Dappnode: Plug-n-Play Decentralization - Eduardo Antuña, Pol Lanski. Ep. 571

Transcript

like scalability isn't important until it suddenly EAS the centralization isn't important until it suddenly EAS and privacy isn't important until it suddenly is the solo Saker is a very particular animal it is somebody that understands the value of the centralization if you're a big pool that has thousands of validators and you're constantly producing blocks you are bound to get some lot some of these big payout big me blocks which are also known as Lottery blocks the inverse scenario is like a solo sticker that has one validator and proposes around two to three blocks per year the chance of this

validator getting a lottery block is Tiny very small so if we agree that home stakers are more valuable in terms of the centralization and resilience for the network why are we paying them less welcome to Center the show which talks about the Technologies projects and people driving decentralization and the blockchain revolution I'm Fred ANS and today I'm speaking with u Antonia and pansy founder and Project Lead of adap note and D note is a solo staking hardware and software provider and who will go into the nitty-gritty of that in just a moment after I tell you about our sponsors this week if you're looking to stake your crypto with confidence look no further

than course one more than 100 50,000 delegators including institutions like Bido Penta capital and Ledger trust corus one with their assets they support over 50 blockchains and are leaders in governance or networks like Cosmos ensuring your stake is responsibly managed thanks to their Advanced meev research you can also enjoy the highest staking rewards you can stake directly from your preferred wallet set up a white label note restake your assets on I or or symbiotic or use the SDK for multichain saking in your app learn more at kus1 and start taking today this episode is proudly brought to you by nosis a collective dedicated to advancing and decentralized future nosis leads Innovation with circles nosis pay

and metry reshaping open Banking and money with hashy and nois VPN they're building a more resilient privacy focused internet if you're looking for an L1 to launch your project nosis chain offers the same development environment as ethereum with lower transaction Fe it's supported by over 200,000 validators making nosis chain a reliable incredibly neutral foundation for your applications nosis Dow drives nosis governance where every voice matters join the nosis community in the nosis Dow form today Deploy on the evm compatible nosis chain or secure the network with just one G and affordable Hardware start your decentralization

journey today at nosis doio he it's a pleasure to have you on the pleasure is entirely mine to so we've we've known each other for a very long time um but it's the first time we're actually talking about dno on the show you've actually been on kind of on one of the on one of the adap con episodes I think but I don't think we've actually ever really C covered that note have we well we talked some time ago when the merch happened being in the center and it was great but yeah I think it's important to cover also T here I guess absolutely before we Dive Right In tell

me a bit about yourselves and uh your backgrounds maybe we'll start with you ad okay great well I've been around since 2017 it's has been a long trip to be here today A lot of thing uh in the middle I started uh working with Jordi making on audit like make th and during this audit the idea of D not comes to my mind and your this one and we started to make this happen and yeah the product has evolve a lot um and we have uh tried to adapt to the market during these years um so yeah really interesting travel to to be here today and a lot of

things to share with you you cut out for just a little bit but kind of you said um that kind of uh ji uh was was this was also J's idea yeah yeah yeah the founders of D are Jordy valina Griff and myself and yeah M the idea that um comes to my mind and J this one G is is about decentralization we realized that at that moment we were creating decentralized technology with but without having decentralized hardware and we started to make this happen absolutely and I mean both of them um are now involved with other projects primarily but kind of they've also been

on the show before so uh Jordy actually several times for uh for ours and polygon um and kind of the ZK stack that he's building ckm and so on and Griff has been on some of the community shows what about you Paul well um well first of all it's a pleasure to be here thank you for uh thank you for inviting us and uh I think I I started in crypto um mining Dogecoin in 2014 fa uh the reason why I decided to mine Dogecoin is because I was late to bitcoin but I realized that it was the same code uh so if I learned how to M

Dogecoin I would just need to scale up my hardware and it would uh it would basically be very a very similar experience um then like I sort of went onto the corporate route I did my MBA I um I worked in corporate for a while and in 2016 I met uh Griff green who uh with whom I collaborated in gith and then uh in 2019 when I was ready to quit corporate uh he introduced me to edu and uh in the first I think in the first five minutes of having a chat with edu I knew that the next day I was going to resign from my job um I think

it was everything that didn't exist in in the corporate environment where I was uh so I I since then since 2019 I've been working with uh with dnot and when uh all the founders took a step back and and went to other projects I decided to uh take over and and try to try to bring it to where we are now what were the things that uh you felt IDU and D note offered that were lacking in your uh in your legacy job yeah so um I think at that point it was like you could go straight to the point

you could go straight to what we're discussing um without endless layers of politics of thinking uh which department would that fall into who respons responsibility is the was this it was just about doing something that was for a greater cause and uh Ado was just so patient with me like I I pretended I was I was in a job interview mode and I pretended I knew a lot and edu was uh super down to earth and uh super uh dactic and tried to educate me a lot uh in a super humble way as he is so um I was uh I was humbled and I was like y this this straight to the point smarts

is uh with whom I want to to be working with I feel yeah I feel yeah I think that's generally a virtue in the ecosystem so uh it's also what I really enjoy ID tell us about the founding story behind uh D node so kind of like it was you ji and Griff and kind of there was the maker audit so what what kind of um what kind of drove you to say look we should make um we should make hardware for people who want a SoDo steak Yeah well that's it's a really good question and yeah has been an evolution to to be honest so the the first idea was just to create software

not hardware and our Deion was to create a kind of um project not to make a business with it but to make it s sell sustainable yes great idea to provide um a public good to people people allow people to have decentralized technology it was not just sticking at that moment because at that moment sticking was kind of an idea only but it was about to have um privacy to be able to connect to the centralized Network in a really decentralized manner so yeah we we started uh the project just myself and and later on we started to ask for some grants to maintain the project we we got grants from uh NIS we got grants from Me

