Nick Johnson I Decentralized Domain Names I ENS I Pragma London 2024

[Music] so for this next interview uh we're going to bring on Nick from ens to talk about how ens has kind of evolved from where it started to where it is now all of us have seen anywhere from something. or. eth on our Twitter profiles to other forecaster lens profiles or just everywhere else on the internet and uh there's a lot more to dig into that in in the world of decentralized domain names so to kind of cover all of this I want to bring on Nick to have this chat so please welcome Nick on [Applause] stage Nick it's uh exciting to be uh talking to you here I know it's uh you don't get to do a lot of these chats often so uh we're we're honored to uh to have this opportunity it's a pleasure to be here so most people don't know this but Ian is one of the longest running projects in in the world of ethereum it's it's been around for a lot of years and uh the thing I'm curious about is uh did you imagine this is how things would end up and is this what was this ever the the exactly the goal and where you want it to be or how much was that from this is cool to oh my God we're running a half a billion dollar protocol now yeah I think that's you know that's been a startling transition um you know ens is far from the first thing I built um and you know everything I build like I I hope it'll succeed but you know most things are you know modestly successful and then to have something like this just take over sorry take off both apply um is pretty remarkable you know um I think uh you know all the time I spent building things that were only modestly successful has served me well and in knowing how to build a good protocol and so forth and you know it all contributes that's awesome um could you share any early day stories of how this was and kind of back in 2016 we were thinking about this as a protocol what was kind of the intention back then and um and would you say that's still the the current definition of what what yenis is well I've always been interested in infrastructure and you know back from when I was sort of age 12 and discovered our se's exist and you know that was my idea of some some good bedtime reading just about um you know until the current day and and particularly interested in things like DNS and the sort of the the infrastructure that makes the internet work uh and so when I came along to crypto and saw that you know we were basically back to the 1970s in terms of usability uh it was one of the first things to occur to me was we need a better solution for this and so I started building it pretty much as soon as I got involved with ethereum uh the very first version was very much an imitation of DNS you know it was almost was just importing it wholesale into ethereum uh but I quickly revised that and and and built on it to Leverage The Unique advantages of ethereum over you know the 1970s internet in order to build something that was better suited to ethereum and to crypto I think initially we were very focused on uh crypto identifiers exclusively so resolving ethereum addresses resolving swarm content hashes but we've sort of realized that the utility and the naming service is much broader than that and been expanding on it since awesome um well kind of from a from a fun project uh when you were doing this thing on your own to to uh kind of where it is now maybe by show of hands just who here owns an enss name that's 70% yeah can I ask uh who here owns an enss name that isn't a e second level name so coinbase ID Unis swap Etc that's great to see and I think over time that proportion is going to go way up absolutely but but but the cool thing here is that it's not just a we're crypto people and that's how we we're doing these things it has actually become mainstream you recently had a large partnership with GoDaddy um tell us about what that partnership entails and and how that came to be uh so you know what it entails is anyone with a DNS name uh which is you know any name in the that you typically buy from yeah like like goddy uh can now enable that name for ethereum and for enss with basically a single click on the GoDaddy site so if you register through them you can go in you can set it up you enter your ethereum address and Presto it all works and the best thing about it is there's no gas fees and it's not thats and godet are paying the gas fees behind the scenes it's that there is actually no gas fees the whole thing is done with zero one chain transactions but still with no additional trust assumptions over what you get from ethereum plus the DNS route it's awesome um how is it possible to do this without guas so through something called ccip read which is a a protocol for fetching data from offchain for smart contracts uh so what happens behind the scenes is when you go to resolve uh you know joe. cb. for instance um it goes off and uh starts the NS resolution the NS uh resol contract says wait I need more data to resolve this please go consult the Skateway server and the Skateway server happens to understand DNS and it understands how to actually query that DNS name and fetch the text records it returns those to the client