try to fix it. So we recently switched everything for deployment over to Nomad, which is HashiCorp's technology for deploying containers essentially out in a well-orchestrated fashion. We also finally bit the bullet last year and the year before it started containerizing everything. So now we have pretty much everything containerized as well. We deploy using chat ops. So you can do all that type of stuff directly from within Slack.

56:23You can query our, we have tools, for example, to determine if particular domains are being hit harder or what have you, get stats on things. And you can do that from chat ops as well. And then for observability, we've run everything through Datadog for many, many years. And so we have the ability to use Datadog to look at logs, to look at metrics. We don't do APM level things because it's very expensive. There's a large cost associated with it. And so we do a lot of our analysis off of logs and off of just general metrics that we keep to determine. Because the failure cases, most of them are fairly well known now.

57:02And so we can alert based on commonly known failure cases for our systems. How many of your headaches are due to bad actors? Are you at a scale that you have a lot of issues with that? We don't have a lot of issues, but I think most of our, many of our headaches do come from bad actors, intentional or unintentional. Unfortunately, I think that there's, that there are still folks out there that try to weaponize technology and use it to, to whatever ends they want to reach. And that's unfortunate because the internet was not originally designed for that. It was designed, it was a, it was a friendlier place in its beginning.

57:35To share with your colleagues at university. Exactly. got any war stories or like anything in particular where like the things were under attack or yeah people abused you wouldn't expect them to unexpected abuse we've had a few so for example one of the unexpected ones was the recent rise of takeovers so what happens is is that if you register a domain and you point it at a dns provider and then that domain gets deregistered, let's say you let it expire, and then somebody else picks it up. what they can do is they can go back to that same DNS provider, create an account, and then they can basically say, oh, well, that domain is still pointing at it.

58:21They could basically not even let it expire. Let's say you still control it, but you haven't done anything, but you let that account lapse and it's still pointing at the service. They can go to that service provider and spin up the DNS for it and basically do a full takeover because now they control the DNS, it's pointed at it, and they'll come in and they'll say, okay, now route email here. And now they actually can start seeing what email is being sent to that or what web traffic or whatever traffic they want to see. Um, and so that's one that I didn't see coming, but actually is something that is fairly just like heavily discussed right now in the world of domains and DNS.

58:58So it's just this whole DNS takeover concept. Yeah. I mean, it seems like a valid use though. Like, is there a way to actually stop it? Like, how do they, are they, how are they gaining the privileges in the first place if they don't? Well, if you, if you're, if you're not going to use a domain, then de-delegate it. Don't leave it pointed at a service that you're not going to keep live. Right. Like if you're delegating it to a DNS provider like us and you say, I don't want that domain to be used there anymore, then de-delegate it. Just take off the delegation completely or you can delegate it often to a sinkhole, basically someplace which just goes and disappears, usually from the registrar.

59:35Or just turn it off completely. That's the right thing to do. You shouldn't leave things just, it's kind of like if you leave the world not cleaned up. So think of software programming. I mean, if you leave memory and you're writing in a thing where you have to manage memory and you don't clean it up, you're going to pay the price for that at some point. Makes sense. I guess you just assume, you know, maybe I started a business, got a domain for it. You know, the business eventually failed and you're like, I'm done with that thing. I'm just going to let it expire. And you just let it expire. That probably happens a lot, I would think.

1:00:06Yeah, I think if you let it expire and you don't care about it, that's maybe a little different. This is the case where folks hold on to a domain and they're not letting it expire. That's why I corrected myself earlier. I want to say it's not really the expiration part. They're still holding the domain, but they've left it delegated somewhere else. I see. So they still hold it. Okay. What was an example of that? There was one recently that was, it was actually, I wish I could remember what it was. It was a pretty high profile case. I think it was for a registry for a, one of the top level domains had something set up that they were using and then they stopped using it, but they forgot to remove the delegations.

1:00:45And essentially somebody started seeing traffic flowing through that was pretty useful for them to attack in another path. So it's one of those risks where a lot of things have to kind of come together and create that perfect storm situation where it can be dangerous to you, but it happens. Like that's the point. It happens. So you have to protect against it. Mm-hmm. So we're making changes, for example, on our side, when we have a domain that used to be with us, maybe, and then somebody says, oh, well, I want to own that domain or whatever, and I'm bringing it back into DNSimple. They have to basically make a proof and show that they actually have the rights to do that.

1:01:30It's a rare case when it happens, but it does happen. So that's an example of an attack that I didn't really consider. We actually have considered a lot of the more human attacks. So we have really strict protocols about how to deal with folks requesting changes to a domain that we can't necessarily prove that they own, which is we don't make those changes. Right. And that happens. We get a fairly regular stream of people trying to say, no, no, I need to change this. Oh, another good one is when I've got stories here. Another good one is when you have two people at a startup, for example, two founders, both using shared access or both having access to an account and they have a falling out.

