How To Fix Crypto's UX Problem with Rachin Kalakheti

Transcript

Settlement is like a very very like a debated term in the space. Um but it's a settlement is very very asset specific. How does this benefit the end user from a user experience perspective? So do you think Salana is going to flip Ethereum? Welcome back to the rollup your home for technical talks and open conversations that shape the modular future. Nothing on today's show is to be construed as financial advice. You can see our full disclosures list at the rollup.co/ventures. Sit back, relax, and enjoy today's episode after a quick word from some of our awesome partners. Innovating finance global institutional and onchain. Deep liquidity to drive capital efficiency.

Now maximize yields on institutional grade liquid staking with me. Unlock trillion dollars of capital via Fitbitcoin. All of this secured by Mantle's treasury of billions in stables and ETH capital pools. This entire ecosystem is coordinated by a vast network of seasoned VCs that are aiming to create the next generation of banking and financial technology. This is Mantle. All right, Rashin, welcome to ETH Denver. You're it's your first time on on the rollup. So, who is Rashin and what what are you building in the space?

how how did you get involved in crypto and uh yeah maybe just a quick introduction before we jump in. Yeah, 100%. First of all, thank you so much for uh inviting me here. Um I must have done some good deeds in my past life to to be here. It's the it's the double earrings, bro. Fresh. Thank you. Um so yeah, a bit about myself. I I have been in like you know the whole computer science development space for a while now. started out as a game developer actually probably half probably um yeah seven eight years ago. Um and uh yeah I grew up for uh most of my life in Nepal and the way I got into crypto was when I

was trying to publish an Android game to Google Play Store. This was 2018 and Bitcoin had just like gone from 19K to 3K and it was just about to bounce back. Um and at that time I was trying to publish an Android game but the problem was to open an account in a Google Play Store um for publishing any content you have to pay like $25 fee or something and the problem is Nepal does not have an international payment infrastructure. So like even if I want to buy a Netflix account today for example I have to ask someone living abroad to do that for me.

Um so that still to this day I think so I'm I'm not too sure today but at that time it was like that. Um so pretty crazy and at that time um I was trying to find some other way to make that payment and one of these guys um on Facebook group chat randomly told me that he could help me buy the account using um Bitcoin on a gift card site. Um, so that was my entry into like what Bitcoin actually is. And uh, ever since then just been in this rabbit hole of like going from Bitcoin to Ethereum, building some applications and uh, yeah did a lot of things in between but now building Twine. Awesome. And what is Twine for uh, anyone who's unfamiliar?

Yep. Um, so Twine is a new way of building applications I would say. Um and uh the idea like the the technical idea behind it is like as a network instead of settling on a single chain you if you settle on multiple chains what you can have is you can tap into distribution of all these major chains at once. So for example you've got constructs like a single chain L2 for example like arbitum optimism base etc. What what that construct allowed for them to do was tap into the distribution of Ethereum that it had already collected, you know, the community, the asset, the users and all that. And uh it made their life a lot easier um while

they uh went on to create their own ecosystem, right? But the thing is um distribution is not um like permanent in crypto especially like if you look at um 2023 for example like Ethereum applications and rollups were printing um and Salana were like was like like applications was not there right like they were basically dead considered dead at that point early on in 2023 but you come to 2024 you see like this complete switch from the distribution going from Ethereum to Salana um so it's Like if you're building an application now, the thing is if you choose like if you choose one chain or like one place to tap into the distribution like there's

this risk that okay that might not be relevant tomorrow and all of that may flow to the other place where you're not connected to. So so that's really so to make the lives of applications easier is is what we do. M so you're saying basically you can have a uh settlement agnostic approach to uh where you are basically like getting your security properties from as a as a chain. Exactly. Security and like you know liquidity and anything like you know distribution related. Okay. So we've been exploring this. So okay first before I get into that what is your Okay. So what is the mental model that you have for the future of of like this module ecosystem Ethereum rollups L2s etc. that led you to believe that this

