Multichain Uniswap - Devin Walsh | Superhack Summit

foreign [Music] all the things that are happening in this L2 ecosystem this is the perfect opportunity for you to think about and learn about how you can potentially also structure your projects to be on multiple chains so tell him welcome and bye Hi how are you thanks for having me today I'm doing great it's been a really fun day I've learned so much and uh yeah yeah very very busy it seems I call it the end of Jessie's talk um and yeah the the bass launch has been super exciting and unchain summer is very cool so um very exciting time to be building in the space absolutely well I'll uh I'll let you tick on with the rest of the presentation and uh I'll get out of your way great thanks so much Protech and thank you um all for being here with me today uh today I'm here to talk about multi-chain unispop unispot protocol and cross chain generally um and um here to fill you in on a few different things first uh I'll introduce myself I'm Devin Walsh and I'm the executive director and one of the co-founders of the unisplot foundation I'll pick off by just explaining a little bit about what that means what is the unit swap Foundation how is it different than companies like uniswap Labs um then I will go into some background info on unisp Cross chains some 101 on how uniswap goes and is deployed on other chains on whether they're l2s or other l1s some some history there as well and then what I thought would be helpful for this audience of hackers is talking through the opportunity for multi-chain uniswap and multi-chain generally for protocols as well as some of the biggest considerations that that I have and and I think the the uniswap Dow has when we deploy protocols on other chains while the opportunity for multi-chain protocols is large we are still early and there are some frictions and things that I think could be made easier to do when we think about deploying protocols cross chain so first off what is the uniswap foundation so uh I along with my co-founder Ken ing founded the unit spot Foundation one year ago approximately in in last summer right now we are a team of seven operating remotely around the world it's possible some of you who are youth Global hackathon frequenters may have met myself or some of my colleagues at some of the hackathons around the world and we are a non-profit that's focused on supporting the growth decentralization and sustainability of the uniswap protocol and ecosystem uh we do that using UH 60 million dollars in funding that we've received from the uniswap Dow that goes towards grants grants to build developer tooling content interfaces cool apps on top of unispop alongside new committees processes groups events that support the longevity and sustainability of the protocol and ecosystem and if you want to learn more you can visit us online or on our Twitter I would recommend following us on Twitter because next week we will be posting our first set of requests for proposals in other words grants funding opportunities to build for uh uni swap V4 which should be coming uh to be deployed by end of this year potentially early next year so uniswap X chains uniswap across chains um unispot protocol as any of you know was originally deployed on ethereum but since then has been deployed on many many other chains both l2s and l1s um some a quick primer on how that happens uh the key inputs to a deployment happening on other chains from the protocols perspective are at a the the open source license on the protocol um called it the BSL or business source license as well as the unispop Dao so uniswap V3 and unit swap V4 upcoming are controlled by a business source license which essentially prevents others from you know deploying uh the protocol for two years unless they've been granted an exemption um uh in this case the Dow was able to Grant exemptions to teams and after two years the protocol transitions to a GPL license so in that two-year period when the BSL is is live the Dow can vote and as you can see on the right hand of my screen you have some recent votes that the Dow has has conducted to to uh you know in this case recognize that uh pre-gpl uh to deploy the protocol on on other chains um and once the uh code transitions to GPL anyone can deploy anywhere and the Dow recognizes canonical deployments um which are supported by the Dow and foundation and I can go into why why that canonical recognition is is important a bit later on some history um so as many of you know uniswap has been around for a few years now uh the most recent live version you just want V3 was deployed May 2021 so a bit more than two years ago uh and swiftly after that was quickly deployed on optimism Arboretum and polygon some of the earliest l2s in the space the first L1 deployment on cello was was conducted a little bit maybe a year and a half ago or so um and over the past two two years uh 11 deployments uh have been recognized as canonical deployments of unisplot V3 on other chains and the Dow recognizes those deployments um by uh through essentially an ens subdomain v3deployments.uniswap.e you can check those out and basically look at the look at the address which has been recognized as canonical by the Dow on all to use all of these chains most recently I know you all just heard from Jessie Pollock on the unit swap was deployed on base and it was also deployed on Linea those are two of the most recent deployments which have been been done um now cross chain usage um so I think it's notable to call out how um popular some of these these X unit swap deployments are so five of the top 15 dexes in terms of volume are um unispot deployment so you can see that unit swap on arbitrom is the third uh most popular in terms of volume at Dex as of as of today also notable um are the fees generated on you to swap you can see that average in terms of average daily fees uniswap is only behind ethereum and notably ahead of Bitcoin and BNB chain in terms of fees generated and you can see that those fees are not only coming from the ethereum deployment but also arbitrum polygon optimism Etc so when you think about unispot deployment cross Chains It's you know not only in in name and like philosophically believing in the power of multi-chain but also in terms of you know true um people are like are showing with with their action how how it needed and valued these cross-chain deployments are uh so one of the things I wanted to talk about uh today as well that I'm sure you're you're hearing quite a bit of are the opportunities in terms of cross-chain deployments so as we could see from the last slide there is you know product Market fit for cross-chain deployments of unispot for sure um our vision or