Epicenter - Learn about Crypto, Blockchain, Ethereum, Bitcoin and Distributed Technologies - Marek Olszewski & Rene Reinsberg: Celo – The Mobile-First High-Throughput Blockchain

Episode Date: December 1, 2023

With an average block time of 5 seconds and sub-cent transaction gas fees, Celo focuses on scalability and ease of access in order to bring DeFi to those that need it the most. Celo’s vision of ensuring equal access is reflected through their mobile-first design, aiming to disrupt legacy TradFi. The upcoming transition to an L2 rollup aims to tap into Ethereum’s security, while maintaining minimal costs through Celo’s scalability and EigenLayer’s data availability solutions.We were joined by Marek Olszewski & Rene Reinsberg to discuss Celo's mobile-first approach to building a low-fee, high-throughput blockchain with real-world applications, and their transition to an Ethereum L2.Topics covered in this episode:Rene’s and Marek’s backgroundsCelo’s design principlesCelo’s scalabilityMobile compatibility for blockchains and dAppsWallet UXStablecoins as gas token & transaction confirmation timesOptimising mobile accessibilitySocialConnect & off-chain social identifiersCelo’s transition to an Ethereum L2Single block finality in an optimistic rollupSequencer decentralisationLowering costs of settling on Ethereum via EigenDACredit Collective & Impact MarketMento DAOThe future of CeloEpisode links:Marek Olszewski on TwitterRene Reinsberg on TwitterCelo on TwitterMiniPay on TwittercLabs on TwitterCredit Collective on TwitterImpact Market on TwitterMento Labs on TwitterSponsors:dYdX Foundation: The recently launched dYdX chain features new governance and token economics, that empower stakers and promote validator decentralisation. Bridge your DYDX tokens and contribute to the evolution of dYdX chain, fully permissionless and community driven. - https://bit.ly/47kqG59This episode is hosted by Friederike Ernst. Show notes and listening options: epicenter.tv/524

This is Epicenter, Episode 524 with guest Renee Reinsberg and Marra Golshevsky. Introducing the next generation of DYDX and the next version of the DYDX token. Welcome to the DYDX chain. New token mechanics mean you can stake to secure the network. Staking is fully decentralized and controlled by DIYDX token holders. All fees are distributed to stakers. Earn rewards from using the DIYDX protocol, with rewards planned for traders and early adopters. this too. New governance means you are in control. Trading has been democratized. You can vote on

protocol improvements, token distributions and more. Bridge your DYDX to seamlessly transition to the YDX chain. Bridge now at bridge.dG.DX. Trade and contribute to the evolution of DYDX chain, open source and community-driven. Run your own validator. Validating is fully permissionless. join us on our mission to democratize access to financial opportunity today. Welcome to Apple Center, the show which talks about the technologies, projects, and people driving decentralization and the blockchain revolution. I'm Frederica Ernst, and I am here with Rini and Maurek, who are the co-founders of Cello, a mobile-first blockchain,

with a strong emphasis on serving underserved communities. It's great to be here. Yeah, thanks for having us back. We had gone previously, but maybe we can do a quick round of introductions and giving a bit of background on who you guys are, how you met, and what makes you get up in the morning. Sure. Yeah, no, a lot has happened since we ran out from the show.

But yeah, I was just doing the mask the other day, Merrick. We've been working together for 12 years at this point. So Merrick and I started a company previously. We met at MIT. There was an AI company, though back that we didn't think about it in that way. But it was, you know, the common threat there was it was actually helping small businesses. So again, sort of, you know, sort of, you know, trying to support the little guy by better competing online. And we sold the company to GoDaddy and worked there for a few years, helped take Goddi go public.

and after that we went back to the drawing board and yeah a lot of the ideas that we were getting kind of excited about were things that you know we felt the technology to solve us was Web3 and so that was you know getting us into the rabbit hole so yeah going on seven years of working on

what has become cello which is which is super interesting Yeah, you definitely make us sound old in crypto when you say that. It's kind of hard to believe. Seven years that passed already. Yeah, maybe just to add to that, you know, I have a technical background at MIT where Reni and I met. I was working on a PhD in parallel computing and working on something called deterministic multi-threading, which is very coincidentally, has a lot of relevance for

scaling blockchain platforms. platforms today. And yeah, coming back to the origin story, you know, we initially started by building a wallet on top of Ethereum. That wallet has now evolved and become Valora, which you can play around with. But back then, you know, we wanted to make something that was just really easy for everyone to use, that was stablequin focused, that could be like a kind of a global, Venmo-like product.

And we tried for a while making it be sufficiently easy for normies, although back then we didn't call them normies. And ultimately, we realized that to do that, we needed to launch or to work with a community to launch a new chain that had some of the features that would allow you to build a really easy-to-use experience. Fast forward to today, the network is live and it's been humming along, really well for the last couple years. Valora is out and used in countless countries.

Yeah, the community and ecosystem keeps growing. And so it's just been really exciting. Yeah, super interesting. You just alluded to the fact that kind of cello had to be built in a specific way in order to kind of work phone ormys and on mobile phones. So what does that mean concretely in terms of technology? Yeah, so I think first and foremost, you know, scalability. You need a chain that can keep gas fees cheap.

Just last week, seller processed 20 million transactions in one day without raising gas fees. Processing, I think, on average 270 or so transactions per second for, again, for 24 hours. So I think that's a good example of just how scalable we've been able to make. and really it's the validator ecosystem has been able to make the cello platform. Secondly, we really wanted it to have a mobile focus. And so we worked really hard to create a ZKSnark-based-like client that can sync on mobile devices.