1:02:11All of a sudden, one founder locks out the other one and you get you as a domain provider get stuck in the middle of this fight. And you're just like, wait a minute, you just stop people. i'm not gonna lie i am logged in as adam at changeout.com right now we actually just had this conversation not this particular one but like how do we even log into dns simple because we i mean it's been a couple years for me jared logs in more frequently and i'm like do we have separate users do we have i hope you do accounts we don't know separate you should we do not we have a shared one password boo there's no excuse for that come on uh what's our excuse adam honestly i just never thought about it um we probably pre-exist multiple users maybe back when we set it up that's actually what i would be if i was thinking i since you were there so early my guess is is that you set it up that way and you just you left it and it's always worked for you so why are you gonna change it right you have 2fa turned on all right that's good that's good so and we've never had a falling out so not yet That's right.

1:03:20Come on, Anthony. You're making this long. Come on now. What are you trying to say here? He's trying to divide us over here. He is trying to divide us now. I would never do such a thing. Yes. I just want you to be aware of the risks. Okay. Oh, the risks. Who wins? If we come to you. Yeah, how does it work? How do you even arbitrage? Arbitrage? No, mitigate. How do you mitigate that? So the thing is, you just said you logged in with Adam's email. I wouldn't tell you that. Oh, you already know that. Nah, it's tough. It's tough. If people get in a fight like that, often we say, look, you need to resolve this yourselves.

1:03:58What we're going to do is we're going to make it so that domain can't move. So we'll lock it. We'll make it so nobody can make further changes to it. You're not going to take it anywhere, but you need to resolve this amongst yourselves by whatever means and agree upon it. And then we can go from there. So do you hear from an attorney? Do you hear from like law enforcement? sometimes yeah and so obviously if they're of any legal precedence whether it's an attorney or law enforcement and they actually make a change you weigh that against its truthfulness and make the change if it's we didn't we generally just don't make the change we say look you you still have access to this thing so we if they say we have a court order or we have an agreement you know here's a letter from an attorney where it's agreed upon and it's and and it's both parties agree, then we will let it move forward generally because we can't stop.

1:04:50We don't want to stop it indefinitely. We understand that people need to get back to whatever it is they're doing, but we, it's not the kind of thing we definitely dislike doing it. I'd rather not get in those situations. I'd rather that folks resolve their stuff amicably, but I know that's not always possible. I was just thinking in that scenario, very particularly, the one thing you would be asked to change would be to either change it back to the original email if it was changed. or, oh, hey, I used XYZ co-founders email for a long time and I've never had my own. So therefore, when they changed it and I don't have access to that anymore, I don't have access to the service.

1:05:28So can you create me my own user? Those are the kind of changes I can think you'd be asked to make versus like move my DNS from here to there kind of thing. Yeah, it's going to be a real tough sell to even get us to do that, because essentially whatever condition the account was in before the incident happened, And that's what we're going to be looking at as the baseline. Because you operated it that way for a long period of time. So that's going to be what we're going to say is like, okay, that was the status quo. Then things started to change. And we've seen this happen. We've seen then they'll change emails and then they'll change it to somebody over here.

1:05:59And then they'll change it to somebody over here. And unwinding that stuff is a nightmare. But it's something that we've had to deal with a few times. So you keep logs. We keep logs on. You're not deleting these records. You're just safe deleting. You're not literally remembering. We keep activity logs for pretty much all those types of actions to ensure that we have a history of it. And those logs don't go away if the account does or the user does. Is it like a separate audit log or is it just like how do you describe it? Yeah, right now we have an activity log for all accounts basically. Each account has its own activity log.

1:06:34Every domain has an activity log. So we keep tracks of the kinds of changes that are happening on those different entities. how frequent is this a a concern for someone is there someone right now emailing you saying hey probably not no it's probably more like it's probably once or twice a year we'll have some sort of contested case like this so again it's not a huge but we don't operate it's imagine now if you're operating to scale it at a company that's running 150 million domains yeah that's a a massive amount of potential, let's just say, vectors of attack, areas where security becomes a real issue, right?

1:07:15And I can't even imagine the kinds of systems they have to have in place to keep from doing the wrong thing. Well, if your marketing is correct, I believe somewhere it said like 200 ,000 or something like that. In terms of domains under management? Yeah, I believe it was something like that. I forget where I saw that at. I think today we have close to that 200K registered domains. And I think we have a total of like 600 ,000 zones under management. So, because again, we have the operational DNS side, we have the registrar side. Sometimes those two match up, sometimes they do not. Some folks only register domains with us, but use somebody else for DNS.