is something that you want to uh build out? Like was it purely about distribution or like on the tech side is is there something that you've like like I'm just curious what your what your mental model for the future of the space in terms of the scaling is that led you to believe that this was an idea to pursue. Yep. So I guess I can give a little context about how we ended up uh coming up with the idea and then built it out. Um so before Twine we're actually building Sedro finance which was a crosschain lending and borrowing protocol at that time and uh this was like one of the first applications to be built on layer zero and we later on like added other message passing protocols as well but uh what we realized quickly was um for anything beyond a simple token

transfer across chains um building a crosschain application is really really hard. It's a pain in the ass for application developers because now when building the core application logic is already hard enough. You have all these infrastructural side of things that you have to worry about like you know handling all the asynchrony non-aututicity um and also having to maintain the infrastructure for multiple chains and if that other chain has a different VM you have to rewrite your application in that VM as well. So it's it's a huge pain. So we realized that and uh and then even then like you know we we did we did hack around and and then went to test net with that and the thing was like the user feedback was like okay you know using it is hard because you have

to keep switching networks you have to um like and have your funds available on multiple chains. I remember I remember using like um synonym finance and like Pike which were similar ideas and it was like you had to like choose which USDC or ETH on which chain that you wanted to do and you you always had to have the right number of on each one of them. Yeah, it was really weird. Yeah, it was a very weird user experience, right? So, so like you know that led us to think like okay there's some deeper problem that we can solve here and uh yeah that that's really when we got the idea of like okay if we can have this one platform that sorry application can build on top of and that platform handles all these complexities of um you

doing the crosschain transactions um then that makes the lives of developers a lot easier right and uh and now they don't have to like you know if if for example like you know tomorrow um a new chain or for example like let's say uh tomorrow um Aptos for example makes a comeback uh now you don't have to and you're an application on Salana for example now you don't have to um go and think about how do you deploy on apttose for example um like you you just let that let let us do that um and uh you just don't have to change anything on your part but tap into the distribution of uh wherever users are. Got it. So how so how does this technically

differentiate for something like a layer or like or like the super chain where it's kind of bringing these different chains together like are are you guys building this like shared settlement layer as like it its own blockchain to facilitate this or or kind of how does this how does this technically like like an architecturally work? Yes. Yes. 100%. This is a great question and I think this is something that I've been asked about a lot uh in terms of like how we differentiate from things like a layer super chain um and even shared sequencers for that matter. Um the thing it's like quite quite different actually it's it's quite complimentary I would say more than it's competitive because like the games like the the the games we're playing here are at like completely different levels. So, egg

layer, um, shared sequencer, super chain, they're all, um, playing at like more of a sequencing and execution layer of the stack. Um, sorry, more of the sequencing layer of the stack um, and block building level. But for us, we go down the stack and uh, think about more from the execution uh, layer part of things. And uh the thing is um with a layer or super chain the the construct is very like a shared collective kind of construct right. So you have this one uh settlement layer or like settlement contract that has been written by a neutral party and now like you know any new chain can come in and opt into that settlement contract which is what allows a lot of the uh unified behavior across

all the chains that is involved right. Um but for us it's like quite of an inverted approach actually. So you we've got this one chain um that is now settled to multiple chains. So, so now the the the interesting switch that happens here is like instead of chains coming in and opting into this shared collective, this chain that is us, we go and opt into um all these chains that we want to tap into. And then the there's like I'm not saying one is better than the other but it's like a layer or like super chain kind of construct works very well for um things like longtail chains where like they want to attach to some major um ecosystem to um bootstrap their

distribution. But for us it's it's more about um going in and choosing the major chains because it's not a very scalable design right you cannot as a single entity go and deploy on all the chains that exist out there. there's like a lot of technical um hurdles there because of that but uh it doesn't really matter for us right the goal really is like tapping into the major sources of distribution and we've started that out with uh chains like Solana Ethereum to start out and been looking into places like Bitcoin as well which is a very interesting landscape now uh with the development of BitVM and and all the stuff where we can now actually verify ZK proofs on Bitcoin um like not not quite there yet, but almost there. So,

