Ethereum Sequencing and Preconfirmations Call #3

Transcript

okay uh welcome everyone to the eum sequencing and uh pre-confirmation call number three uh I guess this is our fourth call now um the agenda for today is going to be in three parts um so we're going to have um a quick update from from my side in terms of the the the based ecosystem so Builders uh that are trying to leverage Bas sequencing and Bas pre-confirmation I have some exciting updates I'll try to keep it very brief uh and then we have Alex glowski from ZK sync who will be uh presenting his concern with B sequencing that we don't have the the time to

present uh last time actually he was on the on the flight which is why he couldn't present and um I think I want to give him a bit more than 10 minutes because uh you know he's representing a rollup and I I think uh that could be some some interesting discussion from the the points that he bring bring up um and then uh in the last 25 minutes or so uh I want to give an opportunity for some of the various projects that have been building cool stuff and doing cool research uh to do very small presentations or or small teasers on on what they've been uh working on so without further Ado um I'm going to uh share uh the presentation okay so uh I kind of want

to give you a a quick overview of what I think the the Bas ecosystem has grown to in 2024 um so at the very beginning of 2024 there was from what I can tell just one project uh in the Bas ecosystem it was it was Tao uh and uh in in the last few weeks uh we've had 10 more uh projects uh look at it so um one which I learned about very recently I think last week is this um this project called uh called rise uh and they gave me a a small blur that couldn't make it to the core so basically they're they're building a um a base rollup uh which is a a hyperformance rust based parallel evm

rollup built on top of of of ref so this would be like uh after Tao the the second um uh publicly announced uh base rollup um another uh uh really exciting development is that the the the founder of one of the the very top uh rollups in production today uh told me a couple days ago that uh you know they will be uh you know pushing to to make their their roll up based um so this is a you know a very uh you know surprising development to me that it happened so quickly but uh there we

go um we also have in the Bas ecosystem um at least two entities doing research uh so these are consultancies that are uh doing research internally but also uh for the various uh projects so I believe uh lime chain is uh is is doing a great a lot of great research and actually um there will be a presentation towards the end of of this call by George uh on what he calls base ticketing which is a a new a new design and we also have nevermind that has been pushing some fantastic research on on on Bas sequencing and also working with uh with various Partners in the ecosystem um we also have infrastructure

providers uh that seem to have uh you know pivoted in some sense to uh to to Bas sequencing and we're going to have various uh presentations and teasers uh from some of these teams um so for example espresso is now building a a a a based shed sequencer uh we have sella um which I think is more discreet about what exactly is building but it seems that the the company has uh you know prived in some sense to to base sequencing uh same thing with chain bound um very very excited with Bas sequencing and I think also um you know still on the wraps in terms of what exactly they're building and then there's uh a Spire which is which hasn't

yet pivoted to to B sequencing but they're basically considering building um this base roll up app chain framework uh something analogous to the op stack for the super chain and that native framework would would basically integrate base prec confirmations and and and base sequencing to make it very easy for base roll LA to be to be created so um oh yeah and then there's a a final uh group of of infrastructure players which are the relays and we have um gataka the team behind the Titan relay sorry the Titan Builder which is one of the largest Builders they recently released the Titan relay and they're very much excited uh to provide

uh additional services in addition to relaying and the pre confirmations is is one of the base pre- confirmations is one of the services that they're very excited to to Prov and um similarly with the with the ultrasound relay uh currently like the the most uh popular uh relay and and looking to expand Beyond just relaying and and provide pre confirmations so uh that's that's it from me um and then I guess I'll hand it over to uh Alex to uh to present to present his concern uh and and and uring discussion around base

sequencing sure thank you Justin we have uh let me check that we have also comment from our team here I'm here here okay uh so I I'll just share a couple of usually and then common you can present the uh the presentation it's fine that's great uh so I just want to say I'm very impressed how far uh the the progress is going and uh it's it's very cool uh it's very interesting development um uh I as a builder of zik sync I uh uh you of like our mission is shared with everyone we want to scale

ethereum while preserving the properties that make ethereum really valuable like what what made ethereum valuable in the first place and um the history shows that if you if the community does not call askar on certain values and pushes for them they can deteriorate with time and so like like one big thing which we discussed with Justin throughout uh the like last two years but in discussions of B sequencing is that it can it might lead to a bit more of centralization uh because you create an infrastructure that makes makes