1:07:51More often, they just use us for DNS and don't register domains with us. Do you also sell certs? We do. So both Let's Encrypt, which we don't really sell. If you're on the level of account, you can get the type of Let's Encrypt certificate that you're allowed to. And then we also sell commercial certificates from Sectigo, I think is what they're calling themselves right now. And really the key difference is the Let's Encrypt ones are all automated and they have a really short 90-day window. And the other ones are like one year, but a lot less automation. So I imagine Let's Encrypt was a disruptor for many people selling certs.

1:08:28Did it affect your bottom line where you're selling way more certs and then it kind of dropped off or not a big deal? It wasn't a huge deal because we never sold a lot of certs. We didn't really position ourselves as a company selling certificates because it was just a side. It was kind of like how domain registrations were in the beginning. It was, okay, we'll add this functionality to support our operational folks who want to manage their DNS, but also want to have their certificates as part of that. Let's Encrypt actually was a good thing for us because it really fit into our thoughts about automation and how to automate all this infrastructure.

1:08:59And so for us, we loved it. we're still huge fans we've been we've been supporters financial supporters of let's encrypt now for i think 10 years since its inception really since early days we are also fans and have helped them get their message out there over the years had them on the show multiple times and have just been beating the drum for years you know finally somebody came around and solved one of the biggest gaps in the internet security space and they made it too easy to say no and actually folks like you also integrating it into the web tooling and stuff so it's just a click of a button versus you know downloading the cli bot and running it yourself and like cronning up your renewals because that's really the big pain with what less encrypt is the 90-day renewal there's got to be that process and so we all had to you know run it on our remote machines but um so the infrastructure is there now and they sure changed the they changed the world of the internet absolutely um and it's cool it is very cool what is the sad story i am i am very happy that they came into existence and they've done such a good job of stewarding this certificate provider certificate authority i think it's been really fantastic to see

1:10:22Well, friends, I'm here with a brand new friend of mine, Kyle Galbraith, co-founder and CEO of depot.dev. Your builds don't have to be slow. You know that, right? Build faster, waste less time, accelerate Docker image builds, get up action builds, and so much more. So Kyle, we're in the hallway at our favorite tech conference and we're talking. How do you describe depot to me? Depot is a build acceleration platform. The reason we went and built it is because we got so fed up and annoyed with slow builds for Docker image builds, GitHub Action Runners. And so we're relentlessly focused on accelerating builds.

1:10:57Today, we can make a container image build up to 40 times faster. We can make a GitHub Action Runner up to 10 times faster. We just rolled out Depot Cache. We essentially bring all of the cache architecture that backs both GitHub Actions and our container image build product. and we open it up to other build tools like Bazel and Turbo Repo, SC Cache, Radel, things like that. So now we're starting to accelerate more generic types of builds and make those three to five times faster as well. And so in simple terms, the way you can think about Depot is it's a build acceleration platform to save you hundreds of hours of build time per week, no matter whether that's build time that happens locally, that's build time that happens in a CI environment.

1:11:38We fundamentally believe that the future we want to build is a future where builds are effectively near instant, no matter what the build is. We want to get there by effectively rethinking what a build is and turn this paradigm on its head and say, hey, a build can actually be fast and consistently fast all the time. if we build out the compute and the services around that build to actually make it fast and prioritize performance as a top-level entity rather than an afterthought. Yes. Okay, friends. Save your time, get faster builds with Depot, Docker builds, faster GitHub action runners, and distributed remote caching for Bazel, Go, Gradle, Turbo Repo, and more.

1:12:23Depot is on a mission to give you back your dev time and help you get faster build times with a one-line code change. Learn more at depot.dev. Get started with a seven-day free trial. No credit card required. Again, depot.dev.

1:12:43So what's burgeoning? What's new? What's next? Can you share some of these TLDs, these secret new ones? Are they out there? Are they published? No, it's too early in the process. All of us who are interested are waiting for the handbook, which is going to be this document that's going to explain the process for how to get a new tail i don't suspect we'll see anything until 2026 probably it's too early for that but certainly there's things coming down the pipeline that ican's talking about or that the dns luminaries are trying to get done yeah what's uh what's out there so they're in the middle of trying to make improvements around the transfer process to transfer domains from one registrar to another that's still kind of tedious in many cases, especially when you do it at bulk.

1:13:28So there's a whole set of guidance coming out and policy changes that are going to be worked through in this next year that are really about making that transfer process a little simpler, a little bit easier to start and finish, and ideally to be able to do that same thing even if you're dealing with large portfolios of domains. That's, I think, one of the big things coming of ICANN. There's always a lot of talk about security. So what's next on it? So DNS abuse is a big thing right now. So there's a lot of push to rein in the use of domains for things like phishing and for copyright infringement and for all kinds of other nasty things that happen on the internet.