But I think even more importantly, we worked really hard to create something that we now call Social Connect, which is a protocol that lets you effectively connect your phone number to your wallet address in a privacy preserving way. This is an optional feature that some wallets implement, and the wallace that do implement it, allow you to effectively send value to people in your contact list, which is just a real delight for anybody who's not cryptosavvy. And then finally, we learned early on through various tests in Argentina

that paying for gas with volatile assets was just too confusing for a lot of users and so we actually built native support for paying for gas with tokens that doesn't require account of traction so this is supported by EOAs

and that allows users to send stable coins to other users who can then forward those stable coins onwards without ever having to think about it you know, what a gas currency is and how to get one. There is a whole lot to unpack here.

Maybe let's go through them one by one, because kind of I want to understand them in detail. So how do you make it scale? So kind of how do you get more transactions per second? Because kind of there's always kind of this scalability dilemma with security. How do you get it to scale to such lengths? Yeah. So I think first and foremost by having an effective.

and preface state consensus protocol. I think. And then secondly, working hard to ensure that EVM execution is happening quickly. I think both of these generally are challenges for EVM chains. And, you know, we've been chipping away

at kind of both of these. The reality is, on the consensus side, you know, there is, you know, obviously this trilemma, and it's difficult to achieve Ethereum level levels of kind of decentralization while achieving, you know, one block finality,

short block times, with a really high throughput. And that's why we're also really, really excited about this coming home to Ethereum, so to speak, through this L2 plan that SLO announced, or I guess that CLAPS proposed

and then the community voted on and agreed upon earlier in the year. And so I think by transitioning cello to an L2, we can inherit Ethereum's decentralization while still offering these really great features, these, this high throughput, this, yeah, just this high throughput that the chain provides. We'll get into that in just a second.

But maybe let's kind of go back to kind of the way that kind of cello started. So basically, are you saying that kind of the way that you could scale so efficiently was by having a fairly constricted set of validators when compared to Ethereum. So Ethereum has like 600,000 validators, right? So basically, how much did you have to constricted in order for you to kind of get these really amazing throughputs?

Cellos right now has 110 validators. And it's actually constricted not by the consensus protocol and scalability limits, but more by the ZK Snark-like client that we have. We built this like client, started working on it. I think four years ago, launched it two years ago, using, I would say, first-generation ZK technology. It's a huge achievement. We're really, really proud of it.

But it does use a very, very big circuit, two to the 27 constraints using a very big curve, PW6, and that is a very big proof that we have to generate daily on a very big machine, and we couldn't increase the validator set beyond the current limit without making that even more onerous. And so that is a current constraint. But as it so happens, you know, we think it's a really great level of decentralization, especially for a decentralized sequencer. as again, cellar transitions to an L2. Okay, so basically that means, like, users have a ZK-like client on their phone,

and that means kind of they don't have the full state, but kind of the full state is kind of like processed by 110 validators, kind of giving you that level of decentralization. Is that kind of reframing it correctly? So, yes, wallets can use, wallets and apps can use, we have some code that works using WASM that works even in a number, or wallets and apps can use like client, or they can use a decentralized RPC endpoint like Lava,

or they can use a trusted endpoint like infura and connect to the chain that way, absolutely. Okay. Why don't we see more mobile first blockchains? Because in principle, it seems almost like a no-brainer to kind of, that there would be at least some, because kind of if you look at the world as a whole, almost everyone is kind of primarily on their phones,

rather than on laptops or more powerful devices, why don't we see more of those chains? Yeah, maybe I can jump in. I do think there's certainly, you know, it's been surprised to us for how long it's taken for more, you know, for the broader ecosystem to kind of think more about mobile. And we're seeing this, you know, even if you look at sort of the kind of average transactions on, say,

uniswap or curve on cello, right, versus on like Ethereum Mainet, amounts tend to be smaller. People are accessing these daps from their, from their phones, right? And I think that's, there's a fundamentally different use case. And for a lot of the early builders, right, I think it was just more lucrative to focus on sort of more high value transactions, right, that traditionally still, you know, tend to happen on desktops, right? So for us, was really it started just with a being you know in a sense being very determined to try to make this work for everyone right and obviously like you said most people have mobile devices and it kind of extends into the ecosystem right where there's the the technical architecture of the chain

is one side but then also you know the kind of wallets that are being built right and there's a rich ecosystem of wallets that includes walora and others that you know have enabled in a sense like this kind of mobile first experience of Web3. But then it extends all the way into the depths, right, where people have built sort of much simpler interfaces, things that can actually be more easily accessed through a wallet, right? Or, you know, hooks into kind of wallets that kind of allow you to kind of intact with these protocols.

And I think there's a whole, yeah, kind of space for innovation that I think is, we're just starting to kind of see happening. And I'm quite excited about when I think about, you know, even going into the next cycle where my hope is that a lot of the, you know, usage and kind of activity is going to come from real world users, not speculating on NASA prices, but, you know, accessing these rails for, you know, getting access to credit, right, or making payments, commerce, all these things.

And that, in a sense, needs to have well, well, at the core. And so we hope that some of the stuff that we've built and we haven't, We haven't gotten through the whole list, so we should come back to some of these innovations like Social Connect that are, you know, or Fiat Connect that are kind of starting to be more widely adopted even, you know, with interest beyond the cell ecosystem right now with the L2 move, can pave the way for that to happen more quickly. Yeah, absolutely. Let's come back to these usability improvements.

When most people think about blockchain, kind of, it's difficult to kind of even explain to fairly tech forward people, right? it's like, okay, download this wallet. Oh, okay, this is your seat phrase. You have to write it down. You can never show it to anyone. Could also never lose it because otherwise everything is gone. So how do you get from there to a place where kind of you can't just give it like less technology forward people?