just to close the the loop on the mental model, I know Rob's got something here. Basically, Twine has its own chain that is plugged into uh Ethereum and Salana right now and could be added into Bitcoin and other kind of uh base layers as like this like as this chain that can settle to any of these uh base layers. And so you can come and build applications on Twine the chain and then you can choose which chain you want to settle to or you can choose where you want to get your like kind of close that loop for me like if I'm building an application on Twine what what why like do I get to choose where I settle or how does that work? Yeah. So so there's like two parts to this. This is like

settlement and the data availability part of this. um Selman is the cheapest part because uh you're just posting the meta data of uh whatever happened as like a like very small summary of the the batch. So so for example like you know settlement on Ethereum, Salana, Bitcoin or any other chain like you know the the aggregate of that doesn't um account for that much and then costwise. So um even the construct wise it doesn't make sense to isolate settlement layers from the applications perspective but DA is where like the actual cost is um because like you know it scales linearly um so so like that that's really where we've decided to go with the violation model. This was the uh design that was

introduced by Starkware four or five years ago if I remember correctly. And the idea is like instead of um you enforcing a single data availability layer as a chain, you allow uh the storage slot holders that can be an account um a smart contract or whatever uh to choose where they want to post their data to. So now like you're choosing the spectrum between livveness and cost. So posting into Ethereum might give you better livveness guarantees than posting on um maybe Celestia or Salana for example but um you got to pay the cost for that right. So so now like you know depending on how much you value the livveness for your application you can choose your place in the spectrum.

So so that's really how we think about it. The mental model that that you laid out makes perfect sense. Yeah, you know, you have you have twine and then theoretically you could have, you know, another chain, call it, you know, Twine too, that also, you know, multi- settlement, right? Um, and so whether it's Twine or Twine too, you know, the the uh the chain is submitting these uh you mentioned metadata, you know, we could also say that maybe there's ZK proofs that are that are sent to these settlement layers and verified on these settlement layers. Um the and I I think it makes a ton of sense. The one hang-up that I have is when we're submitting these proofs to multiple layers. Um what happens in the case of a roll back what what happens when you know Salana goes

doesn't find anything malicious but Ethereum for you know has some sort of uh contradiction with with the state you know how does Salana get notified of that roll back how do you reconcile the two? Yep. Yep. So that's very very interesting uh question and something that we spent a lot of time thinking about. Um and for that um like what we do is like we've created this like very virtual consensive construct consensus construct where we call it like meta consensus which for which we use like these underlying sment layers as a way to as as like validators for this virtual consensus to come to an agreement on a given state. Um so for that for us that is like the latest

stage for a gi latest state for a given batch. Um and let's say like you know I'm a sequencer. I'm posting the ZK proofs on Ethereum and Salana and Bitcoin um for batch B. And now in batch B+1 what happens is it has to submit like the sequence has to submit this proof that the the state for batch B across all these chains are the same. Um that should be embedded in within this proof that um that the sequence is submitting. So that way like we can always make sure that that the chain uh is always in sync and if like you know one of these chains roll back that's really when where we introduce this idea

of like voting weights where um we can allow so it's like very similar to how classical BFT consensuses work right so you've got one um threshold up to which you can um tolerate fault and then beyond that you allow a fork to happen so so like you know uh let's say like you know we reach 90% consensus on um on on a given state then like you know that 10% might be like depending on how you design the construct right for permissionless um consensus mechanism that number can be a lot higher and for permission consensus mechanisms like ourselves um that number can should be a lot lower because there's like less room

for error um and uh and like okay let's say like you know the the number we set is like uh 15% % and there's disagreement um of like only 10%. Right? So in that case we can enforce the majority view uh onto the minority view. So like if 90% of the chains say that it's like uh state S and the the 10% minority says it's state S prime then we can enfor you we can use the proof of that to enforce the state S on the chain that says the state is S prime. But but like if if it crosses like 15% mark which is like okay let's say it u 20% of the chain say that it's state s prime and 80% say that it's like um state s right so in that case what happens is