1:14:14And that has to be a global effort because if just one TLD works on it, it gets shifted somewhere else. if only one country works on it, it gets shifted somewhere else. So it has to really be a global effort. So there's a lot of talk about DNS abuse and improving policies. The Whois protocol got basically deprecated. Yeah, it's gone now. What's the new one called? RDAP. RDAP. Yeah. And why? I mean, Whois is such a cool word. It was a cool word. But as a protocol, it was pretty basic. It was just like, here's a chunk of text. Exactly. Right. And that chunk of text was never standardized. It was like pseudo standardized by the biggest players.

1:14:56You just get good practice at your regular expressions, you know, extract the important parts, please. So RDAP basically said, no, no, we need more structured data. We need access control. So, for example, certain levels of access need to be given to registrars that don't need to be given to consumers generally. Certain access needs to be given to registries that maybe registrars don't have access to. certain access needs to be given to a law enforcement right so there's all kinds of things that have to happen and i think that was one of the major changes that kind of evolved over the last seven or eight years yeah and our apps been out there i read it just it's just now they're actually deprecating who is like they kind of were available correct simultaneously for a while and just people didn't know about it and now they're starting to actually put it out there.

1:15:45Stop using who is start using RDAP. Yep. Yep. That's kind of how things work when you're talking about having to, to move from some legacy protocol to what we're going to use. It has to be a really slow conversion. Certainly we're definitely like, that's something we know very well when we make changes, often it will take us a year, two years to fully realize that change just because there's so many factors involved and we don't want to break things for our customers. speaking of ipv6 is that is that a big thing nowadays yet i mean it's definitely grown in usage i know that the last so we run ipv6 internally for a lot of the routing uh from from our edge services we are our architecture basically has edge caches that sit in front of our origin servers for dns and so we run ipv6 internally for a lot of that um we did find and that in some areas, IPv6 isn't supported well enough to be able to route it correctly every time.

1:16:48So it's just like, we would, latency would become an issue because it was being routed through so many different routers to try to find a path. Interesting. Or there just wasn't a path. Like there was just, there was just no path. Hey, like there's not, there's no router. There's gotta be a path eventually, right? Not necessarily that there's, if there's no, there were certain cases where there was no way to route. And I don't know if it was temporary or or what have you, but it became an issue. But I think we're on the cusp of needing to have it. So I expect at some point we'll see it come the de facto standard for new addresses.

1:17:23Well, I was in college in 2004. I recall my teacher saying IPv6 is rolling out, baby. And I was like, what's this? Because they were teaching us IPv4. And I was like, and they're like, this knowledge is going to be useless soon because everything's going to be going to IPv6. And so I was very much expecting it. And I'm still using those four, you know, the IPv4 addresses. And I think a lot of the adoption is, okay, obviously NAT trying to change things because they didn't see NAT coming where everybody would be behind a shared IP, you know, in a local network. And that really delayed it. I think the other thing that honestly has delayed it so long is like, dude, those addresses are just too long.

1:18:06Like, we don't like them, right? I mean, is that too basic? I can't remember IPv6. I mean, you can put it in front of me 17 times. You can't remember that. Yeah, but really, should you be remembering IPv4 addresses either? Come on. Well, how are you going to SSH into your Plex server, you know? You're hurting me right now. Well, how do you do it? The truth hurts, man. The truth hurts. Don't SSH into my Plex server. Okay. Well, I do think there are sometimes funny, weird explanations for why things do and don't take off. I think PHP is a great example of something that took off for a reason that nobody would have planned, which is like dead simple execution.

1:18:55I mean, really. Deployment. Like PHP deployment strategy was FTP to a server. That's right. Change the.html to.php and now it's dynamic. I mean, that is why PHP is what it is. Language itself, all kinds of problems, et cetera. It's better now, I know. No shame, PHP people. But I honestly think that a lot of the IPv6 holdout is because it's just ugly and we don't like it. And so I'm going to have to be dragged to change it. Is it the colon in there? What is it that gets you? What is it that makes it, like, describe ugly, Jared. visually with your hands and everything just look at it i'll just look at it it's like uh that judge who says you know describe porn on the internet you know when you see it it's like that you know what everybody knows what ugly is and obviously it's subjective but i don't hear anybody disagreeing with me when i say it's ugly they're like yeah ibb4 is better if you just just like your robot overlords you need to accept your hexadecimal overlords okay yeah well i mean i'll accept it when they force me to you know just like with the robots i'm not gonna accept them until they make me.

1:20:04I just see it as an opaque thing and ideally I want to automate it all the way. So in my ideal world as devices inside of a managed network come on they should just automatically get names based on whatever your configuration management has set up. So if you're using whether it's a configuration management like Terraform or something, Ansible, whatever it is, that should just that should all happen together. I don't have to think about this stuff. And when we get to that point, then IPv6, IPv4 you don't care as much about anymore. yeah i mean it should be an implementation detail that lives at layer three or whatever and i should never have to look at layer three i completely agree with you but we still have to every once in a while because how else are you going to ssh to your plex server adam you know the way i do it is you use local dns don't you can i just introduce you to a wire guard protocol please and stop doing what you're doing okay let's talk about that then because that's uh what replaced it for me so for a while there i would assign particular ip addresses to particular machines now this is home lab stuff so this is not enterprise this is home lab yeah we're just watching flex yeah the rule book's out i mean you're not i mean unless you're like home labbing to be enterprise and that's a different story than rule backs rule books back in but in my case i would have certain machines i would give a certain ip address to and i would remember okay uh you know this subnet is dot 100 versus whatever you know whatever it might be and so i would remember the IP address pretty easily because it's the IP address for the home.