You can give it like actual others to God no means. I can give another example. I'm sure you have a punch too. But I mean, this is kind of a very recent example. So Opera, very successful browser company, right? I mean, their market share in Africa is like 90%. And Opera Mini is a browser that's built for Android and is extremely popular because they also subsidize bandwidth

for their African users. They built a seller wallet into Mini called MiniPay. And the onboarding is quite magical. I mean, people, I mean, they're on Android, right, so people can very easily back up their seat phrase on their on the Google account. And, you know, we've seen this since the launch of the product in mid-September. You know, no one has lost their seat race, right?

There've been like, I don't know, thousands. I don't want to reveal like sheet out numbers that they're not ready to reveal, but there's been a very high number of users restoring their wallet successfully without kind of a glitch. And that's a massive success for Web Street, right? That's something that a few years ago, we would have all been just very worried about, right? that that would just, you know, result in a bunch of, you know, messages or support calls or people just, you know, being frustrated and giving up maybe.

And I would say that's a problem that, you know, today really, I mean, we can sort of say it's solved, right? We have the tools to really onboard people that, you know, maybe other than sort of a browser or WhatsApp don't really even use, you know, much, you know, on their phone, right? And so that's amazing, right, because that opens up a whole new population to participate in this digital economy, which is, I think, why a lot of us are here. Yeah, absolutely. I had no idea that opera penetration in Africa is that high.

That's crazy. Yeah, it's over 100 million people. It's pretty impressive. And you said earlier that basically you let people pay gas in any token. How do you do that? Is it just via relay network or how is it implemented? Yeah, so it's actually implemented natively in the platform. So no relay is required, no paymasters required.

And the way it works is the seller blockchain client, P3 supports a new envelope-style transaction. Envelope transactions were introduced with the IP 59 in Ethereum. and this new type of transaction basically takes an extra field that is a address for a token that you'd like to pay for gas with.

And then we technically don't allow paying for gas with any token. It's only through a limit set of allow listed tokens that are governed through on-chain governance. Selo has a kind of a Dow that has a lot of different

functionality and one of which is, I guess, retaining this allow list. And then basically the client right before executing your transaction will check that the token is that the gas currency, desired gas currency is in the allow list. And if so, it will debit your balance in that token by doing an EVM call into the token to deduct how much you're bidding for gas. And then at the end of your transaction, it will do the same to credit you any balance that owed you for as a refund. And that's it. And, you know, naturally to help validators order these transactions relative to each other,

we do currently have each of these tokens have an Oracle that reports on the price relative to SELO. It's technically not needed because validators can choose to order these things however they want. using whatever price that they want, but we just wanted to make their lives easier and just have something that they can look at to make it easier for them to order the MMP pool. And, you know, the end result is really easy UX for end users, for wallets, for DAPs.

You don't have to have all of these extra complexity around, you know, reliers or new types of ways of signing transactions and additional folks playing a role in the, platform. Everything just works natively really, really easily. And honestly, it's been a really big success. People really like being able to cash in into CUSD and then be able to, you know, buy ETH on a Dex without having to go to a centralized exchange, right? If you think about kind of the dream of enabling fully decentralized, you know, capital D, D, DAP, experiences, then allowing people to cash into stable coins, ideally to a stable coin that's

pegged to their local currency, and then to be able to start transacting straight away. That's really a nice experience, a nice self-custodial experience that's possible on and with cello wallets that helps and support gaps and more deep. decentralization. Yeah, absolutely. How do you think about confirmation time? So how long is acceptable kind of for a mobile user who actually uses it for like a real world thing? Yeah, it's a great question.

So we did some testing early on and we found that five seconds is still feels snappy enough. And so Stella has five second block times with one block finality. that gives us more time to achieve consensus, which is good for processing large amounts of transactions per second. But I think realistically, given that there are chains with lower block times, I think as part of the L2 transition, we're talking about moving that down to something closer to seconds. But the end result for users is going to be pretty similar,

I think if you transact on the cello platform, everything feels very, very snappy. Cool. So you guys cater heavily to the global south. Are there challenges that you wouldn't have in the global north? For instance, I heard that that kind of device memory is often a problem. So that basically, for instance, people can't install new. apps because kind of they're out of, out of memory. And like problems we don't usually think about because kind of most people in our bubble

have like the latest iPhone or, you know, top of the line Android device. What has experience taught you? Yeah, maybe just expanding on the kind of mini pay mention. I mean, that was a big factor also, I think, where. when we started chatting with them about sort of this product, and they told us how they launched products into the market, right? I think it mapped with a lot of things that we had heard and learned. In pilots or projects, right, we have been a part of, as part of this hello journey.

And there were some really nice benefits, right? So I mentioned that opera mini kind of subsidizes users bandwidth, which means that if you have a wallet that's in the browser, right, you also basically get subsidized, you know, for your wallet use. So that's really cool. They built the actual wallets to be only kind of two megabytes package compared to, I think, like, I don't know, Mena Mask is like 68 megabytes and Corbiz wallet, like a lot more even then, right? So that's where you're starting to get into challenges when you have users that are living

from, you know, sort of 50 megabytes, 100 megabytes at a time, right, to get through the week and do all the things they need to do online on that. The cost of sort of updating or even just installing in your app to try it is kind of prohibitive for a lot of users. So being really mindful around that and, you know, orienting the experience around that has been, yeah, definitely kind of a key factor. And again, this extends to the ecosystem, right? And thinking through sort of every touchpoint, every bit of interaction, right?