Foundation uh Aragon M these three actors were the first one to support us and it has been a small project uh from the beginning only four people at most uh during the first three years maybe because yeah the idea was just to try to evolve and and maintain the project and survive I mean we have survive I think three markets I don't know the number now but for that you need to be very careful and and and just continue building um so yeah we we started this initial idea of having this open source software for for anyone but then we realized that we need something more because people at the end of the day

wants to have a kind of um um autox solution makes uh their life simple because at the end D note the idea behind D is to make easy for anyone to run decentralized networks that was the initial idea so that is why it was important us to combine the software with the hard Hardware solution obviously running a hardware company is not easy at all is is a pain for us but um but I mean I think it's needed and I think it's providing value to the user that is the first goal of the of the project to provide value to final users to make them um access to this kind of technology that without this facility would be impossible so yeah mainly um

some years working on this then uh uh something that changed everything it was uh the the well the proof of stake and and the merch obviously because it's when people start to using D for something more than than privacy and at that moment we start to Sol to solve more and more hardware for people that want to run solo sticking at home and yeah really really Pro of the of the success that we achieve with this to Hest so who who are your users because kind of like um the the Privacy I mean I totally understand um the Privacy point that you were making so kind of like whenever you kind of um connect to say and at that time it was mostly metamask

so kind of like that would go to infer so kind of infer would kind of like have all your data and even if in principle kind of like you had been super careful about kind of like where the funds on your address for gas and so on came from infura might be able to kind of infer who you are and kind of and so on so I I totally understand the the value proposition than principle in practice kind of very few people are actually willing to kind of pay for privacy a in funds and B in time and upkeep and so on so kind of who who were your users then and has this changed um yeah sure I think at the at

the beginning it was uh a lot of uh enthus iasts a lot of uh people who really really believed in ethereum really believed in the centralized Technologies ipfs um and they wanted to contribute to the network it was more a contribution to the network than to uh than to like take really use it of course there's always been like a fair share of developers who prefer to have their own node because then you have like limit unlimited querying and um you don't uh you don't need to pay uh an API or being rate limited Etc but uh at the beginning it was mostly people who who really believed in ethereum um and the the

shift really came with the first test nets for what was then known as ethereum 2 uh the medalia test net the all these test Nets um that that came out people started realizing that proof of stake was going to become a reality and especially those who were not particularly technical uh they figured out that with that note it could be an amazing solution to just uh oneclick deploy and and be running those test Nets so I think like in the matter of like the 3 months from the first test net going out uh like we like six or six or 10x our the people in our Riot

Channel we were using Matrix and and riots back then uh so it was wild it was we decided that yeah we that that was probably a good pivot to focus on uh on staking above uh anything else yeah and I mean what what maybe just as background so kind of before proof of stake obviously we had proof of work um and there you kind of you needed very specialized Hardware right so kind of it wasn't something that kind of you would run on your on your regular machine at home so kind of like people build people build and bought specialized Hardware so kind of offering that um out of the box kind of for people to kind of just install on kind of like the machines they never turn off in their homes that

wasn't kind of an option so kind of it was literally just running a full Noe kind of getting the complete State and then kind of after the um the transition to um proof of stake or kind of like even the leadup of that um transition kind of people were were actually uh validators on the network properly so I hear that kind of you have had a lot of idealists sing is also can also be a profitable business um has kind of this crowd shifted somewhat from kind of like the people who just wanted to be part of the network yeah absolutely um there's so I think the solo Staker is a very

particular animal it is somebody that understands the value of the centralization and of course uh in general it searches for the yield that ethereum provides but it's looking for something more it's looking for a higher level of contribution and a higher level of exposure um the to to get access to the yield of ethereum it's very easy to just buy into any liquid staking protocol and and then you'll have access to that yield um so definitely like if you need to invest in Hardware or you need to invest in something else is because you want a higher degree of control and here like there's there's

multiple multiple personalities or multiple uh overlaps there's the the developer that wants to run their own node and again query it and at the same time manages to use it for validation there's the OG the that that has massive amounts of eth from the early days and uh maybe they were even stockpiling 1,500 e that it was supposed to be the first staking uh specs uh and and then all of a sudden they found out that they could run a lot of validators with that so so yeah mostly it's somebody that that definitely wants access to that yield and wants the economic interest but that also believes that their work um on validating at home is valuable for the ecosystem as well is that also how

you determine um which network to support because kind of like you support networks Beyond e theorum so I know for a fact you uh support nosis because I validate for nosis uh um I also know you support ipfs and Hopper and a couple of others so kind of like how do how do you determine which networks to support that that's a really good question to be honest obviously we we are trying I mean as I mentioned at the beginning we are on a small team uh so there were help Ming several networks several packet is is hard for us so we need to be very careful choosing them um especially we take a look to the value that provide to

the user and also if they are align with our values that I think is more the most important thing at the end of they D this kind of a general purpose machine right so we will like to have as much as possible but then obviously uh these are the metrics kind of right uh that they are providing value to the final user that is uh decentralized that is offering yeah uh several things and and go going back to the previous question or linking to it is is is um we see how people are not getting too much about privacy but uh also privacy is related to censorship

resistant that was also one of the initial values and it's sad but people only realize about the importance of this when they are in trouble when uh government start to censorship you and that is when you real realize this that you need this technology otherwise it's like people forget about it and from day to time this happen around the world and it's a a big push for us probably but uh it's important to keep in mind that we are building the centralized networks uh no is uh I don't know the the number right now Paul maybe knows how many validator solar stickers are in no but I think it's 30% or something like that yeah just in except of