1:21:32And then it's just, you know, is it, you know, the last three, where is it at in this grand scheme of things, in this network? And now because I use Tailscale and because Tailscale is really installed on everything, it gives it a network-friendly name that is accessible via search. And so when you set up Tailscale for every new machine, when you do SSH blah, it's going to search on the Tailnet for a machine named that so if i ssh into cine cineplex for example which is my plex machine it's named it now has a name on the tail net and so sshing into it is just like really just too easy now do you know his ip address ip address yeah i don't have none of that stuff i mean if i wanted to uh you could memorize it if you wanted to well no to go one layer deeper like my raycast uh setup plugs right into tailscale so if i needed to know which ip address was anything i could just use raycast to conjure which machine via tailscale and just copy it to my clipboard and paste it wherever so i don't ever have to remember anything in or even look it up it just comes to me it's like neo in the matrix you know i know kung fu yeah what do you think i love it anthony does that pass your test is that wire guard enough for you yeah i feel that's good enough thank you good job he's not literally typing in i come from the days where he used to put all of our local machines in our etsy hosts you know so you know i'm not gonna put ee colon 1f colon 92 colon colon you know it's not gonna just get out of my face with that stuff okay you're supposed to be shaking your fist at the cloud right now while you're doing this jared come on I'm just trying to explain how simple human reactions are sometimes.

1:23:18And I hope that it someday occurs because I agree with you. We should not have to think about this stuff. And most of us don't have to most of the time. I haven't actually SSH and a Plex server probably ever. I was representing Adam. Yeah. Okay. Where do we go from here? The future. We're looking at the future. We're looking at the future. So LLMs and Gen AIs on a lot of people's minds, obviously. How can you use that with DNS? Is that a possibility? Well, I mean, the generative part of it is not so interesting. The part of it of machine learning from the amount of data that we get could be interesting, but we don't have enough capacity in terms of, as a company, we're small enough where we haven't said, oh, well, let's apply machine learning to all the data coming through.

1:24:02But I think it's feasible. And our team is looking into how they can use it to do their job better without putting things at risk. So that's definitely something that at least it's on our mind. It's one of our objectives this year is to figure out the place of generative AI at the end simple, if it even has a place other than for internal use. It's the right question. I mean, the answer could be no, but the answer could be no. As they become more capable, it probably fits in somewhere. But I think, for example, an example of something that just, that when I finally started using it a little bit more that I said, okay, this is an interesting use case, is we implemented the ability to export your domain lists recently as a CSV file.

1:24:45Seems like a silly thing, right? Yeah. It should be pretty straightforward. But what happens if you want to just search through your domains and you want to filter based on things? You want to have rule sets. And those rule sets, you want to have the flexibility to be able to not have to actually set up all the interface for every field that you might want to search on and all that. So traditionally, you might use a search, the off-the-shelf search tooling. But frankly, generative AI might be a place where that could be used. And more importantly, it might be able to take the context of the rest of the internet and be able to use that at the same time.

1:25:24Right. So this is an area where I feel like it could be interesting, but we're definitely still at the let's make sure we have a safe use policy. Let's make sure we're not going to put any customer data at risk before we even start trying to implementation. Because there is so much there's so many unknowns and it's moving so fast. So, so fast. Llama 4 just dropped or is dropping, which has a context window of, I believe, 10 million tokens. And so that's making things like what you just described way easier because you don't have to do any sort of fine-tuning. You don't have to do RAG. You can literally, obviously it's pre-trained on the internet, and you can drop so much information into the prompt and just go from there.

1:26:06A good example of what one of our team members did as an experiment was he was looking at a list of domains and he said, just some eight and like count the number of top level domains that are in this. Now, what's interesting about top level domains is they're not always one TLD. Sometimes they could be considered a T, like they're considered top level. For example, co.uk. UK is the top level domain, but a lot of people consider CO UK. So now you have the context of, do you want it with that or without that? And he wrote a Ruby script initially to do all this filtering. And then he said, I wonder if I could use a mod, like a gen AI to ask it to do this same type of stuff for me.