centralization easier if you make infrastructure it makes it harder then you know you you can for example like you can make design choices that favor solar uh Stiers or you make you can make design choices that uh don't favor them and like make it very very easy to delegate and um uh restake and then you will end up in situation where like a lot of um stake is is governed centrally right so uh if you make base sequence qu ing like if like let's imagine Bas sequencing wins it would be very important to make sure that the uh Builders cannot be evil because if Builders can be evil and all of the

rollups are re like sharing the same building infrastructure and effectively acting this one huge sharded rollup we might end up in this yeah sorry about that uh the present's been muted now I see uh so like the the the the way we see it with uh like if you have shared sequencing if everyone is using shared sequencing infrastructure that effectively means you have like one huge system that is sharded and that is uh operated essentially by like a few big entities and there is a really big risk that then ethereum turns into like

big Tech chain right where those guys have a lot of subtle power if if they process all of the transactions from the users they could be censoring users they could be doing a lot of bad things right so which uh we know that that there are Tendencies towards this we know that most of the BS in ethereum are being uh processed by entities that are for example a f compliant right and then today it's just tornado cash but tomorrow it can be anything and like the list can grow and then it can be real time and like specific users can be targeted for political reasons or or maybe just because you know like it already happened that the uh like AA compliance leads to uh using of certain companies that provide this and then

those companies can have like false positive and false negatives and they can exclude your users without them actually doing anything illegal and then then th those companies get a lot of power all of a sudden because they they start to dictate and they can like Sly tweak the parameters in certain way which will favor certain players over the other players right so uh my first like concern or or proposal would be like if if if this is the the route to which ethereum is going then it would be really really important to focus on first making sequencing um trustless like work on the mechanisms that would make it impossible

for sequencers to be evil meaning some form of embedded uh sship resistance um like either through cryptography or through some mechanisms but that are working at the sequencer layer I think it's not enough to say like uh CIP resistance is going to be solved by like extra mechanisms where oh if you're being sensored you can send this transaction to the validators because if Shar sequencing becomes or sh sequencing with based rollups with prec confirmations becomes ubiquitous then most most wallets will switch to this most payment providers will switch to this and then like users who cannot participate in this like one second

confirmation thing uh will be essentially excluded from the economy so like the this is generally like it's a very general concern like we we can uh like I mean just want to kind of like leave it open and and uh for discussion and and see what you guys think but I feel that it's very very important to first focus on those like with with rollups we have this experience where uh you know rollups space started as something very uh focused on preserving the security and decentralization properties of ethereum and then the first rollups started essentially like some some rollups started in a way that

did not really work right like rollups right did not have frog proofs for example uh and then everyone said like sure like well it's just going to be implemented in in in a very short time it's it's okay like let's just you know build fast iterate build MVP and and and try things uh but it got sticky right and like some of the rollups still don't have fraud proofs I think most of the big uh optimistic rollups still don't have permissionless fraud proofs which makes them essentially um you know like very similar to side chains with very high threshold of uh of operation uh and this puts a lot of pressure on everyone this puts a lot of pressure on

us like when we have to prioritize our road map we have to choose between transitioning to stage one stage two roll up and like making sure that it's properly decentralized or like building something that will have you know like the space is very competitive and like building features that will create some like you know interesting advantages for for the product for for the platform for network and I I think it's not in our power to change this like it's in the power of the community to create pressure on everyone to comply with certain values of ethereum to comply with this decentralization approaches and like I I I'm I'm just like being having gone

through this with rollups I really really want to make sure like from my side like I will be doing everything to to uh uh to coordinate Community like push Community to actually enforce those fa before they actually get implemented so that's one concern and the other concern is when we build this system like uh I I would love to understand what's like what's the end game is the end game that everyone switches to to base rollups or is the end game that ethereum is a plurality Network where uh the the original idea of rollup Centric world where rollups can permissionless innovate and go into different directions is still valued by the community and and and embraced uh and if