30% yeah it's it's amazing to see how this is is this network recentralized right and how we are in our helping to make that happen and and and at the end of the day we are creating decentralized Network for this Reon not because it's more efficient I mean if you go to centralized sequencers you are going to get much better performance right but then why we are building decentralized technology just to have this these properties I think if I may add to this um one very smart uh design choice that uh that Adu and Jord did at the very beginning was to really think about dbot as an ecosystem and to to create the SDK so actually anybody can create uh any sort

of apps or any sort of DB note packages or dab note this or DBS as we call the centralized apps right um so it's not so much who do we try to support but also who is able to to be part of the ecosystem and the answer is anybody U all of the networks can uh like if you can dockerize your node it can be deployed on dap node through a through a Dap node app through the SDK and for and then then of course like the ones that we have supported since the beginning are the ones that as a said are completely aligned with uh with our values right those are the ones that that we've been pushing through and maintaining ourselves when the teams um the client implementor teams were too busy or something um then we've been

putting our own efforts into making it available on that note uh I want to come back to kind of like the sing philosophy and Outlook later but before that can we maybe kind of quickly cover um the tech uh aspects of dep node so kind of like if you look at the tech stack that powers dep node um what are the major components and how do they work together oh okay I could do this one probably so yeah uh from the very beginning we envision this project like a totally centralized platform so that is why uh the first thing that we did it was to create um to deploy the AR smart contract the mpm um the APM just to be able to

distribute versions of the package in a decentralized manner so we l link this smart contract to ipfs uh plus ens and this allow us to make a kind of a package manager distribution in a decentralized pure decentralized manner so that mean that if one publish a package no one can block you no one can um remove that package from the network because it's completely decentralized and that was one of the core values of of that employee and then just the other part of this is what we want to distribute and then is is doer so we create a platform of Dockers that uh are Distributing in in this in the system and that makes

quite easy to to develop an application for dnet as Paul mentioned because the process is just to take the d DK just to create a Docker image uh uplo to ipfs and then link to Smart contract and with that anyone in the w g install it and and I think that that's great but also I mean we have been talking a lot of about H blockchain but that is not just that uh at the end of the day something that we are trying to do is to bring home back the computerss because it's like people um are just not having this kind of computer at home they are using laptops and other things and we are trying to force them back to have this

and at the moment that you have a a computer at your home you can do more things like uh time machine uh you can have your own cloud for uh photos you can have not just blockchain and and mainly that is the the the idea and having a decentralized uh distributor system for this I think there's a there's a naval tweet that I that I really like um I think it goes something like this like it like scalability isn't important until it suddenly is the centralization isn't important until it suddenly is and privacy isn't important until it suddenly is so I think TB node is ready um to to make this happen at any single

point and for any platform to be to be ready for uh for either needs for scalability the centralization Privacy so what can we through this so I'm a vanilla user so kind of I I say I want to use I want to say sake um on nosis and ethereum um I have a DB node I kind of plug it in I kind of connect it to the internet um what do I need to do then once you connect your DB nodes um we we've tried to design the whole flow to be extremely easy and to not have to use command line at any point so basically when you plug the machine to

your router and you turn it on it will start emitting a Wi-Fi hotspot and you will connect with this with your device we be your phone or your laptop or whatever you connect to that hotspot and then you have to visit my.dab note which is being served by the DB node itself so basically you use your DB note from your browser and from this browser you can go through the configuration flow and then you can change the password of that initial wi Wi-Fi you can uh configure access through VPN and this is important to have access to your dab note from wherever you are in the world not when you're next to it and and in range of this initial Wi-Fi thing um and then

after this initial configuration uh and configuring the way you want to access it then you can start installing packages right and installing apps and for this there are two modes there's the DAP store and in the DAP store you can find all of the packages that people can create through the SDK and that we have published as well um and then there's the stakers UI and the stakers UI is a very specific flow for ethereum based uh blockchains right now we have uh three of them we have ethereum nosis chain and Luxo and basically the in the flow it it guides you through um configuring your

execution client your conent client then you are forced to install a web3 signer that's where the key stores uh and of your validators will live and then you can choose whether you want MV uh MV boost or not so this is monkey proof basically like it's it's you cannot go to the next step unless you have chosen one of each and and then you click uh apply and dab node then uh installs all the clients and configures all the connection among etc etc etc um so yeah that that's sort of like the flow that you would do to start validating then you would need to create these keys this validation keys somewhere else and then

upload them onto onto your dab note and you'll basically start validating Pro approve of stake chain this is always kind of like when you hear this from techy it's kind of like monkey proof it's usually not monkey proof and I don't think it's monkey proof but I can I can attest to the fact so I mean I I I'm attacking myself but I know plenty of uh people who are not devs who actually run dep noes so kind of like within noses we actually encourage everyone to kind of run noes and a lot of the large parts of the Nev team actually successfully run noes at home um so this is very much this is very much attainable for uh

interested nevs for sure um so so tell us about the DB store so kind of um is it is it fully permissionless so could I could I um uh could I kind of uh upload some sort of malicious doer there yeah this is um a really good question um we have uh two repositories in dnot one is the DB repository that is the one that we control uh is the one that we take care of and we are 100% sure uh that everything is fine and nothing can be breaking because of the repos we are publishing then and then we

have a completely permissionless repository that is the public uh that doties so yeah the idea is anyone can publish on the public um and then obviously this is the jungle you need to be really careful what you are installing but I mean there are really interesting things also in the public like I don't know optimising package or or Community M packages that are trusted so yeah at the end we try to combine these These Little Things uh one more trustful and one more trustless and with the idea of of creating an ecosystem because at the end of the day is like um Regular repository is like the the App Store that you have