1:26:52And it understood the context of what the internet thought as top-level domains and helped him do that far faster. So that kind of opened my mind to the potential and the possibility for using generative AI. Not looking, like leaving aside for a second all the challenges with it. the potential is there to do something super interesting where i would see the use of an llm might be in configuring or hypothetically configuring something so maybe reconfiguring my dns or maybe adding a new service rather than clicking the one click maybe it's you know a prompt box that says i want to add support for this service and then the ui comes back with this is the change i can propose to make to your dns this is what's going to be added this is what's going to be subtracted etc where it's more of a natural language driven option to a user potentially i don't know i mean you got a lot of nerds and engineers maybe they want that less but sometimes like especially as you get an enterprise maybe you need to offer something that's a bit more just flexible maybe natural language you know for configuration i i have no doubt that we'll see tooling around ops that's going to be that's going to have to be context aware of the entire corporate network and then we'll use that context awareness to alter like to suggest where configuration should go where improvement should go when new systems get added it can basically just go this system goes here no need to write you know code to configure this whatever that looks like because it'll just do that for it.

1:28:30And initial will be in your editor. Like the engineers, that's what they have right now. Think about it, right? Like their editor has this, this thing that's helping, helping guide them through this. And if you take away the syntax of the editor, but still have some visibility into what that model is going to do before you apply it, I think you could take away a lot of the complexity of managing some of those enterprise networks. So I imagine we will see services starting to take advantage of this pretty soon. until it's going to happen. And there's going to be some very funny and not so funny incidents that are going to be due to choices made to use this type of tool as well.

1:29:09Beyond that, I'm trying to think if there's anything super interesting I see coming out. I mean, the internet's kind of due for a shakeup, right? Cryptocurrency has become normalized now, right? Crypto is normal. It's traded on the stock exchange under certain funds, right? That's right. There's actually ETFs where you can buy Bitcoin and ETH, I believe. So what's the next shake up? Is it a shake up? Is this it? Is this where we're at? Or is it something that we haven't seen coming yet? I don't know. I love being a futurist sometimes, but I'm not very good at it. Well, it's hard because none of us know the future.

1:29:46Well, something obvious here, but none of us know the future, by the way. Yeah. That is pretty funny. I was still thinking about interesting uses of LLMs on your thing. Where I could even describe what I'm trying to do and domain names get suggested. Yeah, I actually had a friend who wrote, who built something like that, who did it. So the funny part about suggesting domain names isn't coming up with the words to put together, to string together and suggesting. It's actually determining if they're available. Yeah, if they're available, yeah. There's really, really strict limitations at most registries about how many times you can send a query, is this available?

1:30:25like in some cases absurdly small limitations so yeah for example there are registries who let you open two synchronous connections for your entire and that's it like you oh my goodness you have to somehow fit everything you're doing through two and it's like how because and and they're stateful connections we're not talking stateless freaking two stateful so one of the things that's being developed right now is a new version of the protocol that registrars use to connect to registries based on actual REST and state transfer through the entire packet so that you don't have to run in these situations where you have to keep a long-lived connection open because it just doesn't scale up to the level that people need to operate at.

1:31:11When that happens and when that starts getting deployed at some of the registries and registrars, I have a feeling that it'll be a lot easier to add search in because it's not the it's not the suggestion part it's the checking part that's really really hard another tidbit from working in this industry for far too long that's interesting i wouldn't have thought about that it's just yeah old school tech holding you back you know yeah yeah you'd think it would have been fixed by now but nope but no yeah just all the incentives aren't there you know i know that you're enjoying this chill biz you got going on here i'm not saying that a domain lookup would make it any less chill but you know that's the area where you have not wanted to explore to keep things the way they have been but i just wonder if there's i hear you adam i hear you i see you adam i've got the message has been received i think i know what adam wants me to work on next i'm just just trying to hypothesize here on a podcast you know just just saying you know best time to fodder it out let's do it right here i'll make a deal if If your podcast listeners flood me with requests to do this, I will know that they will agree with you.

1:32:17And I can tell the team, look, we have an incentive. Look at all these changelog listeners who really want us to do this. Well, I mean, I think our dollar spent alone should be some indicator, right? I mean, we've been a diehard customer for years and none of that money has gone to you. And I would probably venture to say, I haven't done the math, but probably thousands. thousands spent of which i would maybe see five to ten percent of that because i'm giving the rest of it to the registries this is the part about that that part of the business it's kind of lopsided all the innovation is happening at the edges at the registrars and all the money is getting sent to the core at the registries it's kind of an it's it doesn't really create a great environment for innovation and it's one of the reasons why we continue to see stagnation, I think, in it is because the money's going to the wrong place.

1:33:09How do you become a registry? If you want to become a technical registry operator, there's some pretty stringent requirements that you have to fulfill with ICANN to get permission to do that. And then it's just like in other business, you have to convince TLD holders, so the registry operators are the ones actually on the rights of TLD, to come switch to you. So you have to prove that you have the stability to do this and you have to be involved in all the things at the at the the policy level at the technology level to prove that you have that you can actually stand up to what they have to deal with that is kind of a whole different business though that you're running though i mean it's it's not it's there it's it's not it's like now it's customer support in a lot of cases on a technical level it's not that different though it is because the it's it's different because it's even deeper at the core.