which gives them back to the resolve contract saying you know here's the data you asked for and it's then responsible for verifying it and decoding the result and the remarkable thing about this is that the client that's doing the name resolving doesn't have to understand any of this it's all just you know here please go fetch this message for me and give it back you don't have to understand the content of it uh and there's no uh trust assumptions because the contract is able to verify the data so in this case you're relying on the DNS hierarchy but you already were in order to have a DNS name uh and you can simply you know as long as that assumption holds true there's there's nothing additional to trust in and using your domain in this way so that means the Gateway can't lie either so the thing I guess I'm curious about I think you did answer this thing and we're totally going on a tangent uh but purely out of my curiosity is this not a scenario where you can still tamper with the data that you're giving back to the contract so the what we actually implemented is a complete DNS SEC verifier for the evm so you know DNS SEC for those who are not familiar is kind of like TLS or SSL but for domain names uh it implements a chain of trust similar to a certificate hierarchy uh starts at the rout and works its way down and as long as you verify all the signatures at each layer you can verify that the DNS response is correct so what we actually have is an ethereum smart contract that can resolve DNS records securely and so the Gateway implements the sort of packaging that up and handing it back verification is still on chain I yeah exactly um but it's done in a call rather than a transaction so there's no gas fee gotcha yeah that makes a lot of sense that's awesome um I do want to dig into sort of uh the the DNS pieces uh but only in a couple minutes um one thing that um is kind of interesting here is that anytime you want to register a domain name or or ens name you kind of have to pay a fee and just like a domain name has a recurring fee um and you have to do that and renew it every year and uh the thing I gu I'm curious about is like was there any other reason for having a similar mechanism for ens here when you kind of had the opportunity to effectively do more creative things with money being native as part of the protocol um or was there the obvious reason being we just needed the the money here to sustain this long term and this is the only way you could have uh implemented this thing what was kind of the thinking around charging for these names uh yeah the the first priority is definitely an incentive design you know mechanism design approach uh you know if names were free then they would have all been grabbed up and and resold and and you know therefore wouldn't be free um and uh if they were a one-time charge then you would have a not issue where effectively um you know everything useful ends up on the resale Market um and you also you know you're effectively you're making it cheaper for people who want to buy and resell a name and proportionally more expensive for people who actually want to use the name uh whereas you know building a a renewal fee system in you know prioritizes the right the right users we believe um it doesn't hurt that it also helps the sustainability of of the platform and the protocol of course has there been any changes to how You' thought about Feast since the kind of the origins so the very first version of the registra um we weren't decided on on you know fee structure and so forth so what we did is we implemented a um a vicory auction and the uh winners what's that for for uh vicory auction is a is a blind auction uh so you put in a bid for how much you're willing to pay but it's secret uh at the end of the auction all the bids are revealed uh the person who bid the most gets the name but they only pay as much as the second highest bid uh and the Insight here is that with a system like that you are incentivized to bid your true maximum because you know that you'll only pay someone else's true maximum um and The crucial Factor here was that with the original version that money wasn't actually sent to a multi- or you know the eventual Dow or anything it was actually just locked so uh if you relinquish the name you would get the deposit back uh and so after 2 years we trans uh we moved to the permanent regist uh and that implemented the now well-known annual renewal system uh and anyone who held on to their name for that 2 years could transition to that got a Year's free registration and also got their original deposit back I remember uh losing a lot of money on my initial domain name registrations um so can I coming back to uh to the DNS pieces um why is it important for you to think about and Main maintain DNS compatibility uh obviously we're we're in our own world with E tlds and uh it works in in the context of uh this our blockchains but why should it matter that anybody else knows about this so naming is ultimately a usability issue and where usability is bad uh you know scams and and so on prosper and I believe it's really important that when somebody types in you know food.