like there's like you like we we've now kind of gone into this zone where we don't really objectively uh know what exactly is the truth right so so in that case we what we do is like we allow a fork of this L2 to happen um where now that we've got two versions. Uh in this case, if there are if there were like n number of states, then there would be n versions of this uh network. Then now like what happens is like like for 80% of the chains um that is um so like your asset is always on on that bridge contract. Okay. So you as a user are not facing a security issue here because um

if for example it was like Ethereum Salana and Bitcoin, Ethereum and Salana was saying state S and Bitcoin was state prime then um then like you know your ETH on the ETH uh settlement contract and your soul on uh Salana settlement contract you are always going to be the owner of that you can always uh take that back with you uh using force withdrawal methods um and same in Bitcoin as well. But now you've got these like two different versions where you can now as a user choose the right fork of uh right fork by yourself by like by basic like you're voting on the right fork by basically depositing uh your asset on whichever version you think is the right one. Um so that's a very theor interesting theoretical concept that we have but practically I

think this will all lead to um the chain or like the the Selman contract that has the highest amount of TVL to being the dominant one. Um so it's like very it goes into this debate of uh like what John and Tugru always argue about what is an L2 and what is not. Uh so so it's it's very interesting uh design space there. Okay. So it's all just chains and bridges. Exactly. Yeah. Okay. So, so to kind of summarize that we have in the event of some sort of reorg on a base layer which assumably so in in the initial stages choosing the most prominent base layers kind of eliminates this risk as a whole. I mean seeing that I mean Ethereum isn't going to roll back over 1.5 billion buy hack. Salana likely

isn't going to roll back for any sort of real material risk but there's always risk that's that's present. But even assuming that aside, worst case scenarios, we enter a position where there's effectively uh a need for some sort of hard fork uh because of a lack of of um acceptance of um the the amount of tolerance of what the states are aligned or like to the extent that that there's two different states there there's a certain percentage that it's I have a question about that part like you know if if we choose Bitcoin, ETH and Salana as the the settlement chains um two of them agree, one of them disagrees. Does that mean 67% and 33% or does that mean, you know, is there like you're taking into account all of the

nodes on each of the networks? If there's a thousand validators on ETH but only two validators on Salana and 100 on Bitcoin, ETH has the most weight, how do the percentages shake out? Yes. Yes. So, so in that case, uh this is something that that we're still uh kind of like thinking about, but we've got like a couple of options that we can go about. So first option is like the naive one where you assign like equal voting weight to each of the chains. But that might not seem as fair um given that as you mentioned like you can have different level of decentralization across different chains and also and different levels of economic security.

Exactly. Exactly. So that that that was what I was getting into where like you know you can get like different level of TVLs from different chain right. So if like you know Ethereum um gets like you know 67% of the total TVL that we have in the network um then like you know it it doesn't make sense to give them only like 33%. When you say Ethereum gets 67% of the TVL in our network you mean 67% of the TVL on Twine is ERC20 assets. Yeah. Yeah. Exactly. Okay. Um so so in that case it doesn't like it it's not as fair to like you know provide equal voting weight to each of the chains right um so so one way is like using this programmatic appro approach of

uh weighing in the economic security by uh by like you know deciding how much you've deposited into um twine div uh with the total number of uh asset that you've got staked on the network um And uh and the other way is like it's it's more it's simpler tech-wise but could be interesting to explore which is like you assign a DAO to um to to like determine the voting weights periodically um for the network. So, so these are some interesting approaches that we've been so again like the worst case scenario here becomes that we split twine into two chains and because of a lack of consensus

uh that's lower than the tolerance that we could be more than two chains. Yep. N chains whatever that that is accepted basically. And then uh from that point forward like users users assets would still be uh able to be force withdrawn. But then what how do we return back to the single single state single chain for the next blocks is that that just comes down to social consensus at that point. Redstone modular oracles are available on 50 plus chains trusted by over 100 clients and over 6 billion in total value secured. Visit redstone.inance to find out more. Yep.

