Optimistic Rollups, 2 Years In Benjamin Jones

yeah okay guys i have a very you know legit i swear talk on scalability but i thought we could all use a break from it so i would like to start you all off with a synchronized clap are y'all ready to clap with me there we go [Music] oh all right we did it thank you [Music] all right what's going on y'all i am ben and i am from a company called optimism thank you and i am here today to talk about roll-ups it's the thing we're doing optimistic roll-ups and i hope you guys know portal or this slide won't make much sense but it's been a while y'all it really has it has been two and a half years since we kind of had rollups coming onto the scene most of that's been during covid we personally have been live on mainnet for over a year now and so it's time to talk about that there's a lot of things that we got to say and it's been a while we've been in hiding we've been in quarantine and we've got to get it all out there so i'm going to talk about sort of three things i'm going to go over what's happened where we're at so far i'm going to talk about what we see coming up in roll-ups in the next uh you know say year in the reasonable short term with a lot of clarity and then we're going to try to make some projections about where all this stuff is headed sound good yeah [Music] okay so what has happened okay i'll start with the shells i'll start with the shows guys we almost we are almost at one billion dollars saved on optimism mainnet very exciting so if everyone did all their transactions that they'd done on optimism on l1 they would have spent almost one billion dollars more so we're pretty hype about that if you guys haven't tried out our system go check it out okay so that's the show part i also want to talk about some of the mistakes we made and i think one of the biggest mistakes we made i think the biggest was we undervalued the evm semantics what does that mean okay well obviously to build these rollup things we need to have a way to secure the system and we do that with this optimistic roll-up design it involves disputes some fancy stuff behind the scenes the point is we came out with the first version of this that was called the ovm the optimistic virtual machine and this was great but you'll notice that o is different than e and at first we thought that was fine but it turned out as we kept going this wasn't quite the case sure you could take a solidity contract or a viper contract and get it running on our system but the reality was that the more we dug in and the more people started trying the more we realized that the small itty-bitty nuances of the evm really mattered in fact all of that tooling doesn't integrate it with solidity it integrates with the evm itself the instructions the byte code all of this wonderful technical jargon right and this meant a lot of tooling broke it also meant that sometimes you couldn't just deploy your contract you had to figure out what the difference in the ovm might be and fix that difference and get it to deploy so these are challenges one thing i think we did do right is we got guest running we've been running guest in prod for over a year now and the reality is for these rollups to succeed we need to build on the very strong foundation that the go ethereum code base gives us so that's something that we were doing right but there were a lot we were doing wrong there so if you haven't tried deploying on optimism the last six months you should try now because it got way easier we did the evm equivalence upgrade right so this is a wonderful uh you know um ethereum yellow paper that we perfectly conform with every little nitty-gritty detail of how the instructions and how your ethereum works works on roll-up so that got us the tooling that got us effortless deployments and that got us even closer to geth the reality is we were running death but we had to have some differences now those differences are tiny tiny and that is really good news because it means we're building on the strong core foundation that ethereum gives us okay so what are some of the other things that's happened outside of optimism one thing is that bridges have become a reality rigid if you've used a bridge okay not bad not bad so you know like three years ago this is my co-founder kelvin four years ago now my god writing about fast withdrawals talking about the idea that we can move money between these different chains seamlessly and quickly and that has gone from concept to reality this is from our webpage there's a bunch of bridges that you can use to move around and that's a big deal it's really cool however it is notable that the two largest hacks not just in crypto but in the history of the universe as you know humanity i suppose i don't know about the aliens the two biggest hacks were bridge hacks and we can't avoid this with optimistic rollups that's the point is to give us increased security to avoid it but not all bridges do so i think that's an interesting thing and it's worth discussing and ultimately i do think that's okay because the reality is you get a choice if you want a cheaper bridge that is under collateralized and has some risk to it you can use that if you want to pay a bit more and guarantee your money is safe you can too and i think that having that choice even though it may be scary is a fundamental thing that we do in crypto we want to give users choice and if you want to make that risky bet you can okay another thing that's happened is we've had an explosion of mev so two years ago we put out this idea that you should sequence roll-ups you should have the people producing the blocks of the roll-ups be independent from the people that are mining l1 blocks now we have a majority of l2s following this scheme and