I believe the future of the space is multi-chain um and that on all of these chains you will have tokens that require swapping and that requires liquidity and we envision uniswap to be the liquidity layer for those tokens across all chains um and as again many of you know many chains have different trade-offs so um speed for one transaction cost and I want to call this out because for you to swap you not only have the cost shoe swap but also to be a liquidity provider so to provide assets which can be swapped against and that is a an activity which for many on the ethereum chain is is not as untenable if you don't have too much Capital to to provide liquidity with um I was doing a test last week where I was providing liquidity on ethereum mainnet with probably about a hundred dollars worth of capital to just test out a new function and went to rebalance and it was fifty dollars worth of gas in order to merely rebalance by liquidity and that is just unsustainable on the ethereum main net but that's an activity which is much cheaper and can thus be done much more efficiently and effectively on l2s like optimism and arbitrum and bass and then something else I want to call out is the fact that different chains and different l2s will also have different Focus areas and different um distribution to serve different use cases and different populations so Sello the first L1 that you know swap was deployed to um is focused on it the use case of tokenized carbon credits and is has raw distribution across the developing World um coinbase is based similarly will have um you know strong distribution into um you know a slightly less crypto native population than some folks who are on chain all the time but there are still many frictions that exist today with cross chain deployments um having gone through quite a few of these myself I think there are definitely areas that I can speak to from from experience that um you know or have troubled Us in the past and I think trouble um also other teams um which deploy their protocols across chain and I'll dig a bit into into each of these but three that I'll call out are bridges and bridge bridge security two is the Integrity of a deployment and three our support for the deployment it's it's one thing to to actually you know deploy code for that code to become useful on the chain is is you know a whole other story and um uh and has a number of dependencies for one Bridge security so so why why does uniswap care about bridges why does the unisop protocol care about bridges well you just swap governance so decisions to turn on protocol fees to create um fee tiers and um you know uh to to do other things but the protocol are um occur on ethereum um if if a decision is made via uniswap governance to turn on fees on another chain like optimism or to add a fee tier like one bit the one bit feature on optimism or polygon that decision is made on ethereum but that decision that message needs to be passed in a secure way to the factory contract on optimism or or polygon um and we as the Dow need to trust that that message will not be be compromised and that the bridge which owns the factory contract on the other chain will not be compromised as many of you have likely seen bridges are a developing area of the crypto ecosystem and there have been a number of hacks and attacks over the past several years when it comes to uniswap a bridge attack could compromise ownership of the Cross chain deplena and you know a malicious hacker could potentially take over that deployment and send decide to turn on fees and send them all to his or her self um which is not not a scenario that the unit swap Dow would want to see happen which is why we care about Bridge security um unfortunately Bridge security is notoriously difficult to assess and it is nearly impossible for the typical delegate even the technically Savvy delegate to do well given the amount of context and time required to do it um we've seen in the past some of you may have seen earlier this year there was a uh you know situation that that kind of um uh came up in in the U swap down which uh you know there were potential conflicts of interest between delegates who had also invested in in Bridge teams which caused you know um some some debate within the community about how these decisions should be made and so what we did at the foundation was funded the creation of a committee of independent experts um on Bridges to define the unit swap protocol use case build a comprehensive framework for assessing the security of bridges for that unit swap use case and then they diligence I think eight or nine bridges in in total and recommended two bridges for use as well as a road map for the other Bridges to to be considered uh usable and and secure for the spot use case um the team also uh supported and is continuing to support and we're continuing to fund development of a multi-bridge aggregation solution so you can think of this as a kind of multi-sig of bridges that adds it so two of three three of five Bridges um from ethereum to some other chain um that need to approve a message before it's sent just adding that even more security to the mix um another consideration number two the Integrity of of the deployment so earlier um in in the the slide I mentioned um the uh the licensing of uniswap V3 so today uniswap P3 is GPL totally open source anyone can can Fork it anyone can deploy unislot V3 anywhere today which is great um it's it's great that anyone can build and play around with this code and use it as they would like um however um so there there are some kind of considerations and things that that we had foundation and the Dow thought were important to address here um given how broadly and widespread usage of the unit spot protocol is um we thought it important for um LPS and users generally to know which which deployments have been reviewed and approved by and are owned by the unit swap down itself and there are a few examples of why this is this is important and I think the the tweet that I that actually sent out yesterday um is a good illustration of this so um the governance lead on my team Aaron basically found that there was a uh there are two deployments of unisplot P3 on on base one um that was voted on by by being swept out to be canonical and recognized in another that was not which again is totally great like base is permissionless anyone can deploy the uniswap V3 code the reason we thought it was important to call that out is that this deployer had turned on fees and was sending fees to themself which again totally fine to do but if you're a liquidity provider and you're providing liquidity to that deployment you might not realize that that you're actually giving