Is there stuff that you can sort of streamline and make work better, you know, on sort of those constrained environments and sort of allowing devices? I will say, though, you know, I mean, as someone in the global north, some of these improvements also, you know, feel quite nice when you're, you know, kind of accessing things on the latest iPhone, right? So I think it's just like some of it is, you know, things don't need to be as convoluted, sometimes as they need to be, right? And sometimes simplicity, even when you are not constrained on bandwidth, can be kind of a nice, can be a nice forcing function for a better experience.

So, yeah, I think that's, you know, it's nice to have the, start with sort of the kind of, you know, the users that will have most trouble accessing the experiences that you're building as an ecosystem. And then I think that has nice sort of downstream effects in terms of all users, right? And obviously we have a lot of users in the global south, but, you know, there's also a lot of people in Europe and, you know, all over sort of the global doors kind of using selling products. Yeah, I think Luke Rubelsky wrote a book called Mobile First back in the day when Web 2 companies were thinking about their mobile first transitions and journeys. And one of the things he talked about in that book is about when you build in a mobile first manner, you know, you focus on bandwidth, you focus on, you know, bandwidth. the focus on, you know, band bundle size, all of these things that Rene just mentioned.

But then the end result is that everyone benefits, even people using your desktop applications. And certainly we've seen that in this early ecosystem as well. One thing that I think people all over the place will benefit from is this user identity linking to the phone numbers. What cryptographic methods do you use to ensure privacy and security in this? verification process. Yeah, it's a great question. And so we have this protocol. It's called Social Connect, which under the hood uses another protocol called Odis,

which is a threshold cryptography system. So we use threshold signatures on BLS12-377, that effectively allow a quorum of Otis nodes to respond to any request to look up a particular phone number or the address for a particular phone number. And by doing this, no note on its own can recover the phone number or the address for a particular phone number without a threshold, I think is 50% of these nodes responding. And I think that's really, really nice because, you know, certainly it offers privacy guarantees. It allows these nodes in a distributed way. It's not quite decentralized because it's a permission set of other signers, but it's a big set.

There's, I think, tens of these nodes. And they're all doing their own kind of attack detection to make it hard for any one attacker to try to identify all of these phone numbers. Okay, so basically, in essence, it's distributed set with a threshold where basically I can ask them for the private key that is somehow connected with your phone number. Basically, they give me that, but basically I can't go the other way around. Is that correct, or how do I think about it? That's close.

I think with the one changes, they give you. you the, not the private key, but they give you a sulk that you can use to derive the wallet address. And so the wallet addresses are stored on chain, but are hashed with a unique salt per phone number. And that salt is what this Otis server, Otis servers are giving you for any given cell number. And you need to request it from a threshold of these servers. Okay, and if I want to update my address, I can do that by contacting them too, right? Yes, you don't even need to contact them. You just update on the mapping because you know your own salt.

And so you're able to update the mapping yourself. And can you do that counterfactually? So someone who's not been registered with their phone number? can I send them, is it deterministic? Can I send them kind of funds and the threshold set can kind of know which

address the wallet's going to be deployed to? Kind of like create two. Yeah, it's a good question. So the salt is deterministic per deployment of Otis. We're on our second deployment now. And so

there was a KG that was done. So that's terministic, and that's the terministic, regardless of what portion of these notes respond, what 50% portion of these servers respond. But the second piece is registering your wallet address on chain, and that part is non-deterministic. You can register any wallet address. If I send it to someone who's not yet kind of signed up, it just goes into Nirvana.

So, right, the solution for that flow. is we have an escrow smart contract and so wallets can detect that the user hasn't yet signed up at the phone number and so they send the money instead to an escrow smart contract and then the escrow smart contract

will only release the funds to someone who signs up with our phone number and then wallets can automatically fetch the funds as part of onboarding. And so I think you're asking about what's the experience like for new for someone who's sending money someone who isn't yet

doesn't yet have a wallet and the answer is that the experience is exactly like you would hope it to be which is that you know the user can send them money anyway and then gets prompted to invite the the recipient you know to the platform usually using like a WhatsApp or a text message and then those users can then sign up and have the the fun claim the funds automatically. You said earlier that Sello announced over the summer that kind of it would, or there was vote over the summer that kind of it would become an L2 on Ethereum rather than remaining its own L1.

So specifically kind of, if I recall correctly, you're going the Validium route on the OP stack with like data availability with eigen DA. How did you arrive at this decision? I mean, first of all, kind of like the general move, kind of like of coming home to Ethereum

rather than remaining your own L1 because obviously it also comes with drawbacks, right? I mean, you do inherit a security from Ethereum to a certain extent, but you also pay for it. I mean, you avoid most of that cost by not doing, data availability on Ethereum natively, but it still comes with considerable cost, which kind of

when you look at serving the global South with kind of like low value transactions and so on, it must have been a difficult decision, no? Maybe, Merrick, let me jump in first, then you can go into maybe more of the detail. But, and I think you had mentioned this in the earlier part of the conversation, right, that's when we first started Settle, we actually built on Ethereum, right? And in many ways, this has been a constant, I think we've been obviously closely following everything that was happening in the Ethereum world. And the sort of idea had come up before, but this was the first time where, you know,

also to what you pointed out, right, there was a potential design where the end user experience, particularly around fees, right, would be workable in terms of not sort of creating kind of obstacles that would be too high to, you know, for the current sort of use cases on Salo, but also for the stuff that we want to continue to see on cello. So when that became clear, I think there was very quickly kind of excitement building within the community. And before we kind of went more public, was it in the sort of broader community. Yeah, I think we kind of did sort of our initial research on it and we can go into the details.