1:34:00So in other words, if I'm going to go, who do you think gets attacked most often? It's going to actually be the people operating the registries because if you take them down, you can take the entire entire team. And some people just want to watch the world burn. Right. So they, they will specifically, they, they're under attack constantly every day, all day. I'm sure all the registry operators are, that's the area where they have to focus. And that's, that's what makes it, I, I just a little bit about the money's going to the wrong place. The money's going where it goes because there is a lot of responsibility that they have to continue the operation of the internet.

1:34:34And it's, they're, like the buck stops with them on that TLD. If something goes down, it impacts a large swath of the internet. For any TLD, even the smaller TLDs, a lot of people would be impacted. Just imagine for a second running the.com top level domain. Just imagine what it entails. It blows my mind every time I even start to think about it because it's so fundamental to how the world works today. Yeah. You take it down, you take it all down. Oh, yeah. And imagine, like, none of what we do is possible. Everything would stop. Maybe the better question for me to ask, rather than trying to force you into becoming a place I go look up domains and then tell you how AI applies to it and, you know, attract you to it, maybe the better angle is or a different angle really is is how do you market seemingly boring domain services that are very stable as you said your co-founder said dn stable not dn simple or dns stable kind of thing like how do you market this bespoke kind of unique you know feature sets you offer to the world effectively because i would feel like the the one thing would be great marketing, but maybe, you know, maybe the loss there is, is not good enough.

1:35:53You know, how do you market it? So we, again, we focus a lot on positioning and how we, we use our words. We're still very, we get a lot of benefit from things like SEO. We've, we followed along as the trends have started to change and make sure that we're good at optimizing for a gen AI tooling. So a lot of people are starting to get their search results, if you will, from ChatGPT or from Perplexity or what have you. And those require a different set of basically things you feed into them to ensure that you get placement. And so we've been doing things like that and making sure that our materials are both instructional and accessible and machine accessible as well.

1:36:44And I think that's one of the ways that we've done it from a technical level. But really, it comes down to positioning who we want to go after and then making sure that our website, our messaging when we're out at events, the way team members talk with other folks, the way the community talks about us, that those things kind of all come together to guide those folks who they're having a specific problem that we solve to come to us when there are some really big alternatives out there. Yeah. So yeah, I mean, I know how we do it. I'm not going to say we do it exceptionally well, but we do it well enough to stay in business and continue that slow growth.

1:37:20But the competition is fierce on both sides, on the operational side and on the domain side, but especially on the domain registration side. If you're into looking at keyword prices and things like that on advertising, go look at some of the keywords around domain registration. They're some of the most expensive keywords in the world because the competition is just fierce. And those companies buying Super Bowl ads, you know, they're competing at that level. We can't compete at that level. So we have to instead compete at a level that's more appropriate for us. What you do is you find out where they're filming the Super Bowl ad.

1:37:55And somehow you put a dude or a person in the behind the scenes with a DN Simple shirt on and then you get on. Yeah, you know, that's how you get in there. That would be hilarious. You know, that's how you get in there. That's one of the reasons why we give out a lot of swag is we're hoping we'll have like an accidental photo bomb in somebody else's advertising one of these days. It hasn't happened yet, I don't think, but I'm hoping. One day. Send me some swag. I'll wear it on the show. That's right. Easy button. You ask, I will give. All right. Two shirts, please. Two shirts, please. I just need to know your size.

1:38:26So someone have emailed the sizes afterwards so we know what you're going to be. Well, too easy. Yeah, I was even trying to Google like how in the world. I mean, this is the one that I probably Google. How do I find DNSimple to know I need DNSimple? Is it DNS hosting? Is it managed DNS? Like what exactly is, I mean, I've been a user for so long. I don't even know what you are, basically. I don't even know what you are. Yeah, I think a lot of folks do look for DNS hosting. Yeah. I think that's a big one. I think a lot of people are looking for a solution with a good API. And we do fairly well there.

1:39:04reasons because we've, as engineers, we've focused a lot on building an API that is reliable, easy to understand, well documented. And I think that brings in a decent amount of traffic as well. Even if those folks don't use the API, a lot of engineers just like to know it's there so that the day they do need to use it, they can. But I hear you, Adam. I feel you. I see you. He's going to go work on this for sure. I really don't know if I'm on the fence. I don't know if it would improve your business based on what I know about it. I don't know if it would probably just be more headaches. I would, if it were me, the way I would consider it isn't because it was my dream to host the place someone goes when they have an idea to figure out which domain to buy.