that's the case then the second big thing that would be very important to to preserve is connectivity between the different systems and um the ability like the right to exit the way the way you can think of blockchains like why blockchains are great as Network States as like the with with all the onchain governance is that you can always exit when layer one your exit strategy is a fork like you don't like what the validators of Bitcoin are doing you don't like they they start sensoring users or on ethereum they they they do something bad you can always say you know what like I don't like this this rules um I'm excluded here I can Fork a system change the the set of validators

and continue there on rollups it's even easier you just with throw to layer one and you with throw to other rollups and uh like the um ideally we build a system that that is much more interoperable between rollups but in a way it does not force everyone to go and become you know a um uh essentially one one single system like Embrace completely the same architecture the same approach to uh to sequencing like it would be great to have alternative approaches like alternative systems alternative uh sequencing methods uh that are still interoperable with with this uh you know

like whatever mainstream movement is and uh here we have uh uh an idea to propose but may maybe it's worth to first open uh the floor for questions discuss generally like the general concerns I think there are lots of people co uh Noah you have your hand up yeah so one of the things that I wanted to bring up with like the censorship resistance around shared sequencing and PBS in particular Alex's concerns is I do think adding the censor assistance at the rule of LEL something that's really interesting one of the things I I know Alex didn't bring it up but one of the things that I've seen recently

guys additional features like that l no what it might be nice if you slow down a little bit because your audio is cutting out in and out a little bit the audio is not good can you use a different microphone maybe is this working better so much better okay great sorry about that so to go back to the start the concerns about our PBS I believe are completely valid and I think that one of the things I've seen recently is CK Sor

collaborated for a rollup called CK lighter which has senstive resistance built into their own amm with CK circuits I think additional features like that could be a great way of essentially having better censal resistance but it also comes around the rollups block building rules at the end of the day so I think a lot of these systems that we're building whether it's espresso shared sequence or node shared sequence or any of these other options on the market mes really want to be maximally um lazy in a sense so we don't want to have too much too many things in shin and too many things added onto the system itself so I think a lot of these Primitives for better sens persistance can be added on and you can still Fork away at least for our system you can use altta and you can Fork away from our

system so I believe that those will help out with the cens resistance and being able to say like hey I don't like this I want to leave but I do think they're valid concerns I do think that over time there will be better systems that get developed and the best way for a shared sequencer or base sequencer in this case to really function is to be maximally lazy in a lot of these things and not tell the rups hey this is the way you have to do things forever and you can never Fork away because just like if any of these things with these large decentralized systems like just how ethereum operates today getting everybody to agree on one thing will be next 10 imposs so I think that having maximal have optionality makes a ton of sense it will really help every rollup be happy with the solution that that

they choose instead of all going with one or the other I guess what you're saying Noah is that it would be nice if the shared sequencers or sequencers be minimally opinionated so it leaves as much room as possible for each individual rollup to come up with their own encrypted men pool or their own inclusion list strategy or whatever it is exactly uh Ben you have your hand up yes thank you um and Alex thanks thanks very much for uh for coming here and presenting that I I think that the concerns that you're sharing are something that we all share too and uh and I think it's very important that we're having this discussion um I think that one thing uh to consider is if every rollup is anyways going to

decentralize its sequencing then and Alex I think you and I maybe discussed this before at some point but it creates pressure anyways for Builders to try to build for multiple uh ecosystems at the same time um and in fact if you uh if you're looking at Builders who are going to try to participate and win uh slots for multiple rollups that are have decentralized sequencing at the same time it's going to narrow the Builder set even more um because the the builders that are most well capitalized and most able to um to simultaneously win these different sequencing slots at the same time are uh are going to be just an even smaller Set uh and because

there's going to be Market pressure on uh on Builders to do this because they can offer these um additional benefits to users when they're producing rollup blocks from multiple rollups at the same time um it's going to happen and so in some ways what I view as shared sequencing is some kind of coordination to allow more participation in this Marketplace which brings me to the second point I think that we have I think we we have sort of made a mistake as a community and calling this Shar a shared sequencer uh and in fact espresso is moving away from the term shared sequencer what espresso is really doing

is building a Marketplace for shared sequencing and this is compatible as well with the idea of based rollups maybe not as the the original uh way it was described but if you think about any of this infrastructure as creating a marketplace where Sovereign rollups have autonomy over their own block space and can sell block proposal rights to third parties including the L1 proposer uh that is really what shared sequencing is so rollups that that are participating in such a Marketplace can produce blocks on their own sometime and if it suits them for synchronous composability or other reasons they can sell blocks uh sell block proposal slots to other common proposers some rollups