we have also seen an explosion of mev on l1 an absolute explosion right when we wrote this very small now flashbots everything else it's millions of dollars it's a huge huge game right and what we're starting to see this is an awesome paper i'd highly recommend by the flashbots team we're starting to see the dawn of cross chain mev we're starting to realize that just like you can do arbitrage on one chain you can buy one asset on one chain and sell it on another and make arbitrage and there's really interesting properties that are coming out of that so it's fascinating there is a lot more going on with mev and it is a long road ahead but it is very very very interesting and it's definitely been very active in the past two years okay the last thing that we've seen is we have seen this uh cycles eos we have seen solana the big block chains and i think this is a really interesting one for the audience and for the people who are here at eat denver because the reality is okay i this is a ridiculous slide it's definitely self-deprecating because i am firmly in ethereum but i think the reality is there are big block chains outside that are trying to scale ethereum differently and while they may not align with the values that are very very important to us and to our community there are legitimate advantages that these systems afford namely they have much better composability because everything is on one chain it means you can do your cross chain call your uh you know cross-contract calls synchronously and the reality is that is what we should be comparing what we're building with ethereum l2 it's simply not realistic to say that values are enough we have to provide the technology and we can do this it's harder because we want to secure things with l1 but it is possible we can compete even if we have tighter constraints and you know in general this is i think there's a multi-chain versus mono chain narrative right now which is kind of what all this is about and i'll talk a bit a bit more about that later namely it is a false dichotomy imo you can have both and the last thing that we have is that the roll-up centric ethereum is going strong so two years ago we put out this post about the role of centric ethereum roadmap today we have literally dozens of l2s shout out l2beat.com you can go look at them all learn about them fantastic stuff the ecosystem is exploding and that's really exciting it's it's the ethereum vision coming alive okay with that being said some things have been lost i would say that if i were to look three years back when ethereum scaling was much more focused on eth2 phase two sharded execution the talk of the town was how are we going to solve the composability problem and since we started saying that roll-ups are going to be an ecosystem that are all settled on l1 that's a great vision but i do think we've lost some of that so i pulled some of these screenshots from a while back that are people talking about cause shard composability and some of that has been lost in the current system okay so that is what's been going on in the past two years what is coming up in the short term okay the big thing for us is optimism bedrook edition just a little bit more shilling here this is our ground up evm equivalence roll up stack so we've basically taken all the lessons that we've learned from building our first version and from running it on mainnet for over a year now we're taking all the incredible abstractions it turns out by the way that eth2 and the merge what the beacon chain does for geth on in the merge is exactly what l1 does for l2 gets in rollups who would have thought and so one of the incredible things is we can use the same apis the same clients really really crazy and so bedrock is our v 1.0.0 that rolls up all these lessons into one thing another part of it is canon oh my god canon canon canon canon is the first public evm equivalent fault proof this means that our disputes and our rollups can behave exactly exactly like the evm itself in fact canon runs l1 blocks it doesn't even run out two blocks and this is fascinating it's a generalization of a lot a lot of research that's gone into the role of space over the past many years and another thing that it does is it gets us ideal gas costs so i could talk about this stuff all day but we've got eight minutes so i'll keep going i think the main thing that i'll say about it though is this came from our frustration with how we were doing rollups before and what were we doing we were starting by writing a fault proof by writing the security system that kept our rollup secure and then implementing a node that would follow the rules that this dispute game that we had invented followed we've realized that this is not the way to go and this is what canon gives us before we were trapped in a cage we had to write a fault proof and then take that exact logic we'd implemented and rewrite it as a node now we just write the node software and the compiler turns that into fault proofs really really really powerful can't tell you how much of a relief it is to not write everything twice and be stressed that they don't match it's really good so we can compile the fault proof of an evm chain okay the last thing that's coming up in the actual proven year term is that we are getting fees much lower ethereum uh as a system is finally coming around to the fact that call data is the primary cost of the secure l2s today and that we need to lower those and we're starting to get pragmatic about these designs we had eip4488 which might not have gone through but at least was a discussion we have dank sharding shot one of the greatest names of ethereum scaling i've ever heard in my life dang it's literally dank sharding okay no it's donk sharing it's because