in your mobile phone uh in Android right so you can install an an application that you pH everywhere around internet or you can just use the official one so yeah it's it's hard because at the end of the day people are restoring value on on this Hardware so that is why we need to try to balance and protect user of not installing uh the thing that are not correct what's the process of um updating your notes like for a user because kind of like as we all know kind of like these updates happen regularly and they are often somewhat manual so how do you how do you kind of balance Security in in usability in these

scenarios the I think the key here is that once you install a package um these this package is signed by the key of the developer so somebody trying to spoof a uh an update or something like this it wouldn't pass the validation checks um for for like the the the does and first of all they couldn't even publish into the same uh into the same repository into the same smart contract because the permissions wouldn't allow it but there is this uh develop trusted developer key system that sort of helps you manage which

developers do you want to trust and this also works for the the public DB store as well right like if in the public DB store um you have decided to trust uh some developer there because you you trust them and you you've met them and there are some people who who have reputation in the space and they have published packages um then you you will not be alerted or you will be a you'll be allowed to download these packages and update them etc etc if a package is updated by a uh an address a wallet or signed by a wallet that you do not trust you'll be alerted so you can uh you can make sure that that this is a person do that you trust or like an update that

you trust now when it comes to the updates themselves um because dap node is uh already online at 24/7 and it scans the this repository the smart contract this Aragon package manager that we were talking about whenever the developer publishes a new version of it uh the user will be Al will be alerted so the DB node will read the smart contract will read the event and we'll say hey there's a new uh there's a new version available and you can even set up automatic updates so it will automatically update to the newest version the newest version of the of the D node package which is really handy

because basically takes the maintenance to to only looking at your D node when something goes wrong um you can of course opt in or opt out of those up a to updates and you can check and and manually uh update uh the the packages yourself so so how often do things actually go wrong so say I set it to auto update how how often would you expect I kind of have to check in and manually do something B people on the D note Discord and so on H that's a good question um it kind of like the the things that go wrong are usually not related to DB node the things that go wrong are usually related to uh like the the power failing at home

and then disconnecting the machine and then the database of the execution layer client uh gets corrupted and you have to resync again um or the internet goes down or they change your IP um and or there's something with the router and then you lose access through through VPN so dabner itself is so if everything goes well you rarely have to look at it and that's actually a bit of a problem because people um are missing out on on new cool features and and new things because they just don't have to look at their dab note ever again uh after setting it up uh but if like if you leave in an area where the light is constantly going on and off like you you

can expect that at some point you will have to resync your execution client because the database will will be corrupted okay fair What proportion of D nodes run on D node Hardware oo that's an excellent question probably my favorite um and the answer is we have no idea righty the reason why we have no idea is because we do not capture any Telemetry by default um and we know the DB nodes that we have sold of course but we have no idea if you installed the free open source software

into your own machine so we we can only tell you about the dnos that we have sold but not about the proportion that these represent over the total amount of TB noes that we have so um and this is by Design because like as as was saying privacy was one of like the the core principle but but you know how many D nodes there are in total right uh no no we don't I mean we know the ones that we have sold but you don't know how many people actually run dap node packages for for kind of validation no um the only things that we can

measure are proxy metrics which basically is so by default if you're using dap note to validate the default graffiti will say validating from D node so then we can go and look at the validators that have this graffiti and safely assume that they're probably running dap node we still don't know in which Hardware like if it's Hardware that we have uh that we have sold them or or not um so we can have an idea of how many validators but again there's a decoupling of validator to hardware and we don't know if two validators are running on on the same Hardware either

so it's very very very hard for us to know uh or to to even estimate the amount of dap nodes that are that exist out there this is what you think so kind of like the the graffiti there being a good indicator do you remember like 15 years ago or so when people started sending emails from their iPhones and kind of it would say send from iPhone but then kind of everyone would kind of copy that kind of like on their Hua alwayss and kind of you know put like scent from my Ione below because kind of just the this this what you did right um okay yeah yeah joking aside so you don't know how many how many people are running dep nodes but how many dep nodes have you sold

uh we have sold around 2,000 D nodes okay okay if you could go back to the very start of dep node is anything you would do different U oh this are really good questions I don't think so awareness I mean we build this with um very strong principles we keep that from the very beginning and I'm really proud of being maintaining that maybe uh I don't know at some point maybe try to find a small uh business model to sustain this project could be interesting maybe because forance as we have been talking we don't know how many people are running down for Val e or or nois we have this sh reported G but at

some point we were running around 1% of the c network and this is a lot a lot and we were not taking anything from that just surviving with Brands and and and and well the selling of hardart but yeah probably probably you has to get a little more for for making this bigger because obviously when you don't have um this bigness B mod business models for for this kind of software it's really really hard to maintain especially at the very beginning uh so maybe try to find something in the middle although if I need to choose I prefer the the the path that I took in the pass but yeah obviously make our life easier to build

this will be will be great at some point you had a pretty large upgrade recently dep node smooth talk about that and kind of how that fits into the broader dep node ecosystem so I think uh we realized the importance of smooth uh when when we realized that we are paying less to solo stakers overall than to Big centralized players and and and you you'll go like what wait what what do you mean we're paying less at these people and the

reason why why I say this is because um we have taken this me path on on on ethereum and we have decided to to to manage the the block building on this like Proto PBS uh that that the me uh me boost enables and what what this does is basically um Meb distribution is follows a power law which means that very very few blocks will have huge M Rewards and the great majority of blocks will have you know not so great MV rewards um what that makes is that if