may choose to sell this all the time right and be like the most based if they're selling it by the way the difference between a based rollup and a non- based rollup is really mostly about whether you're reorg with the L1 or not so if rollups are building off a an older L1 uh older L1 State uh for the benefits of stronger finality then the L1 proposer is not able to construct even to buy the rights to propose for them at the same time because the L1 is is is ahead of the rollup in some sense whereas rollups that reorg with the L1 and sacrifice this faster finality they can sell block proposal uh slots to the L1 proposer so anyways I know that's not necessarily how everyone is thinking

about this but that's how I'm thinking about this and I think I'm coming from the very same place as you are Alex in in the sense that I think enshrining one infrastructure to control all rollups is definitely concerning and and and and you bring up a very good point but I I I don't think that's the direction the industry needs to go in it can be it can be in in this more ad hoc shared sequencing model that I'm describing here so if I was to summarize one of your key points Ben you're saying that if there is demand for synchronous composability there is going to be demand for shed sequencing and that's going to happen no matter what and the question is is this going to happen through like extreme sophistication with people building like additional layers

and like extreme centralization or can we build this coordination infrastructure that actually dramatically lowers the bar to entry to facilitate this shed the shed sequencing and in some sense we're removing complexity and reducing the barrier to entry yeah that's that's Point number one and and I think Point number two is that shared sequence it's not shared sequencers as one infrastructure even base rollups don't need to be enshrining the L1 as the sequencer for all rollups it's it's more that Sovereign rollups can sell proposal slots dynamically um and if they're reorg with the L1 they can even sell it to the L1 peros gotcha so you're saying you know basic

on say is not all or nothing you don't have to like always uh be based like you can choose when you want to be based and you can open it up to the market to to decide where to allocate uh these resources as optimally as possible uh George you have your hand up yeah thank you Justin hey Alex nice to meet you um I'll actually go on the other direction and and approach this from engineering standpoint and I'm not totally Sol on uh on your point where you say well once we are figuring out how to do best sequencing whatever is sequencing let's figure out also sensorship resistance so for the past I

don't know weeks or months uh we've been trying to figure out how to do just sequencing that is possibly enabling Universal synchronous compuls ability and it's possibly as decentralized is possibly thecentral or at least has minimal training wheels and I am very skeptical that um if we also tag along the requirement for base sequencing to be completely sensorship resistance uh we are not going to be uh super over complicating the whole thing so from engineering standpoint I'm not totally so that we can actually do this uh without hampering the rest of the requirements that we have for sequ

uh as as an engineer I have to agree with you as like someone who cares about ethereum like this is exactly my concern like to put it Wonder ethereum into a KC chain I I I wonder whether this can be sequentially addressed though it can like once you do something that that you know like uh that is easy to do like you you and like takes ction it it's like so basically today ethereum is is decentralized right like whe you have validators you have like everyone is building something in a in a decentralized manner you have some people who follow the the censorship

some people who don't like the tornado cach blocks like I'm not advocating for tornado cach for the record right but like just using them as an example example build right in the world where the uh ethereum is essentially sequenced by a few big Tech providers there will be zero blocks with tornado cash and with like whoever those three guys don't like or whoever their bosses don't like like this is this is just reality like this is what's going to happen like take me on record this is what's going to happen if if this is the future of going down like without first building something that like makes it makes the sequencer not able to be evil they will become evil we've we've been there with

the internet we've been there with blockchains here as all right so like this is precisely the concern like this is why ethereum has spent so much time andit Bitcoin I mean the entire Bitcoin philosophy is based on this like Cypher punk in and out like above everything else right uh ethereum initially has spent a lot of time in thinking about this specific problem how do we prevent censorship how do we prevent capture by big Tech by by like the centralizing forces like how like uh why not just like increase the block space uh as much as you want so that you have a few entity you know like you could do the same thing as like Bitcoin cash right

that was specifically deliberately not done in order to keep ethereum Cypher Punk or like free you know Sovereign decentralized this development for for the reasons George which you mentioned is exactly what I'm concerned about because like it tends to to like we tend to go down the route which is easier in which like Market forces are pushing us and if there is no push back from the community this is exactly what's going to happen Okay Mak sense so Alex I guess what you're saying is that there's there's path dependency here so either we figure out censorship first or we risk like never figuring it out um I mean from my perspective I