e they shouldn't have to worry about you know am I using the same eth as my friend who gave the name to me you know is is there going to be a collision between these two you know maybe these names resolve fine but this other one doesn't because it's doing some sort of fall through mechanism and so forth and so I believe it's really important to to the greatest degree possible to have a unified name space and so we you know we believe that the best way to do that with is with DNS because we also believe that ignoring existing technology is not the best approach you know building on top of it uh improving it is very much in the spirit of the internet you know every new technology has succeeded by embracing and enhancing previous ones I mean I absolutely respect that um has there been any kind of unintended kind of side effects of of ensuring this uh I think you know it imposes some restrictions upon us you know at times it may have been very convenient if we could just launch a new TLD you know if we could say you know we're going to launch on this other chain and and convenience you know technically it's much easier if we just have like a whole new name space you know with a different extension um you know doing this properly forces us into some some more difficult technical choices and to to do it the right way the first time rather than the easy way how do you think about now kind of the emergence of a lot more new tldd so we obviously talked about the coinbase cb. ID which again is is compatible actually is part of ens um but we're seeing a lot of other new tlds appear um how should we think about breaking compatibility or rather not breaking compati like them breaking compatibility as more and more of these things appear and um what should people keep in mind as as they see a lot of these things uh come up to them yeah it's it's really tough and it's um you know I I guess my my first you know obviously self- serving note is to to try not to legitimize the obvious copycats you know the ones who are BAS basically taking an existing project minting their own TLD and have no you know no concern of their own for for avoiding these sort of collisions um but obviously just like everybody should just use enss you know and we're fine is not necessarily a sustainable position um I think that you know each of these projects needs to be looking at what they're doing to minimize their overlap and that can be through you know through minimizing the number of tlds they create like enss has um it can be through choosing tlds that they know won't be issued globally um it can also be just through using different uh syntaxes for their their naming systems you know you don't have to use dots separated for everything if you're building your own name system um you know what whatever you do I think you should be looking at you know assuming that people are going to uh you know have uh multiple naming systems around and making sure that under no circumstances can you get two different people into the same name into different wallets and get different results um and I mean it only gets more complicated when now people can just bid for their custom tlds in in the real world now for uh for TNS um let's kind of change some topics um obviously there's a lot more to still do on the protocol side so the thing I'm curious about is how how do you usually interact with the ensa and and the rest of the community how do they usually influence the direction of the protocol whether it's features or fixes or just getting General feedback uh what is that process like I think it's been sort of more organic and less sort of proposal driven than I might have anticipated you know a lot of the uh a lot of the benefit from the Dow has being been having a highly involved and vocal Community who provide the sort of feedback that when you're creating a new project you know you go out there and you say you know please please give me some feedback on this idea and you get radio silence and then you build it and then a year later somebody's like oh you should have done that and you're like great where were you a year ago you know but with an active Community with the da we've we've been able to get really incredible both sort of Technical and non-technical and usability feedback and build things better the first time um and that's been a much more organic process than just saying oh you know somebody put forward a proposal to you know change this parameter to this value there's certainly been a degree of that um and you know that's that's led to measurable improvements but I think the the input from the community is what I've really valued oh that's awesome um I mean that's a very healthy sign of an ecosystem that wants this to succeed what does this change in terms of your role as as ens Labs uh you know I think 's process through the whole the whole period since launch has been gradual decentralization um you know I think we're pretty damn decentralized by this point you know with with all the protocol being governed by the Dow um but you know enens Labs need not be the only service provider building thing you know infrastructure for for ens and we see you inally don't want this is that right yeah you know I think uh I think we're really smart uh I think we're good at what we do uh and I think we'll keep building good Tech um but I think there's room for more than one set of of Builders you know and and innovating on the platform and the recent uh approval of the streaming uh funds which was incidentally something that came entirely from outside the Dow from Alex vanders sand uh demonstrates that the N is ready for that too that's amazing um and want to have a well I do want to wrap up but I do have a couple other questions here um one is um what what's kind of making it hard right now for an to be more readily available on different l2s and uh and what should be be thinking about especially as we see hundreds and hundreds of these chains pop up in the next year so our short-term approach has always been that uh just like most people uh use a gmail.com or or Outlook or whatever email address most people don't buy their own second level domains to represent them and so our Focus has been on making it easy and cheap and secure and and Trust minimized for people to buy uh uh second level names like cb. like uni. e Etc um and to make that as efficient as possible uh in the longer term you know we we are looking at moving more infrastructure off chain in order to to minimize the onchain costs but our Focus really is on onboarding the next million users and making that as easy as possible rather than just optimizing what we have is that a stance you're taking on you want to just get the next million users to be on ethereum Main net and then that's kind of the priority or is it just overall users is what matters and if that means you do the cb.

ID appro and have other chains spin up these identity platforms that's the way to go or what's the stance that ens itself takes here yeah we we want them to be using these uh you know these second level Integrations you know things that that you utilize l2s for assuring subdomains things that utilize offchain Solutions and so forth got it um and maybe kind of my final question here is um uh what what is currently missing from the inis um whether the protocol or or your ecosystem and U what can we look forward to from uh from the community or or Labs uh I think the more maturity in the L2 space you know which is rapidly changing but uh I would love to see you know ZK based rollups further along to the point where it's easier for us to integrate with them and and provide that to users um and in terms of what you can expect is you know a lot more work in that direction um both from us and and Partnerships with with teams to to sort of build out the infrastructure needed awesome well Nick thank you so much for this uh conversation and uh can't wait to have a few million more people use enss and for 30% of you do not raise your hands don't worry we'll we'll get you an NS domain name by the end of next week thank you very much all right thanks s