you're a big pool that has thousands of validators and you're constantly producing blocks you are bound to get some lot some of these big payout Big meev Blocks which are also known as Lottery blocks so if you get one of these Lottery blocks with these big pools do is they do not keep that to a particular validator right because the validators within a pool are fungible so basically they distribute these huge reward among all of the validators of the same pool and that brings the execution reward higher for all of the blogs in that pool the inverse scenario is like a solo sticker that has one validator and proposes around two to

three blocks per year the chance of this validator getting a lottery block is Tiny very small um so most likely the most likely scenario is that in a year one validator that is a solo validator will not have exposure to a lottery block and will have a small reward whereas uh one validator that is part of a big pool will have a higher return because they will have exposure to the lottery blocks that have fallen into into this pool so if we agree that home

stakers are more valuable than one marginal validator in a data center because they give more resilience more geograph rical decentralization more more geopolitical protection more uh censorship resistance so if we agree that home stakers are more valuable in terms of the centralization and resilience for the network why are we paying them less but is it really the expected value so kind of like if I mean if I kind of it in the infinite margin so kind of like uh I kind of I run a solo validator forever kind of I will hit upon a lottery block and I won't have to share with anyone I will get to keep it all for myself so doesn't that kind of make up yes exactly um but at like in an

infinite timeline you are bound to get a l block but you might get it in 200 years uh or in 80 years or who knows if ethereum Will exist so in a 5 years timeline you are much better off joining a smoothing pool because you in in five years you would produce 10 blocks for example 10 to 15 blocks and that's still very very very little and that is if we continue at uh at our current level of validators which it seems that it hasn't stabilized a little bit lately okay so after we realized that that the uh solar stakers were sort of

like being short changed on their execution rewards we decided to like build a pool for their execution rewards and that's literally what it is like the solution to this is basically we all pull together our execution rewards into the same smart contract which uh then distributes these rewards among all of the participants of the pool therefore if one of us gets a lottery block everybody else gets their share of this Lottery block so yes if you get a lottery block when you are in the pool you will have to share this Lottery block with everybody else but the the thing is you cannot possibly know in advance that it is going to be you who will earn this Lottery block so that's

the trade-off you either decide to have higher than expected execution rewards by joining the pool or decide to play the lottery and have overalls or most likely a scenario where you have less rewards than if you had joined the pool but you get exposure to this huge Lottery block that you might get so it's kind of like Insurance inverse insurance so I get a payout every month but kind of if I become very sick kind of uh kind of uh I uh I also kind of the insurance pay that for me right yeah yeah exactly tell me about uh kind of uh dep

node smooth and the business model it also changes things on your end right yes yes certainly I think um so we've decided to charge a fee for smooth because it basically so what happens with the numbers the actual numbers of joining smooth um is that you will you're expected to double your execution rewards um so basically you if you consider execution and and consensus rewards you you basically maybe like a you increase your total rewards of a Val validator by 20 25% something like that if you're participating in these and you can double your rewards we've decided to take a 7% fee of all of the

rewards that go into the pool um and that sort of like changes what AO was saying before right like uh up until now U we we didn't have any business model built into tap node and that's because tap node makes it extremely easy to to run validators and stuff but uh we didn't quite develop any client implementations and basically what you're doing with appn is you're you're basically running other people's uh software right so charging for this I mean could have been an option but we decided not to go for it um but here it is something that we have created it's something that actually gives you value gives actual value to the user it gives double the expected execution rewards so

we decided to charge a fee here and that brings in or starts bringing in this business model that can make dap node sustainable are there technical challenges to kind of um develop and Implement dep node smooth because it kind of from from the face of it it sounds like a pretty straightforward pooling contract no yeah maybe maybe the most challenging thing here is the orle because at the end uh most of you know uh there is a disconnection between the thing that are happening on the secution layer and the thing that are happening on the consensus layer so it's not possible to to correlate this data so that is why we needed to build an vle um obviously well

we have been talking here we are not uh too much too much aligned with oracles probably but I mean we create a system of uh different Oracles providing the same value and obviously we envision a a future in which we can replace this uh Oracle for th knowledge proof of the of the result but uh the most complex part of of this uh smoothing pool has been the the regle because he the one that take care of the the the wrestles of the things that happening on the cons layer and posting this data on chain um but yeah following the the culture of dut we obviously did everything open source and anyone can run his own Oracle to double check that

everything is fine so just trying to to be aligned with with our spirit is I mean we we have a mantra in D is decentralized until inher uh centralized until it works and we try to follow that from the very beginning and because at some point there are something that you cannot decentralize and I'm just thinking about for instance um you mentioned about automatic updates well we have a connection to telegram so if you use a in telegram you can get uh notifications in telegram uh but obviously this is obtained and there is no way to send a decentralized notification to our user so at some point you need to find this Val right in the and and yeah a balance

between the centralization and centralization to make the things works so what's been the feedback like from the community for dep node smooth H yeah that's a that's a very good one uh people who have joined they absolutely love it they some people were saying like hey I haven't even proposed a block but I'm already accumulating rewards and I don't know what the reward of my next my first block is going to be but I I know that I know exactly it gives me this predictability on exactly how much uh how much am I going to get right um I think most of the bad feedback has been uh like not really understanding the difference between the expected

rewards which are around the median and the uh rewards that you can or or how smooth gets your rewards closer to the average uh people have some some trouble understanding the difference and I have a I've come up with a with a quite an easy way of explaining it um how how the median and the and the average are different so imagine that there's like um there's a company and uh we are all part of this company and the owner is very fair the owner pays uh €1,000 to each employee including himself um so every month everybody gets

the same so the the median salary uh the median is defined by if you put all of the uh the numbers or the salaries in order from the biggest to the smaller the one that's right in the middle is the median right and it's clear to see that if we're 10 employees the the median will be like 1,000 it's uh and the average which is defined as uh adding up all of the numbers and uh dividing by the number of instances of salaries so uh adding up the 1,000 dividing by 10 uh that's also a th000 okay so that's these in ethereum are the consensus rewards they're equal for