think we already have uh you know a censorship problem at at layer one with with with the builders um but that leads to what I call weak censorship where basically transactions have delayed inclusion but they're not completely censored from the chain um and I don't think we'll ever have strong censorship now weak censorship even though the the name has the word weak in it is still you know extremely important to tackle and I think the two best technologies that we have are number one inclusion list and number two encrypted men pools um and I think Think Like These are being like worked on and also prioritized so for example Mike Neer is

really making a big push towards inclusion list um now inclusion list admittedly does require some sort of special path some sort of uh mechanism whereby you know transactions can be gossiped to a public manool and then picked up uh by the by the proposers um as opposed to the pre-confirmation system which is more of a peer-to-peer endtoend encrypted connection with a well-defined entity and so I think you're right like we need to be building both paths in parallel and not just one otherwise we risk ending up with just the one path which as you said there's the power of the default where like wallets will just bother integrating the simplest one

which covers 99% of users and then a minority risk getting sens it yeah so like I I think the uh inclusion lists work for specialized players like for for rollups who want to submit transaction like if a rollup is being censored inclusion list is probably a good way to enforce that this doesn't happen uh but not for the end users and so it's even more important to make sure that like if this is the market forces are pushing us down this route let let's just make sure that there are alternative ways to build rollups and people who want to build them are not excluded completely from this economy and and there are ways to

interoperate right I mean I guess one observation is that this concern around censorship applies generally to decentralized sequencing not just to shed sequencing if you have a decentralized sequencer with PBS you want to have a censorship resistance picked in and yes and so I guess that's I think there's consensus that we want to have decentralized sequences and so as a community we definitely need to to solve this and we have spent some time I'm thinking about this problem and we came to conclusion that you can't really decouple s ship resistance from the protocol itself so you cannot build a uh like a sequencing layer which is completely like agnostic to towards the fact

whether like it must be part of the protocal mirat you have your hand up yeah thank you um so we we've we've been thinking about this particular issue as we had applied for uh The Tao grant for the pre-confirmation on L1 and based on on some of the discussions that came out of this group actually we positioned those to be a potential relay providing these uh pre-confirmation and then creating um essentially a constraint for the builders that the relay is working with so that's kind of the basis of our proposal there that Tao transactions would essentially go to um you know a relay with which would then uh create some some constraints on the builders

that it's working with and this sort of works as a as a you know minimum viable inclusion list if you will so I'm wondering if there's a a line where we can draw somewhere saying hey look uh this this this path is satisfactorily kind of censorship resistance enough to move forward or um and and the question is to Alex do you think that falls under the same kind of path dependency where uh it all breaks down because it's not fully censorship resistance path I don't quite yet understand the idea I would have to turn more okay yeah uh maybe we can do that uh offline but then but yeah uh for the community uh you know there's active work in in this area where uh we're working with we're

talking to relay to figure out if we can enforce some rules from the relay to the Builder uh perhaps in a in a less kind of um profit decaying way with with L2 sequencing and and and um and other kind of uh less defi or mevy uh you know constraints common did you say you had a Pres presentation ready yes it's if you want uh you can have uh 5 minutes if you can fit it in uh can you see the screen yes we can see the screen yes the presentation right uh not the we can see your calendar ah okay just a

second uh how about now uh it's loading uh yes okay so I'll try to be very brief so a prerequisite for Universal synchronous composability is layer to interoperability and tying into the previous conversation we have to create uh shared sequencing interoperability standards in a way that not everybody has to participate in Bas sequencing but we have to let every flow room basically so uh interoperability is making about layer two's

borderless and layer two standards and interoperability fit into our previous conversation based on this diagram so it's kind of adjacent to it it's not directly uh the same topic as sequencing and I want to present the deposit sharing idea today which would allow interoperability between different layer twos but just this ties into our previous conversation with universal synchronous composability because Universal synchronous composability is only possible in the intersection of the two two topics so we need both based sequencing for Universal and L2 standards and inability and L2