Yep. So at that point it's like what happened with Ethereum and Ethereum Classic uh during the Dow fork, right? So um yeah that's that's how we think about it. Um and uh okay cool. And and so any suppose n equals three, right? Bitcoin, ethan, salana. Does okay. And then twine we have twine 1, twine 2, twine 3 is 1, two and three representative of the settlement chains. For instance, we have twine one is the ERC20 chain. Twine 2 is the bitcoin asset chain. Tw and twine 3 is the salana asset chain. Is that the reality or is it such that there is a mix of assets on Twine one, a mix of assets on Twin 2, a mix of assets on Twin 3. So,

so like you know if if like you know let's say we've got like three forks um with like you know one's one one with Bitcoin state, one with Salana state and one with Ethereum state. Uh well in that case what we do is um so you've got one version of the chain uh which has which like for which bitcoin is the only valuable asset and other version is where like ethereum is the only valuable asset and salana is the only valuable asset on the uh third one right so um so that's how it operates like you still have the representation of your assets on each of these networks but um it's not valuable and then like you know by uh like going between these different versions is how you I see. So then you

suppose that the the the iteration with the most TVL would end up achieving social consensus and so you would you would say like you know say the Bitcoin chain has 51% right the ETH assets from Twine ETH move over to the Bitcoin fork and then the the soul assets from Twine Soul move over to Twine Bitcoin and then you ultimately arrive and converge these three back into the Bitcoin uh version uh by moving the assets into that one twine iteration and that's the one that that continues to pr proliferate. Yep.

Exactly. Okay. Okay. Just zooming out of the catastrophic reorg situation here when you mentioned about uh the DA component where um there's can be some opinionated choice as to how uh either smart contracts or users or applications basically smart contracts on Twine can uh choose which DA provider they're using. So what you're saying is that on Twine there can be simultaneously applications that are settling to different base layers using the same DA layers and then there's applications that are settling to different base layers base layers that are using

different DA layers all within the same chain. Yep. Exactly. So like technically there's a lot going on there but what is the what becomes like the advantage for builders as to why that's like a breakthrough for them compared to just you know using a stack that gives you like one choice and one base layer like what is the technical breakthrough or for development or what what net new ideas does this concept enable for application developers or um yeah so so what what like instead of like you know going from a single base layer and a single data availability layer layer.

What now happens is like let's talk about the settlement layer, right? So you can now tap into multiple base layers uh like you know the trust minimized um bridges from multiple chains at once within the same assets within within the same state space and uh with DA layer as I mentioned it's it's going to like for it's it's more for users than builders I would say because um users are the one that's all at the end going to be burning the cost for the DA. So, so builders like so for example, if I'm building a gaming application, my DA doesn't probably need to be as secure, right? So, so I can just choose like some cheap DA uh in that case and if it's like some high value, yeah, for a financialized use case, but like you know, you use

Ethereum, but um me as a user am I'm supposed to be able to choose the DA. Um no, no, application chooses that. Okay, understood. Yep. But you can also like you know because like there's two when you're doing a transactions there's like two different versions two different accounts that is being changed right uh you've chosen you've already chosen your uh your DA you want to um get your data posted to as a user and you've like you know the application has already chosen its uh DA space as well. So so like when you do a transactions both of those like gets uh posted back to its respective D.

