Fragmentation From First Principles | Arjun Bhuptani - Everclear

[Music] yeah I think it's pretty clear that it's the time of liquid funds having uh good tools and data for them is important next up we have the amazing Argin uh bani from Everclear talking about fragmentation from first principles hello um hi everyone my name is arjin I'm one of the founders of a project called Everclear uh this is actually started at my last slide so we're going to go all the way back you're getting a little bit of a preview um I have been a little bit about me I've been working in this space since 2016 to try to figure out how uh we can fix some of the key problems that stop us from making ethereum more widely adopted uh obviously we have not yet succeeded because otherwise we wouldn't all be here um this talk is called fragmentation from first principles I want to try to dive a little bit deeper into what fragmentation actually is because I think one of the biggest problems with fragmentation is that no one seems to have the right understanding of the problem in the first place so um what is fragmentation uh unless you've been kind of completely out of the crypto Twitter sphere you are probably aware that fragmentation is really the biggest problem that's facing ethereum today um this comes from uh just the massive explosion of chains that we're seeing and the the like uh move movement of liquidity to a bunch of different ecosystems um and really what we need to try to understand is what underpins this this trend um at the core of it is is this like original sin uh where when we were first thinking about ethereum first thinking about IR c20s r20s were not really designed to operate in a multi-state machine environment right we had we had this like idea of ethereum at the time where ethereum scalability was going to be centered around the Shing road map where all of interoperability all of scaling was going to be built as one unified thing um this is not what has happened and it's good that this is not what's happened because it's allowed us to expand a lot more quickly but it has led to this pattern where in order to bridge tokens to other chains we have to do this really strange thing where we go and we put tokens into some sort of deposit contract and then we mint a IOU on another chain this sounds reasonable enough but the problem happens when you're dealing with many different deposit contracts many different Bridges many different rollups that are actually going and putting tokens on ethereum into different into different places um ultimately at the end of the day it doesn't really matter if you're talking about like a burn and mint system um or a a like lock in Min system you are still depositing token somewhere you're putting that value somewhere and putting it out of circulation and if it's not in a place where it can be uh used as a uh kind of in a single place where every bridge can interact with it you have this form of fragmentation where uh a token that is minted by one bridge and a token that's minted by another Bridge are different token tokens even if on L1 they are the same token you can really understand this problem when you when you start looking at things like usdc um so usdc has this like very interesting Paradigm at the moment where cctp allows you to burn and mint usdc between all chains it's a it's a fantastic experience this is like how we get to like institutional level scale of uh of liquidity movements between chains um but it doesn't work everywhere and this is the big limitation of usdc today where if you want to expand usdc to certain kinds of new chains or or bridge usdc to C certain new chains for example um from arbitrum to ZK sync you have to go through L1 you have to actually go and uh move liquidity back down to L1 go and lock it in the ZK sync L1 bridge and then Bridge a representation of usdc which is usdc on ZK sync now of course usdc is not the same as canonical usdc and so when cctp eventually goes and deploys to ZK sync you will now have two tokens which is obviously nightmare as we've seen from just about every chain where they've done this tldr deposits are really what matter all fragmentation on l2s really comes from the fact that L1 deposits are in different places if you have L1 deposits in the same place um you can do things like cctp ideally without it being just like controlled by a single Central entity but uh activities like shared sequencing um like more interesting ways to move liquidity between chains ways to be able to bridge arbit amounts at zero cost become available and ultimately if we actually want to solve the fragmentation problem we have to find some way to mitigate this issue okay so tldr we are all building the the same thing um messaging protocols L2 Frameworks Ras providers shared Bridges everyone that's building an L1 deposit contract is technically competitive with one another even if they don't want to be do intent solve this intents are of course a paradigm where uh you kind of have a solver who is this like mythical feature um that goes and uh gives liquidity to a user on the destination chain immediately so that as a user you don't have to think about bridging you don't have to think about like going and interacting application somewhere else um but unfortunately this this actually just just further kind of this just pushes the problem to someone else um uh we kind of talk about this as like a game of hot potato where like users go and give the potato over to Bridges and bridges give that potato