standards would allow uh asynchronous inability between different chains which would already be a huge benefit for them so let's look at the basic problem in this area the biggest problem is liquidity fragmentation if you have two different chains then they will have two different bridges on layer one and if you're transferring value between one rule up to another then you have to transfer it on the bridges as well this problem has to be solved and the biggest question when it comes to Layer Two interoperability is how do you solve this uh layer one transaction

problem so for a single ecosystem just to get everyone up to speed inside a single ecosystem inside a different uh single governance and Trust Zone you can use a shared layer one shared bridge and a shared transaction route this transaction route commits to all the transactions happening inside the ecosystem and using this you can do what we call hyper bridging other ecosystems call it LX ly bridging but it allows cheap and trustless transactions to go from one roll up to another so this is possible inside a single ecosystem the big question is how to do it between different

ecosystems our solution for this is an shared Bridge coupled with an asset verifier this is Alex's idea and basically what this would allow is uh permissionless VM verifiers on layer one which would be decoupled from verifying the fact that each chain will have its balances verified together in the shared bridge and using this system uh the VM verifier can be permissionless deployed arbitrary uh this the transaction route would still commit to the

uh transaction inside the all ecosystem and the bridging process between the different rollups would still be the same so this is the first idea and there's another idea and this is jordy's idea which is to have mass deposits between different Bridges basically uh in this idea a roll up would commit to its transactions in the transaction route and then there would be a mass transfer from one bridge and it would go into a mass deposit Handler contract on layer one and the destination rollup could then withdraw these Mass deposits and uh

could then receive the transactions from this Mass deposit Handler so basically I wanted to show these two ideas to get this conversation started and to uh have people understand the two basic options that will be necessary that are the two options for Universal synchronous composability we think that asset verifiers are better than Mass deposits they do require a bit more coordination because the asset verifier uh which is a proof verifier contract has to be standardized this could be a multi verifier situation making it less easier uh making it easier to coordinate but it

has uh multiple advantages and uh basically the next steps are to decide on how to solve this liquidity fragmentation problem so I think this is the first step we should discuss it go over it and then there are further steps but because of the shortness of time I will uh skip these so thank you very much thank you cman so I guess what you're saying is that in order for Universal synchronous composability to really be practical you know in terms of transfer Assets in and out of these roller ecosystems we need um the asset transfers to be much more efficient if you go through the L1 every single time

it becomes extremely expensive and there's basically two solutions we can either do batching and you have these Mass deposits um or we can have some sort of shed deposit layer with a minimal verifier who's only is responsible for for for the asset set and not the rest of the of of the logic um thank you so much I think uh this could be a great uh discussion starter for for off offline discussions or or future discussions on these I I created a group I was share link in this uh chat here and then uh you can guys add anyone else who is not participating in call and maybe we can start discussing there I think this would be an important addition to the

like work in parallel to Shared sequencing to make sure that we're make okay perfect thank you Alex uh we'll make sure to include those in the in the PM notes on on GitHub um I guess now is the time for some uh announcements and and uh and updates and teasers from various teams doing cool stuff so I guess I'd like to invite uh espresso first to give a a three minute uh pitch on on on what they've been working on sure Justin I don't think I really need three minutes um but uh I and and I was able to highlight this a bit in the

context of Alex's um uh you know presentation as well but what ispresso is creating here is more of a Marketplace uh for rollups Sovereign rollups to be able to participate in what I call ad hoc shared sequencing um and uh you know I I don't I know that there's there's been um you know different debates on is there going to be value in shared proposers proposing for multiple rollups at the same time maybe there's no maybe there's no value I mean this is this is uh something that ye brought up in the chat as well maybe there isn't cross domain meev I think the value in being a shared

proposer for two rollups goes beyond me it goes It goes to everything that has to do with pre-confirmation uh as we've been talking about and maybe between some rollups there won't be value in having a shared proposer um a Marketplace is something that can determine this right um You may have rollups that are not selling block proposal rights because nobody is bidding on them um and they're just producing their own blocks you may have other rollups which are selling their block proposal rights to Common third parties that end up um proposing for multiple rollups at the same time and this is the framework through which I view uh based sequencing as well because base sequencing is a special case where the