up a portion of your fees to this deployer um because you you may think that you're you're providing liquidity to the canonical uniswap V3 protocol which does not have turned on fees and um NG who just may not realize the fact that there are are multiple deployments some of which are have different parameters that that impact your return in this case um this this was not the case here but it's also possible for um a a deployment to um have code uh rearranged deleted or maliciously updated to rug users um uh and of course that's something we we really care about um or would we care about preventing um we would not want folks to uh you know go to swap or provide liquidity on a deployment of unispop on base or any other um any other L2 or other chain and be harmed by by that usage so that all gets to why we as a as a dow the unit swap Dao has decided to take this step of recognizing the canonical deployments the deployments where users can can trust you know the the code has been deployed by trusted sources it's been reviewed by folks in the Dow it has the same parameters as as um that you would expect a deployment of unispop to to have um and I call this out because as we saw yesterday um you know me tweeting about this is probably not the right long-term solution to to this problem or um to this I guess like consideration that we have um so so that's another point I'll call out and the third is um chain middleware and interface support so so I call this out here for um if you've you've you know approved and are excited to uh deploy cross chain that's only the beginning of your journey in some cases um even if another chain is evm compatible um some evm chains may not have the same EIP support as as mainnet um which for you to swap will be relevant for V4 I'm unified V4 is dependent upon an EIP EIP 1153 which will be introduced in the Cancun hard fork and um if an evm compatible chain does not have that EIP V4 will work differently on on those chains and will not have the gas efficiencies that that 1153 chains have middleware may not be available across across all chains where you deploy unit swap so um uh or the middleware May function differently and in a one example I'll call out here is um for folks who have used the graph on the graph is an amazing tool that uh is you that we use with uniswap in order to index and make queryable unit swap data we found that on arbitrum the arbitrum Chain was too fast for for the sub graph um so we had to well one for a year or two to be honest some data was actually broken on the info.unispop.org website until the foundation provided a grant to um to a developer to do some custom optimizations to the arbitrum subgraph to uh speed it up so it could keep up with with the chain and provide data to the users to our users um and third uh interface support requires additional Dev work so um so in the union swap realm you may be very familiar with the unispop uh app front end um which is maintained by unislap labs and um you know you that in in some cases in this case not not all um chains which unit swap protocol is deployed on are um actually readily available on on this interface yet um uh popular interfaces may not have the development resources required to maintain um all of the the chain support given some of the considerations I've mentioned before um uh so so all of that being said this is leads to you know um some some frictions per additional deployment of a protocol um so so with that I think I'm actually I just uh I'm just at about 20 25 minutes time um wanted to say thank you again for having me and um I am happy to chat this weekend or you know over the next week however long folks are hacking um reach out to me on Twitter I'm at Devin a Walsh um and you should also follow us on Twitter to see some of the grants that will be excited to support uh next week that was I was actually very informative this is uh I learned a bunch of things around how uh you structure the the grants and the foundation here but uh I mean you've brought up a really really interesting point which is we will see some malicious behavior here especially when the code is public and being able to actually streamline that and at least trim on the communication if not like the resolution prevents these uh scam potential situations is a very key requirements that not only just for unit swap but every single protocol is thinking about going on different chains um so having some better Frameworks is uh is something that I feel like you and and the rest of your team and a few others will Pioneer very soon and others can uh you know steal from uh for the better but uh like I'm really happy to see this is already happening and uh and we don't have to worry about or well we have to wait for something to go wrong before we start thinking about these things actively which is a really good pattern to follow here yeah yeah I think threading the needle between you know embodying and embracing open source ideals while also you know doing what it takes to ensure user is like definitely a tough line to cross um it's something that you know you don't really think there's a great playbook for today but we're trying to figure it out and and I'm also excited that's why I'm so happy that you guys have these hackathons because there are very smart people here who can help us figure these things out yes I think the the people uh people usually think about hackathons as a way to accommodate ideas in codex but I exclusively think of them as a way for us to figure out what's broken and uh that also with more net new progress than uh than the first one yeah there's and there's there's quite a list I think our yeah I heard the end of his talk where he was saying they they in the past three days that they've like aggregated and I can I can I second that like many things a lot I think for some of us who have been in the space for a long time I think there are a lot of things that like have been broken for so long we don't even realize they're broken exactly yeah Power defaults for for us and our attackers versus people like oh this is annoying and I wonder why and like oh it's how it is but you actually get to see that when you bring in way more people who are uh reacting for the first time so it's been a great pattern and hopefully we'll uh we'll see some better guidance on how to think about multi-chan deployment stuff from that yeah yeah that's great well yeah good luck talking to everyone here and yeah again feel free to reach out to me on Twitter uh and hopefully you know if you think you have a solution to some of those issues I brought up please reach out we have grant funding we support you amazing thank you so much Devin thanks take care bye um