over to solvers and solvers give that potato over to Sexes and someone ends up having to take the L on fragmentation right now it's Sexes with large deposit pools but they are also kind of waking up to the fact that that this is this is a problem for them and so really this is a Band-Aid solution right this is a way for us to fix this for users so we can go and figure out what to do on the back end so how do we really fix fragmentation um well there's a lot of answers uh but none of them are great uh option one is basically we kind of kill the entire interoperability road map we we kill any sort of like uh differences between l2s alt together and we have everybody adhere to a single enshrined message transport that is a core part of ethereum uh this kills the super chain this kills elastic chains this kills like any sort of differentiation between rollups it also kills every single inter interop protocol but it works and it's actually would work and solve all of our problems um unfortunately this requires a shelling point right everybody has to agree to use one deposit contract and one messaging framework and of course everyone goes and builds their own deposit contract and their own messaging framework so you end up with 15 of them uh this is unfortunately what's kind of happened with all of the different messaging systems and agair and etc etc um option number two try to find some Middle Ground where you have a a deposit contract um but you are interacting with many Ming protocols this is also something that has been tried and unfortunately hasn't worked um there's an EIP that that I I actually helped write a couple of years ago called xc20 um that aimed to create a shared deposit contract on L1 for a bunch of different messaging approaches unfortunately again the incentives are not super well aligned right because if you have a shared L1 deposit contract you you still kind of lose some of the the like competitive advantage that you have as a project in this space another tldr the incentives around fixing fragmentation are really broken and that's a big part of the reason why it's such a big problem um so what is the theoretical best solution uh the theoretical best solution is we don't we just sort of push everything back into L1 right we make L1 responsible for uh uh sequencing like all of the rollups we make Lan responsible for uh moving funds between all the deposit contracts and this is basically the based rollup proposal um but based Roll-Ups themselves suffer from trade-offs and the big trade-off here is that it's a it's a long time to get to prod right we can't afford to wait around for several years for Bas rups to ship we need something that works today and ideally something that is actually usable with things that are not going to adhere to a very specific specification right so what happens for rollups that are posting elsewhere what happens in the case where you have rollups that just don't want to become based for whatever reason um they just are kind of shut out of the system so what is potentially helpful is to Define more precisely how what we are actually even looking for in a solution and uh and to be honest there isn't there isn't a true answer to this yet but at least understanding the problem for everybody here can help us get a little bit further and what's missing right now is a solution that actually optimizes for two things two very like present practicable things if you were here for Danny Ryan's talk this is the time for ethereum to kind of become practical right we need to figure out how to make this a usable product that can be sold to trafy um and in order to do that we need something that preserves sovereignty for all of the projects that exist today that are actually going and doing a lot of the BD efforts to go and work with um a lot of web2 and and and trafi systems um and so this means preserving the sovereignty of rollup Frameworks of messaging layers of agay uh super chain Etc um and then two something that is actually usable today right something that does not forecast out solving a set of research problems in the future that can potentially get bogged down in unfortunately what has happened with like ZK rollups right um ZK rollups are are such a promising technology but if we had not shipped optimistic rollups we would still not today have any way to scale so it's good that we had an interm solution in the meantime um I don't really have a real answer to this uh this is this is like one of the most interesting and open uh in the space today and if it's something that you're interested in talking about definitely come find me afterwards I have a hypothesis um and this is something that that I have been working on uh with my organization Everclear but I think it's something that's starting to see a lot more traction elsewhere as well people are starting to believe in this notion of netting um uh netting is a process where Financial systems uh basically find coincidences of wants between different parties in an ecosystem such that the total amount of movement of funds you have to do is a lot lower so instead of having to go and Bridge uh all of the funds that are there you can't really see see how much that is but I think it's a I don't know several million dollars um instead of having to go and Bridge all those funds between these rollups you can net and have them just move only a final amount of funds um I'm going to leave things there and if you want to think more about this with me then come talk to me afterwards thanks