1:39:50That would be cool as a standalone thing. But I think I would do it if I were in your shoes, if it raised the bar of my awareness. And even if I only broke even on the registration thing, does it help me, you know, address a bigger market? Because more come more people come to me and there's lots of enterprise out there and you've already crossed that chasm from engineer to enterprise. So you're looking to get other customer types in there. To me, that would be one way you do it. Just become, you know, more. It just drives more awareness. I and I see where you're coming from with that. I think there's, as you pointed out, there are a lot of things that happen when you open up to a different audience that are unforeseen consequences of any choice like that.

1:40:39So, and we have to weigh those choices against what we want, our goals are as a business. And that's kind of like, we're constantly doing that. We're constantly saying, if we make this choice, does it have, is it going to be overall positive, negative? Do we have any idea? And we run a lot of experiments and we'll probably keep running a lot of experiments as well just to see if we can change things in a positive direction. So, like I said, I hear you. You hear me. You hear me. He hears you and he sees you. What is the coolest thing you've learned about, I suppose, in the world that because you operate this managed, hosted DNS, like what are some of the coolest stories that you would have never had if you didn't run this business?

1:41:19Yes. That's an interesting question. There are stories about characters in this business who have built up these massive domain portfolios and sold them. That's just so weird. Like there's so much weirdness there. Okay. People that have made tens of millions, hundreds of millions of dollars that I would have, you'd never know unless you're in this industry. Stories of being blocked out of countries completely at the DNS level. We actually had that happen to us at one point. Which country was it? Kazakhstan. It was so weird. It was great. I mean, I laugh at this, but it's kind of, it's just such a bizarre story.

1:42:00I got an email directly to me one day, which looked like it was just somebody, it was all lowercase. And it was like, you need to take down this site or we will shut you off. This is not, this is illegal in Kazakhstan. I was like, and it was from an email address. It was like a KZ email address, but it was just, it didn't look like it wasn't a government address or anything like that. I was like, this is really weird. I was like, well, we have a very clear policy about taking things down. Here it is. And he's like, I don't care about your policy. I'm taking you down. And sure enough, a couple of days later, we had some folks who were like, it's really weird.

1:42:32We have customers in Kazakhstan who are not seeing our site now. I was like, holy crap. That was real. That was real. It was DNS. Yes. It always is. And it took me two years to get us back in Kazakhstan. And I don't even think I tried reaching out. We tried. And eventually I reached out. And one day we were just back. Because that site that was their problem had left on their own accord, obviously, once getting blocked within like 60 days. But they didn't care at that point. So weird stuff like that just doesn't happen unless you operate in these types of spaces. This is the domain in DNS space.

1:43:14It's just, it's just been weird. It's great fun, but it's such a weird space. Well, I've enjoyed this conversation, Anthony. Is there anything that we haven't plumbed the depths of DNS or DN Simple or Anthony Eden that you've been waiting for us to ask? No, I think I've enjoyed it as well. I like, I love nerding out about this stuff. That's the thing. I don't know. I've learned a lot along the way and I've probably forgotten more than a lot of people ever know about this space. So it's just always fun to talk about it. So thanks for having me on to do it. Appreciate it. Absolutely. Thank you. We appreciate your time.

1:43:49Been fun being a customer all these years. If you ever need anything, seriously, like one of the things that we always do is we, everybody on the team watches our support inbox. And so if you have a technical problem, an engineer will get to it. If you have a non-technical problem, it could come to me. It could come to sales. It could come to marketing. It depends. So don't hesitate to reach out. And this goes for anybody who's a damn simple customer or who wants, is even considering becoming one. We're here because we service y 'all and we like doing it and we'll keep doing it because we actually care about what our customers want.

1:44:23Very cool. Well, I'm going to go open a support request. I'm going to put my t-shirt size in there and hit send and see what happens. I know who that'll go to. I'll fast track to the right person alright awesome thanks Anthony alright thank you both appreciate it

1:44:40well 15 years now where they've gone 15 years I don't know I sit and I wonder sometimes where they've gone little Bob Seger there for you yes 15 years with the Nsimple loving the Nsimple such a good platform a true focus on developer experience, a true focus on keeping it simple, which I think is awesome and admirable. Even though during the show, I was pushing for a real-time search for domain names and all the money that they've lost from us and others, I'm sure, because they don't have the service in place. But you know what? I don't have the inside scoop or the inside lens like Anthony does so of course i can speculate well if you haven't yet go to dnsimple.com that's dnsimple yes dnsimple.com check them out that's what we use and of course our friends over at retool our friends at heroku and our friends at depot making them builds faster 5 10 15 20 you know 20x faster could you imagine that your builds being 20x faster depot.dev baby check them out and of course to fly.io our partner our friends and the home of changelaw.com fly.io okay then beats are banging because break master cylinder brings those banging beats the best beats in the biz well that's it this show's done we'll see you on friday

1:46:39Game on!