So does Twine then have a multiaceted execution engine? Like is it like you're you're able to then use like Bitcoin and Bitcoin assets and Salana SPLS and ERC's all within the Twine chain itself? So then what is the execution uh like environment that you guys are operating? Is it an EVM chain? Yeah. So we're starting out with um an EVM chain cuz that's really where a lot of the developer tooling is available. And then I mean like the goal really was about making building easier across chains, right? Um and uh it didn't make sense to introduce new VM early on when the goal is like you know to get to make things as easier for developers as

possible but like you're able to then like Eclipse has Salana has a SVM on Ethereum. You guys are able to have like Salana assets you used in um EVM in EVM environment. Yep. Exactly. And those are like wrapped Salana ass assets and like wrapped Bitcoin assets or because you can settle because it's settled to m to each of these they're they are the native assets. I mean it depends on how you define native and wrapped right uh like you know the the uh ETH we use in arbitum today or base today is also wrapped ETH. So, so define, so I mean if you define native as being like you know the trust minimized bridge um issued on the L1 is what I mean. Yep. Yep. Yeah.

So, so then like you know in that case everything is wrapped that's beyond uh Ethereum or like beyond Salana. So, so I mean like what we can optimize here on is like you know make bringing down the trust assumptions to as low as possible. Um and that is the supposedly the industry standard for calling native assets. So those are the the advantages for a developer of an application. Um let's let's talk a bit about the advantages for the end user, right? Because ultimately you you built this so that you could have a crosschain lending market. Um when you have you know multiple assets, multiple uh settlement engines uh settlement layers um you know how does this benefit the end user from a user experience perspective? So, I mean, like I think we can start start

this off by kind of like debunking the idea of like what a lot of people think that crosschain doesn't really matter as much. Um, it's like it's it's kind of ironic to me because um when you think about Bitcoin and why it existed, it's because uh at the end it's it's all about the value you hold as a user or like the holder. then the abstract then this like abstraction of um like you know the the the local currency right so me being in Nepal and holding a Nepalese rupee um is not same as uh holding USD in America right so so but but then at the end that that shouldn't be the case

like at the end like we're all all doing the same work uh so so that that's really what ended up like you know uh making bitcoin relevant especially across like you know outside the states Um and now like you know when we think about okay like you know users on like holders of Salana uh assets are not they don't care as much about Ethereum assets or Bitcoin assets or like the the uh activities that happens there and that doesn't make as as much sense to me. And so in in like the application of a like crosschain lending market is probably the easy example to to picture. Um, you know, for instance, say you need to liquidate some collateral, right?

Because, you know, someone got margin called. Um, you need to find a venue to to go sell off that collateral, pay it to a liquidator. Um, where do you do that? Like, do you do it on Bitcoin? Do you do it on soul? Do you do it natively on Twine? How are are you able to do it in a synchronous contract call? Because you have all these things tied together. H how would that like scenario play out? Yep. Yep. So lending market is very interesting and I would say like you know I think the kind of application that benefits the most with twine uh and uh the reason is like liquidation is always the hardest part about building um building like a lending and boring protocol and especially in a crosschain environment and uh in this case it becomes a lot easier. It's like you can

just deploy an a fork um like you know at the bare minimum and also like be able to serve um soul ETH bitcoin assets um and then the reason is because all of that is living within the same state space right so so like all of that can happen synchronously um and also like you know even even if like you know you're not building on Twine um like all of your application on Twine one of the things that you can do is like you know even if you're building a crosschain application um like a crosschain lending and boring protocol using um like with the construct similar to how you do with layer zero where you deploy your different instances on multiple chains.

Um what you can still do is like deploy your logic on Twine uh which then handles all the liquidations. So there's like two ways to go about this. The easiest one is like deploying all of it on Twine. But even if you don't just by deploying your logic on Twine, you make things a lot easier for yourself and and the users for the end user too. Y and so just to to put some like concrete, you know, examples behind this, like you know, for instance, say I I borrow Bitcoin against ETH, right? And uh I have ETH as collateral. I have Bitcoin as my debt. All of a sudden, you know, Bitcoin goes up, ETH goes down. We need to sell Bitcoin and use it to repay my debt in ETH, right? So I need to swap Bitcoin for ETH. Where do I do that?