everyone now imagine that the end of the year comes and me as the owner decide to pay myself dividends and I pay myself uh like I pay myself like 90k of divid it's a huge payoff not not 1,000 but like like 90k so basically at this point um on that particular month everybody will have their salary of 1,000 including myself but I will have a massive bonus of 90k so let's look at the median for that month the median for that month is going to be take all of the salaries put them one

after the other and the Medan need still 1,000 but if if you take the average you add up all of the rewards that you've been given that's going to be 10K or like 100k so if you make the average the average of the 100K divided by the number of employees divided by 10 it's 10K but this number is meaningless to you as an employee if you hear oh the the average salary in or the average compensation in that in that company is 10K well for you that is not the owner that like that means nothing and what you're interested story what you're going to get is 1,000 not 10K right um so something something like this happens

except that we don't have owners um in in ethereum or like not not in the traditional sense not in the same sense but the like we what we have is people who are lucky enough to have one of these Lottery blocks and this if somebody gets a lottery block it pushes the average up but that doesn't mean that it means anything to you who most likely are never going to get one of these lot blocks so if you sort of like trust the pro your probability or like look at the probability of you having one of these Lottery blocks and on a as you very well said before on a time span of 5 years how many blocks am I going to propose how many of those have a chance of being a lottery block what you're going to end up is that your reward is

going to be closer to the median what you can expect is closer to the median um now what smooth does is really pulling this 90k of bonus that I gave pulling it and making it and giving it to everybody it's like instead of being a a company it's a cooperative and we all divide all of the all of the benefits amongst everybody which really does bring the total compensation to uh 10K instead of uh for everybody yeah so so what proportion of dep node users have opted into smooth well that would be an amazing question to answer if we knew how many DB users we have I think it's important to mention that

uh ding pool is not only for D user is for any okay and and right now we have uh almost 2,000 uh validators on it uh and yeah we we're getting considerably considerable smoothing effects already like we we are hitting some Lottery blocks I think the biggest one that uh that we got was around 13 14 e something like that wow um let's talk about kind of like staking kind of from a bir side perspective right so kind of solo staking um as

admirable uh and virtuous as it is um it's kind of um economically it's often not the best way of kind of like using your e right so kind of like if I um use a liquid staking protocol I can kind of earn uh rewards um on my e and at the same time kind of use it as collateral somewhere else because kind of it becomes fungible right so I don't I don't have um the the the uh uh the opportunity cost that I kind of have when I stake um how do you see that yeah I mean this is a really good question so um the the thing that we are

seeing or at least from my point of view is that is at some point we are putting the incentives in the ground place I mean as Paul mentioned before um we should incentivize more the the solar staking but the thing that is happening is is the pro I mean the people that are not doing solar staking are having more benefits are having um less uh less or more facilities to make this happen and especially with the leg as you mentioned now um what I expect in the future is to have these Lu at scking Solutions combined with DBT because now right now one of the big issues for for D not for solo sing is that 302 is is a lot of money and not all the people can

afford that obviously so if we are able to combine these DVD Solutions and liquid taking and make anyone to be part of this decentralized network validation in um in a way that um make easy makes is easy for them and also provide this value of of liquid taking I think it's the the feature that we need to SP or the thing that we are trying to achieve because otherwise we are going to end up in the same situation in the past like in Bitcoin where most of the the minor are controlled by same people and and from the very beginning of the calization was always uh one go there and and that is why today we are able to

run validators on consumer Hardware because it was always one of the principle of theum so I hope that in the future we can start to go this direction of of being able to to take the semit or even more for Sol Taker and and not benefit the one that are going in the centralized Direction centralized path you just mentioned DVT um can you maybe uh explain that yeah sure I mean the thing is currently um the keys to make aatur uh ofor um has they have a lot of power and and they can sign a message they can slash you so you have an issue you

cannot um decentralize this key or give this key of validation to anyone because they can um exor you they can do bad things with this key so the hand of DBT of the DBT technology is to create a way in which no one has the control of thisy and with the signature of three or five or uh n of uh um members of this committee they can make this in a Tong of the validation this allow us to give this power of validation without exposing and and to be exposed to the risk of of exposing the key so theidea Hand of this div is just to create um a small groups of people sharing a key but not having the

control of this key allowing to anyone to uh to trash on this decentralized manner of signing things and if you put the together with liquid taking you can take ether from people give a liquid derivative of this ether and put this ether to validate on a decentralized system of people running this new technology so that's distributed validation right so kind of like you need less than 32 e um and you can kind of pull those together and kind of you can do it in a trustless manner is that more or less it yeah yeah much better than my explanation for sure thank no no I yeah um and um yeah so kind of why why would that kind of eliminate the opportunity cost that

people currently face when they solo stake yeah uh the thing is that currently you can run 32 uh I mean one validator do is I mean obviously with no is is is just one so it's much much easier talking about the difficulties that mainly that is why we have more validators in rather than C probably because it's more a affordable is more accessible for for users to be part of the network but in the cas you can only run 32 e 64 so is a lot of money um probably if you have a hardware uh you were willing to run I don't know 100 uh 1,000 validators in the same Hardware

right but you are not having access to this big of either or maybe some people but not not the regular people so the thing is to find I mean it's like we have pool of people having and maintaining Hardware willing to run validators for other ones and in the other side we have people with e willing to decentralize this so at the moment that you are able to trash on this decentralized Hardware ler of people validating maybe you at home are not going to run only a solo Staker I mean you can run your validator this 32 Plus validator for another people from this pool allowing you to get more benefits