L1 proposer is able to create Surplus value for rollups by being able to simultaneously construct L1 blocks and rollup blocks at the same time you might have some Roll-Ups that don't benefit from that right you might have rollups that uh really don't benefit from synchronous composability with L1 contracts um in which case those rollups may not end up selling their proposal rights uh to the one proposer and you know what maybe those rollups will at some point um find that there is a reason in some slot to do this and that may dynamically happen so all in all espresso is really viewing this as let's just create a Marketplace

what does a Marketplace require well it requires coordination um at the level of we can advertise our block space and if some common proposer is able to buy block space from from multiple rollups at the same time to become a shared proposer then that need needs to be recognized right so we need some kind of finality over that um what what enables the L1 proposer to be able to participate well the one tricky thing is that if you uh if you run this Marketplace too far in advance then the o1 proposers aren't known right so on the other hand if you if you don't run the marketplace too far

in advance then pre-confirmation um you have a you have less time to run kind of this sale process and so that you might run into efficiency issues and also uh you have you give less um you know of of a look ahead time for for parties to start giving out pre-confirmation of what they're going to do during their turn so if you want to run pre-confirmation you know further then say sorry if you want to run this proposer uh Marketplace further than 32 slots in advance then you can always just give the L1 proposer what we call a right of first refusal which is the ability to buy uh any proposal rights from any other player when it becomes known that this will be the L1

proposer um and so finally uh what makes a rollup a base rollup versus a non- based rollup within this Marketplace is simply their decision on whether they are going to reorg with the L1 so allow the latest L1 block to uh um have an impact on the rollup versus not reorg on the L1 if you don't reorg on the L1 you could still sell your uh block proposal rights to the L1 proposer but it would not allow for synchronous composability with the L1 um the the synchronous composability only comes from the ability of the L1 proposer to uh you know simultaneously construct the L1 block and the L2 Block in such a way

that actions it takes in the L1 block can actually have an impact directly on the rollup you can do a deposit and withdraw let's say from the rollup in the same block uh it also relies on realtime proving um or perhaps something like proof aggregation if as long as we can get it down to less than 12 seconds um so that's a that's a quick summary and I I'll be posting a I I I've shared the uh preliminary version of a hackam depost that describes this idea in this channel multiple times but we're actually just publishing a Blog on this today um it should go it should have already gone out in fact and I'll send the uh asend the link towards the

end perfect very exciting looking forward to read the blog post and we'll add it in the the show notes uh links I guess next up for a real quick teaser is Alex from the ultra sound team yeah hi um yeah I'll I'll keep it really short and just share what uh we as ultrasound see for areas where we can contribute i' say we're very much like in the exploring phase uh for how to contribute to this and interested in building or helping build an MVP of of doing this um so it say like pre confirmations or they is really close to uh the block building and pricing especially the tips like it'll require a

lot of Hardware a lot of specialization and that's not something that all proposers will have like equal access to and as a proposer you'll have to pick a party that you're doing these pre confirmations with um and then even if you only know relays for how we're trusting them and that's what you don't like them um if you want to minimize trust really are already this party that many proposers are looking at closely to decide whether they uh they trust them um and then this is a role that that a relay could fulfill as well and then second of course is reliability um especially if on the user side we get very used to having these very quick pre confirmations um but also just for the the MV to arrive uninterrupted to make sure that that block Builders are informed on on what transactions are

prec confirmed um and then finally of course that this this product that they build together that that lands on chain this is all stuff that you want to be done like very very reliably I think we recently hit like a million blocks landed so very happy with that um reliability seems to be important I also Imagine I heard censorship resistant came up um like in protocol much nicer but it would of course be very unfortunate if one of the most successful one of the most liked prec confirming parties is also the one that likes the sensor a lot um so I think that's another area that that we take very seriously and where we might be able to contri contribute by you know following the ethereum ethos of uh doing

something about censorship um also the the pricing of pre-confirmation tips the the rebates that people might get um relays are maybe not the part to do this but we're already connected to the parties that might be really good at this the block Builders um so I see that as a benefit um and maybe realas might even be able to facilitate a sort of neutral tip rebate uh auction mechanism I think that could be really cool so yeah finally as um if you as a proposer want to unlock these these things for the ethereum network the virtually instant transaction free confirmations uh the better composability the better atomicity um and you want to then you're going to need some