Where do I where do I execute the swap? So, we have we'll have um an AMM or a swap contract within like Twin uh specific chain as well. Um where now you can do all of that within same within a single call. Okay. Within this within a single call, you can make uh uh synchronously composable swap between Bitcoin and ETH. Exactly. So, where are you guys at in development now? So we've achieved the multi-chain installment between um Ethereum and Salana so far. Uh we'll probably have our developer test net ready by early next week. Um so lots of work going on in the back and Salana was especially hard um because Solana is not a friendly chain for developers. It's it's great

chain for applica uh for users but not so much for developers which is why you don't see a lot of like developer activity there. Um but uh like and then then like you know one of the re one of the things that we require for our construct to work is the chain has a light clientish support uh in ZK and uh Salana not only didn't have light client uh not only have like not only didn't have like ZK light client but it didn't have any light client at all um because of like how messed how like integrated tightly coupled heavy its consensus um and its whole like um network is u So we had to build that out first and um and and like you know go move forward in the

stack. So we did that took a lot of time uh but I would say like that was um the 80% of the work um for uh what we want to achieve and uh and the thing is like we really wanted to do Salana for a couple reasons. It's because a like Salana is like the best settlement layer you can think of after Ethereum. Um, and B, it's a different execution environment, right? And it's an SVM. So, like to truly like showcase our power as to like what this construct can do, Salana and Ethereum was like this two perfect uh choices for us. So, we really um wanted to do that and uh I'm glad that we were able to do it. Um and uh

yeah we can dive into this later but uh I also want to call out that settlement like you know settlement is like a very very like you know debated term in the space um but it's a settlement is very very asset specific so as long as there's demand for um for ETH the asset like you know people will build L2s on top as long as there's demand or like this there's this like huge liquidity on sold the asset people would want to build uh L2s on Salana, which is why we're seeing like a lot of these network extensions and L2s coming out on Salana over the past year or so because the the demand for Soldi asset grew so much that and then the distribution it it carried with itself grew so much that um there's

like a lot of people incentivized to do it. So like Selment, what I want to say is Selment is not something that's very unique to Ethereum and can be extended to a lot of other chains and Celestia actually is like a very interesting one in this because uh started out as a DA layer but with their ZK accounts which allows you to verify ZK proofs. Now you can actually build a trust minimize bridge and store your state there and verify your execution. Um so as long as you can verify your execution on this chain you can call that a settlement layer. Um and the demand for that depends on whether there's like this demand for their native asset and TIA is is like you know really uh it's it's a

big asset right it's a major asset. So so which is why you know like for me like the most exciting thing about Celestia is the the lazy bridging or like the ziki accounts they're introducing. Yeah. So is Celestia next on the We talked a lot about Bitcoin, but I take it Bitcoin needs OPCAT in order for you to verify on Bitcoin L1 and settle to the L1. Yeah, I mean uh like OPCAT makes things a lot easier, but even before that bit VM2 construct, it doesn't use bitm. Yeah, but it's an optimistic uh setting. So it's like we need to do a lot of modifications here and there, but uh it's still possible.

So Celestia is a bit easier in that case than Bitcoin. So, h how does how can you leverage ZK accounts and lazy bridging uh to to really just make this multi- settlement uh ecosystem more robust? It probably help if they shipped it. ZK accounts and lazy bridging. Probably help if Celestia shipped it. Yeah, that would be nice. Yeah, that would probably be helpful. Yeah. Yeah, that that would be the the the most important thing. So, so suppose everything goes according to plan and is shipped. How how can you leverage that technology? What we can do is like like Selmet like at its core is just about verifying the execution of whatever happened offchain was correct or not right and ZK is like the be ZK snark is has been the best way to do that so far um and now like you know

until now like you know people have been using Celestia to dump their data right um and nothing else it's like that's why they call it lazy chain dump their tokens too yeah I don't know yeah moving on and uh like you know the reason like that that was the reason why they didn't even have any execution layer at all right but now as soon as you introduced um a feature that can verify zk proofs on this chain now you can verify that the execution for these transactions offchain happened correctly so we already have the like clients for celestia now you get the proof verification for the execution offchain that's all you need for uh a chain to be settlement layer so as soon as that's live uh we can use Celestia as