But I will just say, and a lot of this work is driven out of C-Labs. And so Merrick can comment more in the sort of play-by-play. But it's been just phenomenal to also see how, for us, it feels like a coming home moment, but also we've been receiving such a lot of support from the broader Ethereum community and a lot of people that I think had followed our work over the years. But now I think felt sort of maybe also more urgency or just seeing that there may be kind of more opportunities to kind of partner and collaborate, which has been really amazing to see.

But yeah, Marigal, I'll hand it to you for some of the technical choices in that initial design and maybe kind of where some of that is now. Yeah. And, you know, I think this has definitely been kind of a journey for the whole ecosystem and community. I think, as already mentioned, you know, we started off building an Ethereum. And ultimately, to accomplish what we wanted to accomplish, we felt that we had to work with the community to create a new chain, an EVM compatible chain, because we believed in Ethereum and we believed in the EVM. But nonetheless, till a new chain. And, you know, over the years, obviously, the.

roll-up-based scalability roadmap kind of continue to get refined, more roll-ups launched. It became very, very clear that Ethereum was succeeding with this strategy. And so given our excitement for Ethereum, you know, I think the community's actually been long talking about ways, you know, going as far back as 1.5, 1.5, 1.5 years ago, you know, talking about what could seller look like as a potential L2. And so coming back to this year, you know, C Labs, I think with the announcement of EugenDA, realized that, oh, actually there's a way for us to become an L2

while keeping gas fees cheap while also doing it in a very Ethereum-aligned way with Eugeny, DA, you know, effectively being run by Ethereum validators. You know, we felt that we could create an L-2 to design that preserves all of the features of cello, all of the benefits of cello, while still plugging into the Ethereum scalability roadmap and being part of Ethereum's scalability plans. And so I think that was this aha moment for us. And ultimately why, you know, C-Lab has put forward this proposal.

We wanted the proposal to be very concrete. and so we actually did the legwork to come up with an architecture that, as you said, leveraged OPSTAC, leveraged IGDA. We also thought long and hard a lot about one block finality, which is a feature that cello offers and we want to preserve, not, and one that's not trivial for L2s to offer. And so we spent a bunch of time on that. We put forward a very detailed architecture proposal that the community voted on. I will say that since then we've got a lot of inbound from a lot of L2 stacks

that have been pitching the community on other L2 technologies. And so even though C Labs has been building an OP stack since July, since that proposal passed, there continues to be conversations at the community level about maybe there's other ZK-EVM stacks. that might be a good fit for cello. And so that's been exciting to see. It's just being really great to see, not just the whole cello ecosystem

get really excited about this, but the broader Ethereum ecosystem, the reception has been just absolutely amazing. So it's been really, really great. How does single block finality kind of mesh with fraud proofs and kind of, maybe it's a false dichotomy, But in my head, either you can kind of, you get shared security from Ethereum

and then kind of you need to let people, to leave them a pass to actually access Ethereum security layer by kind of escalating a conflict on the L2, or you get single-slot finality. How do you tow that line? Yeah, it's a great question. And I think the answer relies on this kind of fundamental observation that for finality, the fraud-proof window is actually not important, and neither is, you know, ZK-proofs necessarily.

Ultimately, finality for most L-2s is accomplished by posting transactions to the L-1, and as soon as they're on the L-1, they're deemed in a final, right? Any full node following the chain-deriving blocks on the chain will, will deem those blocks as final. The fraud-proof window is actually irrelevant. That's mostly for the bridge itself, for the native bridge. Yeah, but not so much for finality.

And so there's two sources of potential reorgs that can happen in L2. One is from the L2 itself. So if you have a sequencer that tells you that the next block will look like this, but then submits a different block to the L1, then L2 clients will have to reorg when they, you know, first they'll get the first block and they'll assume that that's kind of an unsafe

but likely block to get promoted to a finalized block, but then when they see what actually got posted to Ethereum, they're going to have to reorg and roll back. And so that's one source of non-the-term of reorgs. And then the other source of reorgs is just a, L1 itself, right? If Ethereum reorgs, then the L2 naturally has to reorg. And so to address this, we came up with a design that, one, it has a decentralized sequencer, and two, relies on slashing for equivocation by that decentralized sequencer. So if the, in order for a

transaction batch to be kind of sequenced, you need two-thirds of these sequencers to sign off on it. And if they ever sign off on another transaction batch and post-ad to Ethereum, then that's easy to prove and a slashable condition. And by having more than one sequencer, you can actually have more stake that is at stake. when these folks misbehaved.

And so that allows you to effectively have what we call cello security to offer a finality guarantee that can exist in a shorter window than whatever period by which batches are submitting transactions to the L1. And then secondly, you know, for the L1 reorg, the solution there is to effectively weigh for transactions to be finalized on the L1 before they are imported into the L2. So that means that any forced transactions, any deposit transaction that happens on the L1.

We have to wait 12 minutes effectively before those transactions appear on the L2. And that way, even if the L1 reorgs, those transactions won't have appeared yet on the L2. and so the L2 doesn't have to reorg. So those are the two changes. And I was talking about this in the context of a roll-up posting transactions to the L-1, but everything is very similar, even in the case of the L2 using an LTA like eigentya. I think that makes a lot of sense. Do you use the validator set that you had kind of as an L1,

and kind of for the decentralized sequencer, or how do you construct it? Because kind of that's been a really hard problem for all L2's kind of sequencer decentralization. And by, in a way, you already had, like, I mean, if you think about it in those terms, kind of the sequencer is kind of your validator set, right? So basically the sequences, that entity

that's allowed to kind of actually build the blocks. and you already had a 110 distinct entity. So how do you enmesh these two things? Yeah, that's a really great question. And the answer is 100%. We will transition the valider set to the decentralized sequencer set. There's a lot of benefits in doing that. And it just ends up being a much more elegant migration for the ecosystem.