and and reduce your operational cost because you are just running your validator plus the validator other ones and getting the of of doing this this job for them so yeah maybe this is data but you can't use the steak de as collateral as you can kind of do when you um have liquid staking right kind of like when when when I liquid stake I get a liquid token back that I can use as collateral elsewhere to kind of take out more e or whatever right and kind of and yield on that as well so kind of I can leverage up yeah as a user I think I would say it's even better what you can do which is you can you you can put you can be running an operating a validator without a bond and so as Z was saying

there's there's these people who have the eth and they need to be running like the pools or centralized exchanges or uh or liquid staking pools they have the eth and there's no reason why before DBT they would they would unrust random user on the internet to to run their validators to run validators for them because they owe it to the people who have surrendered their liquidity uh their actual eth for staking they owe it to offer them a proper yield and if like they they need to ensure that the operations of these validators are okay now with DVT what they can do is they could they can give it to groups of random people on the internet that they've gathered together they've created a DVT cluster and and now they can present themselves as credible node operators because if one of them goes

down as it's as it would be perfectly normal on a home setup the rest of the cluster picks up the slack and what they can do so as an operator now let's let's look at it like as an operator you don't even need to actually be a Staker to benefit from the like the the rewards of operating so you will get paid as an infrastructure provider not as a Staker anymore so you're basically you don't even have to put up any Capital so right now there's only uh one program that uh that does this or I think maybe two programs that do this that they require no no bond um there's ethery with the

operation solo Staker and they manage their DVT clusters uh and they they GI E from The Ether protocol uh for them to to to run validators on and then there's the simple DVT module of Lio so in order to putchase IP in either of those you don't really need to put up any bond so there's literally no cost of opportunity you just need to run the hardware okay I I I understand that kind of if you're the person kind of running the hardware but if you you're the person with the eth and you still better off kind of doing this in a way where kind of you get a fungible token back that you can use in things because kind of like when you kind of stick um kind of in a DVT um setup you get no fun

token back that you can you can still utilize right yeah that is correct you you don't if you want if you have the eth and you want to have a liquid representation for it uh you're obviously better off like DVT is not useful for that then you have like other things like you can create like a uh a volt on stake wise and uh that will create a liquid representation of the E Stak in that bolt and then you're what you're introducing is you're introducing a layer of smart contract complexity in here but if you're okay with this then you can transform the eth of your Vault into OS e which is which then you can

use on defi so you you can totally do that as well by not necessarily by using DVT but by using other Solutions like stake wise volts okay yeah I I I understand the other the other thing that kind of gets discussed very critically is reaking what's your take I he that yeah maybe just very quickly in not kind of reaking is the idea that you can use the same collateral twice to kind of stake four different things so kind of like if you already have e staked on ethereum you could also use it to validate on seven other protocols and kind of be slushed on the same eth um which kind of allows you to kind of

multiply the yield you can generate on that e it also possibly overloads the security consensus so um yeah what what's it take yeah so uh I think it it is a great idea in practice uh having all of these e sitting there um is uh not being slashed is could be like could be seen by some people as like a a waste right like all of this eth that only serves to validate ethereum um and it could be reused for for other protocols so but mostly I think it's given way to Greed and the ability to or the the

willingness to get more rewards uh out of the same eth and this basically endangers uh endangers ethereum as a whole like in the case of a mass slashing event on the reaking protocol um like eth could see uh like like this let's let's bear in mind that this is actually not e that it's not staked at the same level as the um as the consensus layer so it's not uh burned uh but it it will be sold and it will basically dump the price so uh a massive re slashing event uh reaking slashing

event would uh would be very very negative for ethereum and I think it's something that has to be like it it it is okay to be tried slowly um but I think it's been overblown and um uh sort of incentivized with all these points protocols and uh it increases risk and complexity in a way that we stop being in uh in control of the financial products that are built on top of uh States uh of of the security of the consensus uh of of ethereum and it exposes the consensus of ethereum to a lot of complexity in the same way as the

repackaged mortgages uh exposed the financial system um to to a lot of risk in a way that we could not possibly understand what was what was going on behind the financial products that that were uh that that were created in 2008 for example U do you agree yeah I mean I totally agree with Paul is I mean is this feeling that we are putting leverage leverage on on The Ether and I mean when the things are in the right direction everything looks good but uh the moment that you are increasing this liage um if something goes wrong it can be completely Sher

so at the end of the day I think it's also important how how much I mean how this grow this this raking I mean at the low level I think we can be more safe but if this become massive I think the the RIS that we are taking is like putting leverage on eum itself so I think we need to take thison um because yeah I am in this U chain effects that can happen is you have link everything so if something happen in one chain uh it can be a massive situation for the other ones and and it'll be a wave that destroy everything so yeah this is feeling that um in principle is a really

great idea I mean but when you start to think about all the consequences that is coming to the theem well it's the feeling is that it's it's really risky that's my my point the other larger topic that's currently very much um under debate is issuance so um kind of the idea that kind of if you already have a million validators having another validators or very limit marginal value whereas kind of if you only have say 10 validators having an 11th is uh of much bigger value so where where do you stand on on that debate so kind of how do you

measure the value added by each additional validator yeah uh the is is um is probably also hard topic I have heard Boys in different directions about this how how how to measure the Su s of a network from my point of view is how much you are able to make it decentralized uh other people is Mission or other things uh for me is if you are increasing the issu of the network you are getting probably more rewards so that means that more people are going to get into the ecosystem but the thing is

since the current network is is incentivizing more the ize way to do the things uh because these natural flows to centralized providers probably the thing that you are going to see is just more people with more money or that is my feeling so at the end even I I I prefer the opposite just to have people that really believe on the on on the core values and and not being here only for the rewards so probably I in the direction of of uh of uh reducing the not increasing theion a lot and but yeah I mean at the depends also I mean if you are able to to create a more decentralized Network that's my