a settlement layer as well. And I think this is really where we kind of start to see the difference between um like kind of see the line blurring between settlement layer and DA layer. Um% yeah and then I think like the way Celestia approached it is like very I think the most optimal way of building a solment layer remove all the debt that's uh uh that's not helping you become a settlement layer. So, so like you know all you need to be a settlement all you need to be a good settlement layer is like you have this uh high data availability bandwidth and this ability to verify ZK proofs that can verify any arbitrary computation offchain. So that's all you need to be a layer and then uh you can just like you know work on that and optimize towards like you

know making that base layer fast enough to provide actual network effects for everything else being built on top. So, so yeah, that that's that's a very that's where like lazy bridging comes in. That's where like Ethereum shipping some sort of native interrupt comes in. Yeah, I don't know when that native interrupt is happening, but I'm very uh I I wish that happens fast. Um All right. So, do you think Salana is going to flip Ethereum? Um I I don't have a lot of opinion on that actually. Um both have been um a great like they they are like doing great at what they do, right? Salana is like there's like this idea of like single composable machine um and that

that's like working well on their favor um on Ethereum side of things. Um like they they shipped the rollup uh centric road map but they they somehow didn't do the native interop ac across these roll-ups which is I think the main thing that's hindering uh from achieving what um Ethereum could achieve. So for example um like you know across did if I remember correctly and like has done like over $19 billion in volume so far across Ethereum L2s um and that's only one bridge and the thing is Ethereum or like the the settlement layer should be the interoperability layer for uh the L2s on top and Ethereum that means

Ethereum could have captured all that volume and all the uh revenue that gen that that was generated from that if it was not slow, if it if it had some sort of native interoperability. I mean there is like you can do native interoperability but it's like very expensive and time consuming. Um so so once all of that is fixed I think yeah I I don't I can't answer that as to like which one will flip which but uh again Salana is also like on this edge of like you know all the all the applications that uh that were like very high valued uh or like that was generating a lot of volume they're spinning out their own chain like Jupiter announced their Jupet uh which by the way is also in the same direction

as what Twine is building um And uh and yeah, so it's like now like Salon also has to think about how do they capture that um how do they keep them on the chain basically? Exactly. So so it's like both of them have their own problem to figure out. So whichever solves their problem first will be the one to flip. Do you think uh do you think L2s should launch a custom like their own gas token? I mean it doesn't need to, right? uh but with things like you know um things like fast finality like just providing it depends on like what services you provide as an L2 like a vanilla L2 I don't think it requires its own gas token um but things like uh this is like something movement is also doing

as like you have this fast finality pre-confirmation post confirmation whatever layer you want to call it something that provides uh on demand better user experience uh in terms of latency for users. Um so that sort of things it it it requires a token. So in that case you would want to uh use um you would want to use like your native token for that. Um and gas itself uh like I'm not too sure as to like how much demand that would u induce on your chain especially early on. So maybe later on it's it could be a good option.

uh but early on probably maybe not the the best thing to do cuz like you know especially when you're tapping into uh Ethereum um distribution like you know the last thing they you'd want them to do is like they come to your chain they're like oh I I have to buy this go buy this token to transact here. Yeah. Well Rashin it's been an absolute pleasure chatting I hope you have a great week at ETH Denver appreciate your insights been a great technical conversation here today and learned quite a bit. So thank you. Thank you so much for having me guys. Um this was amazing and uh yeah if you guys have any questions please feel free me feel free to reach out to me on Twitter, Telegram um wherever. I'm terminally online.

Always excited to chat more about Twine and what is possible here. Um so yeah again thank you so much for giving me this honor to be here. Yes sir. And and uh where can they find you at twine.xyz and on Twitter right? on Twitter, TwineXYZ. Uh on Telegram, on on Twitter for my personal account, just my first name and last name, uh Ratchin Kalaki. Uh and Telegram, you can find me at ratchin_k. Yes, sir. So, yeah. Thank you. Being me. All right. All right. Thank you so much, guys.