And things like slashing for equivocation, we actually get for free if we do that because we already have slashing for equivocation on cello. And so all of the finality guarantees that we offer today, basically the L2 inherits, and then adds to that Ethereum security and finality just with a longer time period. So if you're doing a small transaction, then maybe you're happy to wait for the cello sequencers

to agree on a transaction batch. But if you're doing a big transaction, maybe you wait a little bit longer until that transaction. the existence of that transaction gets logged on Ethereum. How often do you settle to Ethereum? Because, I mean, so basically, if you look at the costs of L2s, I mean, it depends how often you you kind of check into Ethereum, but typically, like, 95% of the transaction costs kind of go to data availability, and then, like, 5% go to, like, checking in.

Obviously, you can kind of lower that by kind of checking in less frequently, but then, kind of you have security later. So how do you navigate that? And did you decide kind of on a ceiling price, say, for 100,000 gas or something that kind of you think would be acceptable to your average user? Yeah, it's a really great question. So we haven't decided on the period or the price right now. I think we're still doing that work. And one nuance, I guess, for us is that we're storing the transactions in IGDA.

and an eigen-DA reference kind of pointer back to us, and we have to store that on Ethereum. And that is much smaller than obviously the set of all transactions, and so the gas costs of storing that is much, much, much less. And so that's how we can keep our low gas fees, is basically by replacing that 95% cost that you just referenced with a much smaller cost that is likely going to cost. you know, something like 5%.

So we might double the transaction fee on cello, but we won't, you know, 20 exit as part of this transition. The exact period by which we'll be posting that, I think, also depends a little bit on eigenda. Igen, DA. Igen, DA just went to Tessnet two weeks ago, and so we're still waiting to see what the production specs will look like, which will definitely impact kind of the design choice here.

So the danger with being an L2 on Ethereum is obviously that kind of your transaction fees are directly coupled to Ethereum transaction fees, right? Because you kind of have to, you have to pay prorator the kind of check-in costs of cello on Ethereum. And currently gas fees are pretty low. But we've seen them spike in the past. And if you kind of expect Ethereum to kind of eventually become the global settlement layer. And if you look at kind of the scaling roadmap of Ethereum, there's not actually, I mean, beyond dangsharding, which doesn't help you all that much because you're already using eigenya anyways.

There's not all that much to seriously speed up scalability on Ethereum in the next couple of years. How do you think about that kind of in terms of the applications that kind of rely on fairly low fees? Yeah, I think this design caters to them. So I think by only having to store the eigen-DA ref, you can dramatically, dramatically lower the fees. And certainly you can also change the period by which you're storing this. So that might also potentially give some wiggle room for the chain. But yeah, I think given this design, I'm not too concerned even as Ethereum prices.

go back and go beyond where they were in the last bull market. I think it goes back to the earlier question as well, right? I think, you know, you can offer cello security, right, for some, like, for a lot of transactions, right, that kind of are below a certain threshold, but that have Ethereum security for sort of the more valuable transactions, right? and sort of, you know, I think in many ways this is a concept that we're pretty familiar with even kind of pre-Webstree, right, in terms of like, you know, the stuff that's in my physical wallet, right, kind of the security of that versus, you know, what's in my bank account and so on.

So I think it's kind of, you know, and I think this is just a very simplified view of this, but I imagine as some of the L2 infrastructure kind of hardens, right, and the improvements that, you know, all the teams that are focused on this are making, even since July, since we kind of openly started kind of that exploration has been pretty encouraging that there's a path towards continuing to lower that cost. But even if you completely discount kind of like all of the data availability costs that you incur, kind of just by settling on Ethereum periodically, you kind of, you have to pay that fee kind of as an L2 and you kind of have to kind of, kind of have to kind of.

of put this on the single transactions that kind of happen on cello, which does actually put a floor on how little a transaction can cost, right? And kind of if costs on Ethereum go up, kind of the costs on cello have to go up by the same amount, just by virtue of kind of ultimately settling on Ethereum. So I think, so you're right, there is a floor, but interestingly this floor is not tied to the number of transactions in the L2. It is for roll-ups, right, because you have to store transactions on chain.

But if you're using eigenDA for data availability, then the number of transactions that you put in any given block can be arbitrary. The size of this eigen-DA reference is fixed. Likewise,

when you update the state route for the L2 bridge, the amount of transactions that executed between state route updates can be arbitrary. So if it came to that, you would just kind of prolong the period you'd wait until kind of updating the state route on Ethereum. Yeah, I would be surprised if we would have to do that because I think even at high gas fees, just based on the amount of economic activity happening on several today, like I don't. I don't think that that would be necessarily required.

But yeah, absolutely. I think more than that, I think we would look to scale cello so that it keeps up with Ethereum gas prices, I think, is the answer. In principle, it also seems like a good route for kind of other chains to kind of decentralize their sequences, right? Because, yeah, it is a major problem. But maybe let's just move on. So we already talked about MiniPay,

the application of OperaMany that kind of gives users natively a wallet without them kind of having to think of it as a wallet. You also recently had something launched on Cello called Credit Collective, which I found very enticing. Can you talk about that? Yeah, I mean, one of the use cases