dunkrad came up with it but so be it and we also have these blob transactions that are going to really improve things in the short term by the way if that sounds interesting to you our team is hacking together blob transaction stuff here at east denver so come find me if you want to contribute to the future of what ethereum needs to be doing all right we've got six minutes left so for the last six minutes i'll talk a little bit about what we see on the horizon so i don't know how many of you guys have seen this but dalek put out a post uh in the last month of last year called endgame which talks about his belief as to where the future of ethereum scaling is going where does the rollup centric roadmap end up and i think there's a lot in there and i couldn't possibly unpack it all now but the main thing that i'll say is this diagram if you take nothing away else from what vitalik's talking about the future of ethereum take away this diagram this is how vitalik lays out three different scenarios that all lead to the same end state completely different approaches that appear to be narrowing in on the same thing and the reality is that we're finally starting to see on the horizon what the end game for scaling looks like what is that end game that end game is one in which block production is centralized validation is decentralized censorship resistance is preserved and we have huge blocks which is really important because the blocks that we have today simply will not get us to the global scale and for us to reach the next million people and the next billion people beyond that what we have now just doesn't cut it okay i showed the compile fault proof before thing with canon what we're realizing is that if we can write a node and get a fault proof for it we can write a sharded node and get a fault proof for a sharded chain this is mind-blowing this is what we're talking about and it's one of the paths laid out here it's the top the top right path as a matter of fact okay basically how does that work quick tl dr you have a huge block producer that can create really big blocks and the block producer cuts that block up into 100 parts and at each intermediate part along the way it gives a hash of what the state is at that part then you take one percent of that block you give it to a hundred different computers and those computers can run only one percent of the power but check that if they line up all of their results that the net result is what the sequencer gave right okay so that doesn't make sense so that diagram didn't make sense just think of it like this we're gonna have big plot producers that give us the solana like properties that we want and we're gonna have a bunch of people that can still run that on their laptops and contribute to security this is a way to get decentralization but still scale massively and composably in the way that we want so remember before i said that the multi-chain mono chain slide was a bit of a false dichotomy that's why it's because we can have both we can have big black producers but we can have decentralized validation and i think that is really powerful and there are other ways that we can get there but it seems that they're all ending up in this model this is a ridiculous slide we don't really know how to visualize this effectively yet but it's i like to i like it this way okay so tldr guys i that that's just about it i would say that the tldrs roll ups have been exploding it is not over yet we've got evm equivalent fault proofs that are really really important we're super glad to have them we're going to have a pragmatic non-three year away sharding roadmap that is going to give us the scalability that we need in terms of the cost of data and in the long term we're going to shard that data even more we're going to start the execution and it seems like we are narrowing in on a vision for what blockchain scalability looks like into the future so that's all i got other than we are hiring if you want to come jam with us and you know play more music and other songs together come work with us jobs.optimism.

io and as predicted we have two minutes so i could answer a question in the next two minutes if you all like not answering that one yeah going up there's a mic right there all right we got a man with a question uh how with the big block producers you want to keep censorship resistance and also for instance legal safety of of these block producers but the sensors resistance is most important so the question was how do we maintain censorship resistance and how do we keep safety so for the safety right that's that's this thing right that's taking a bunch of laptops and splitting up the compute to keep every piece safe for censorship resistance it's by using l1 right so just like we can take a rollup and send a censorship resistant l1 transaction into the roll-up we can do that with these big roll-ups too all right it's 12 29 we got one more question to answer what's your guys um stance on the future ck roll-ups and how that will uh compete with you guys and maybe what you guys want to do in the future what we're seeing with zk rollups right now is very similar to what we were going through a year ago with that ovm what we're realizing is that that evm is fantastic and it remains to be seen whether zk will ever get as cheap as an optimistic system because fundamentally you have to do this crazy moon map behind the scenes hopefully for many applications it is useful and we definitely see that integrating into the larger picture of chains but it's not going to be the only thing because rollup because optimistic systems fundamentally are always going to be cheaper all right guys i think that's all the time i got thank you so much for jamming with me take care [Applause]