bra and and that's my feeling I think it all stems from the impossibility to know um what what is a decentralized validator that actually adds more value like if we had a a proper way of measuring this um then it would be very easy like okay we we keep the centralized or we have some sort of like asymptotic rewards curve on on how many validators uh you you can have so the first validators increase your reward quite a bit and the like after like you reach a certain plateau of validators then your issuance or you like the eth that you mint on your

validators your consensus rewards uh lowers or tapers down the like of course of course this is an extremely naive approach and it would uh being at the completely decentralized platform what this would uh what this would do is like it it would give uh raise to like a lot of civil attacks and uh and people would pretend to have low validators so that's not something that's on the cards right now and like if we like I I honestly don't think reducing issuance is a good idea because um um there is a li like we see it in our community a lot of people in DB node um even though they

are like we we said that they are ideologically motivated at least in part they do also want rewards and it's not that they hate money you know like they they do like they do like their rewards and they want to maximize them and if the rewards go down above like below a a certain threshold they will stop validating but other big institutions or people who are less Hands-On or people that have less cost of opportunity um for for running their own validators those they they might be okay unless the issuance is uh negative so with a non with a non- negative or non-zero uh issuance or reward they they would be

okay on keeping this so I see reducing issuance in a naive way just reducing it um I see it also as a negative for the decentralization of the of the network now there are some things that that I think could be done so right now we have correlation penalties for uh slashing events um and it is suggested that we could have correlation penalties for missed attestations as well for being offline and I think it would be really funny if there would be a way of reward so there's there's this thing there's the unicorn unicorn chaos day I

think it's the status team that that implemented it unicorn chaos day is a day where people try to hack their own the things that they are building themselves they they it's just a day for testing security I think it would be amazing to implement something like this in ethereum something where if you can prove that you are making the an entire group of validators go down at the same time and me at the stations if you can prove that it was you you get part of these correlated attestation penalties you get them for yourself so you're still like only playing it's it's a safe game because you're still only playing with like mest stations you're not uh you're not slashing the principle so to so to speak um and it could be targeted at like players that have or

like yeah like more having more than 1,000 validators for example so if you manage and you can prove that it's you who like destroyed a data Center not destroyed sorry like a went like effectively um attacked a group of validators that that are an easy target because they're all together then you get part of the reward and that would be an interesting case where people would be like actively watch toering um the be W towers for for these uh centralized actors right and that would mean that it's less profitable to run validators in a centralized spot because you're risking yourself to being attacked and if you have to decentralize them in different in different spots that will

also increase your costs so it will be less that you'll be able to give less profit to the liquid staking representation so uh at the same time solo stakers do not are not affected because they usually run less than 1,000 validators so uh they would they wouldn't be a target for for these sort of people so effectively what this does is it increases the costs of securing their own infrastructure operations for big players and it has no impact on solo stickers I'd like to see something like that implemented so what do you guys have in store um over the next you know months or years with that

node yeah um uh I think the so we've been talking a lot about sing uh and the issuance right now is quite low and we're not seeing that much influx on on staking right so a lot of What's Happen happening internally in Dap node is looking at ways to leverage these Hardware already existing in people's homes um to give them access to uh privacy preserving Technologies or uh self- Sovereign Technologies in a way that they that that they're using that Hardware that's giving them some some sort of return um in order to provide more value out of it and here what we're talking about is we're talking about

running local models for example um we have a proof of concept where we have local models running on a dab note and you can communicate with them through a telegram bot the same way as you speak on a chat interface with chat GPT but instead of uh chat GPT and giving your away all of your uh data to open AI you're talking to your own local model right I think the the future of Open Source and local AI is is really brilliant because um we there are certain things that we are not going to uh we we are not so basically AIS are are this sort of like exoc cortexes like

this third- party brains that that we pay to access right and um like for some things like work um we're we're going to be using stuff that is that is the maximum the foundational model that's the maximum performance but at some point we will want something that is ours some agent that is not temperable some agent that really works for us and is controlled by us and maybe decides whether um we can ask a foundational model or uh we can we can have our own um or or we will use our own data to do to do something like this so that dab note I think is perfectly positioned to to provide those Services um and that's sort of like what

we're looking into right now super interesting yeah my point of this is just to think about a lot of applications that uh uh in which privacy and EA Mar like for instance imagine that um you can connect your local model to your Google Drive um and you can make um index queries on the data and ask for I don't know anything in your in your documents or in your photos or in your bank accounts or IM to connect um II to your uh emails and make uh Su of all your emails or your telam conversation or just to REM I mean I see a lot of applications in which um and I can help you a lot but it's so sensible the information that

you want to take a look is so sensible that you have no not you are not going to never connect to a centralized provider you are not going to bank accounts and this point and also because it's also part of the work Cod values mixing the Privacy with this technology that provides a lot of value I see a lot of potential application that makes totally sense and and can bring a new paradig on this on these new applications cool fantastic guys um where can people learn more uh about dap node uh download the software by Hardware even um and how can they be in touch with you guys right so um db.com is our website the

a node n o DB no.com that's uh that's our website and all of our socials are also listed there uh and we have a the super great Discord community that really helps both with the installations of the open source software and also with support and also like uh providing feedback for the hardware Etc and with guides on to on how to do things on DB node so we have an amazing Community uh our Discord is probably uh the best way of interacting with this community and see if dabn is for you from the website as well you can you can download the software to your own machine or you can uh access the shop and buy a plug

andplay Hardware that will bring dbot to your home without any technical knowledge perfect thank you both for coming England thank you for having us thanks again for inviting us here