I've personally been, and I think many in the ecosystem since the very early days, is giving people access to credit. And so this actually was really an entirely led, kind of community-led effort where a bunch of people that were working on credit protocols got together and started meeting and organizing. And, you know, some of the activity was centered on cello, but it wasn't like, you know, that wasn't sort of like, oh, you know, you can't, you know, can't be here if you're not on cello. So actually, a lot of the early members were kind of building in different ecosystems. And one thing that was kind of interesting was credit collective. It started really just as a bunch of people that shared that kind of interest

in bringing these solutions to market, meeting up and exchanging ideas without sort of it's starting, you know, like usually a lot of these things start with like, you know, funding or some other thing right here. It was really just, it was kind of like a, you know, a group of people getting together and exchanging ideas. And then at some point, Credit Collective sort of formalized that and actually applied to the Selo on-chain funds for funding, for effectively funding to help bootstrap adoption of these protocols. So they requested, I think it was a total of 2 million cello euros, kind of cello backed stable coin or kind of linked stable coin, pegged stable coin, to be used as kind of seed capital on these protocols to extend. in many cases, what is, you know, direct to consumer or kind of, you know, small enterprises

in the global south, you know, kind of credit. And just to be clear, these are, you know, for the most part, under-cradelialized or non-conecordialized loans, so not sort of the ones you would get on Avi or compound. And really sort of extending, I think, the utility of WebStre, right to kind of the real world, which is something that's really exciting. Yeah, I don't know if you had specific questions about it. we should also, you know, maybe to connect you to some of the people working on that. But I've been pretty excited by seeing something like that kind of take form. And, you know, so that self-organizing and, you know, creating sort of funding umbrella to it, you know,

was really cool to see that happening. Yeah, and it's been cool to see some of these projects take off, right? So Impact Market, which is a UBI project on SELO, research. he launched a micro-lending campaign. I think they have hundreds of thousands of dollars now that they've been lending out in the form of micro-loans. You can think of this as being analogous to Kiva in the Web2 world. I think Kiva has had a lot of success

in having this kind of donate-to-lend kind of strategy, and I think it's had a lot of impact globally and impact market is now replicating that, by using crypto rails, allowing them to operate at, you know, a fraction of the overhead of something like Kiba. And so, yeah, it's been really great to see. I think these are the types of use cases that we have long talked about in Web 3 and that have gotten a lot of people, you know, myself included, excited about Web 3 early on. and it's just, you know, so nice to see them, you know, finally working on a large scale. Yeah, it's been great.

Are these sort of local initiatives? So kind of is this kind of like, say, for instance, I don't know where impact markets or kind of the keyword equivalent are based, but are these kind of bi-locals for locals or is this kind of by well-intentioned global noisness for the global shows. It's a mix. So there are definitely projects that are originating in, you know, narrowly manila and, you know, sort of right where the activity is. But it's but it's also, yeah, it's people coming out of the Kivas, right?

And wanting to bring that experience to solutions that I think have the potential to scale, a lot better and leverage sort of the infrastructure that WebStreet can provide. So it's actually interesting because I think that cross-pollination is also something that I think benefits both, right? It brings more of that local expertise to, you know, maybe the teams that are coming at this, not with that first-hand experience. And also, you know, the other way kind of helps the teams that are kind of local, right, kind of exchange ideas with, you know,

in terms of like, yeah, product development, other things that are maybe, you know, harder to kind of find peers in sort of local markets, especially around some of these very specific use cases. So, yeah, that's been fun to see. How do you foster and engage with your local communities? I imagine that, I mean, you are both, you know, global north and us. So kind of how do you build communities in those places?

Look, a lot of it happens. at events, community events, sort of, you know, nothing is replaces in person. I think in many ways, Merrick and I joke about this, but the pandemic, I think, has probably, you know, put us, you know, back,

you know, kind of definitely quite a bit, right? Because in those years, it was impossible, kind of, to travel, right? And spend time with communities on the ground. But, I mean, personally, I try to spend time, you know, going, going to meet with local entrepreneurs and like even kind of see how the products that they're building being used and just try to stay as close to that as possible because I think it does

help you know when you talk to enough entrepreneurs and they tell you kind of their honest opinion of what needs to improve for them to be successful right that kind of gets you the clarity on sort of the more infrastructure level right what are some of the things that are not you now that maybe you can help sort of move, move on, right? And so that's kind of how things like Social Connect or Fiat Connect, which is, you know, protocol for helping people on and off ramp more easily in these different markets have come about. And that's, I think, something that you only, I think, do get a sense for if you do spend a lot of times with the builders in local markets. And I assume you kind of, you have community people that are employed by the

foundation or one of your companies kind of like in in in in those local communities right we do yeah i mean a lot of the team members also that are working you know at c labs or walora or foundation or mentor you know some of the kind of early ecosystem projects are also based uh you know kind of all over the world so there's there's also a lot of that kind of expertise um but i'll personally say i you know i try to spend as much time as possible with with with founders in the ecosystem and And, you know, I think that's been really, I think for me, the best way to stay up to date on what's happening. You just mentioned Mento and reminded me that last week you guys announced that there's going to be a Mento DAO spin out. Tell us about that.

Let's get into it. Yes. I mean, for people that have followed Sallow for, you know, since the early days, they may remember that through on-chain governance, a series of states, Cable coins where we're launched, Cil the dollar, Cill Euro, Cil Real. And, you know, that's become definitely one of the differentiating factors for the ecosystem. And I think we've been very fortunate that there's been a team that basically said, hey, you know, we want to go and grow this much more quickly than, you know, I think otherwise would be possible as just kind of a public good under the sort of general

ecosystem umbrella and basically created a proposal or kind of a process for taking this protocol and really making it standalone. And yeah, very kind of recently just, I think it was got last week or so, the team has put forward a proposal to kind of start talking about how they envision governance to transition from SELO to kind of a mentor token to do that. So pretty exciting to see, you know, in a sense of taking some of these things that, you know, initially maybe we're kind of smaller features that may tell a little unique that are becoming protocols in their own right. And I think, you know, because of that, are able to have a much more ambitious, you know, kind of agenda and potentially impact on the broader ecosystem as a result.

So that's been, that's been really exciting to see. Yeah, especially just thinking about this in the form of just Dow governance, right? I think having a potential Dow spin out, I think it's just an exciting thing to be happening. And so, yeah, I'm just eagerly awaiting next steps on that. It's pretty cool. Yeah, super cool. So you guys have been around with Sello for, say, six or so years. Where do you hope to be in six years from now?

Wow, probably still working on cello. I really, I do feel there's with every sort of, you know, inflection point in terms of the technology or, you know, how the ecosystem's growing. I think there's, you know, you kind of, it's like you're climbing a mountain and then you kind of, you get to the point and you're like, oh, man, there's all these kind of new pathways to explore and like different ways to grow. And, and so I really, I mean, our.

I think we talked about this in the past, right? Sort of the cello as a name comes from Esperanto means purpose. And, you know, early on, one of the first things we did, even before writing a land of code, was to sit down and write down, you know, kind of mission for what we wanted to build and a set of values. And I think that has really created a lot of space for within that what cello can be. And of course, now it's no longer just the two of us and Zebauer, our fellow co-founder and some of the early team, but it's a whole community, right? And so a lot of the stuff we kind of get for free on wherever people are sending us much,

hey, I built this on cello. It's like, wow, yeah, this is like so cool. I don't think we would have ever gotten to this, right? So that's nice to see. But I think there's still a lot of white space, right, in terms of if we take sort of prosperity for all as sort of the guiding mantra, you know, that can mean a lot more things than the things we're able to execute on today, right? And I think that's really exciting because I think the metric that is probably least talked

about in these things is sort of the brain trust you can build around the project and sort of the ecosystem, right? And I think about like some whatever, you know, recently we were kind of going deep down the rabbit hole on D-Sai and I just, you know, happened to be in our community space here in Berlin. And I was like, there was a founder and was like, hey, have you talked, have you looked at this? And immediately like, oh, yeah, you know, here's some ideas and immediately pointed me to some kind of other things.

And it's kind of like, it's incredible to see how much more quickly one can move as an ecosystem on some of these pretty hard, difficult kind of concepts and ideas, right? And some of these things may take years to get to full fruition, right? but it's possible when you have kind of a mission aligned group of people that are all sort of, in a way, like, oriented around the same kind of shared set of goals. And so, yeah, six years from now, probably, you know, like, yeah, working on things that hopefully expand sort of what SELO can mean in terms of bringing more prosperity to more people, but probably also some very, you know, like stupid things that we thought would be solved by then.

you know, I mean, back six years and we're like, okay, on off-ramping seems pretty easy, but ultra-lact lines seems pretty hard, right? And guess what? We're still tackling on off-ramping, right? So I think that's sometimes also hard to predict, but I think a lot of, you know, whatever it is, it will be under sort of that kind of general umbrella of the Mission North Star. Maybe just to offer my take, people are frequently surprised to hear that Tron is the number one chain by daily active users. and they do that primarily because there's a lot of payments happening with Tether on Tron globally.

And so my sincere hope for this cell ecosystem and for the Ethereum ecosystem is that, you know, we can reclaim, the Ethereum ecosystem can reclaim that top spot. And, you know, that seller can be the L2, that where, you know, that type of payment activity is happening. We're already seeing a lot of that happening on cello. I think we're frequently a top 10 chain by daily active users. But let's work together as an Ethereum ecosystem and take the top spot. We don't need to wait five years for that. We can do it sooner.

That's a lovely closing remark. Where can people go and find out more about Zello, how to build on you, or how to kind of use products that I build on Sello, or kind of just get involved with the community. I mean, definitely celloorg.org. So our website for the ecosystem. Twitter as well, I think is sort of one where the, you know, we try to kind of capture things that are happening in the ecosystem.

There's regular community, you know, kind of meetings, some of them kind of organized by the community. usually at most of the major conferences there's, you know, Ethereum conferences there's going to be some event or meetup as well. My recommendation would be if you're someone who is interested in SLO or getting involved in the ecosystem,

just, yeah, find out where the next event is and show up or ping someone, you know, in the community, a founder even. I think there's so many touch points and I think I'm, you know, happy to vouch for, you know, really, you know, the ecosystem founders that they're all a friendly bunch and, you know,

will be very happy to chat and, and sort of give their perspective and help and support new people coming into the space. And, yeah, I think those are probably the sort of more general things. I do probably also want to plug, unplugged, which Merrick, you started with Kobe a while back just to share updates on the progress around the L2 work. and where you're sort of bringing folks on to give sort of more detailed kind of tech deep dives on some of the different, you know, kind of the different areas kind of across that stack

that that's being considered. And I think, I mean, I even like every time I listen, you know, and those I learn something. So I highly recommend that even, you know, if you're not just interested in cello, but just generally following L2 sort of developments in the broader ecosystem. Super cool. Thank you both for coming on. It was super interesting. Thank you. Thank you. Yeah, thanks for having us. This was amazing. I love the questions. Thank you for joining us on this week's episode. We release new episodes every week.

You can find and subscribe to the show on iTunes, Spotify, YouTube, SoundCloud, or wherever you listen to podcasts. And if you have a Google Home or Alexa device, you can tell it to listen to the latest episode of the Epicenter podcast. Go to epicenter.com.com. Go to Epicenter.com. for a full list of places where you can watch and listen. And while you're there, be sure to sign up for the newsletter, so you get new episodes in your inbox as they're released. If you want to interact with us, guests or other podcast listeners,

you can follow us on Twitter. And please leave us a review on iTunes. It helps people find the show, and we're always happy to read them. So thanks so much, and we look forward to being back next week.