infrastructure to do that and then if you uh if you want that to be infrastructure that you can trust minimally um we want at least allow proposers to be able to do that with us so I think that summarizes it perfect thank you Alex um next up I want to give you a quick announcement on some of the events happening in London uh in the next few days so on the 12th 13th 14th and 15th there's four different uh me events so it be a packed week ahead of e London so if you are in E London uh please please send me a DM uh the one on on the 12th is me. Market on the 13th there's a there's an event that's kind of tailored

for the uh the rollup project so if you're part of a rollup team uh there will be a whole day of workshops dedicated to to base sequencing and base pre-confirmation and we have uh most uh rollup projects represented um most of them actually with their their Founders coming so uh if you're part of a rollup team but not yet coming and you are in London please uh send me a DM and then uh next up um we have George uh from lime chain who will be talking about uh his new design uh teasing based tickets

George I think you're muted sorry yeah you see my screen yes we see it all right uh we'll try to put in in five minutes um so over the last couple of weeks we've been working on an idea that we call the base ticketing grup um it came out of several realizations first of all is that we like a lot uh based sequencing uh rollups however uh as we were exploring more and more about prec confirmations we figured out that like there are certain pieces of the concept that uh make it hard uh to start with this

concept and also uh make it a little bit hard to actually adopt this concept one of these things is for example the gap between L one proposal that are actually going to be to proposers um and with the uh help from the folks from from Skull namely Mo um we were able to come up uh with a concept that's called Bas ticketing and I'm going to give you the tldr of this uh of this concept it is essentially combining the base r with execution tickets uh concept uh any rollup can decide to offer two types of tickets uh one type of ticket is called The Proposal tickets uh the other types

of tickets are called the enforcement tickets so the proposal proposal tickets uh the the parties that have hold the proposal tickets actually eligible to be drawn to be the next proposer of this rollup the basic rollup and uh they can essentially propose the transactions list the enforcement ticket holders actually have the rights to constraint the block that the the block that the proposal uh ticket holder is going to propose so the enforcement ticket holders the enforcers can actually provide R SX services like prec confirmations but not just regular inclusion prec confirmations but they can now start constraining the uh

upcoming block to uh things like Post State uh commitments and even to um stuff like Universal synchronous composability so if there are multiple rollups that are based ticketing growups if a single entity actually holds the proposal and or enforcement tickets of multi of these rollups at the same slot they can actually ensure atomicity of both uh both rollups and they can ensure the um Atomic execution and um messaging between these between these rollups um so this concept essentially tries to leverage the best from both base uh sequencing and execution tickets

obviously there is much more than this uh I'll throw the link to the first draft uh We've we are currently is supported by thean foundation for This research so all of our research is going to go live uh quite soon even even even this draft uh we are pretty much halfway there in defining the various uh flows that are there and also exploring the various mechanistical infrastructure that is uh needed in order for this to to happen uh and the changes around the M pipeline so yeah stay tuned perfect uh thank you George um I guess we have um a couple minutes either for

questions uh or for another project who's been doing uh uh based sequencing stuff who wants to uh to to have a an opportunity to present their work Muhammad yeah just want to ask a few questions from the audience so uh one is the when we talk about synchronous composability are we talking about the roll different rollups being able to synchronously read and write to the state of the other rollup or are we just talking about this specific type of you know estate manipulation that's uh sharing liquidity and transferring assets which is only

possible if we put that subset of State in very specific location because like from the presentation of the cman and some some thoughts shared by by Alex I'm not sure which direction we are going uh that's one and second we it seems like there's some discussions in the chat around like uh whether meev is the biggest thing we want to get out of this Shar sequencing like C chain mu or whether Market forces that are forcing us to do this sh sequencing are press Ms and so on I want to put this thought out that uh like if we enable synchronous composability in

the sense that you could manipulate other change State uh then like we enable U A New Market Force that makes you know Shar sequencing very sexy and like without that it's quite kind of useless in my opinion and like it it boils down to figuring out if we want this like a strong notion of uh like synchronous compossibility or not and whether this is you know in conflict with other itos like sensorship

resistance and so on I guess like focusing on shared some layer for shared assets or easier transfer of asset is like missing the bigger point I'm not sure what are import about right okay let me try and answer those questions so I guess the first one is like what is the definition of synchronous composability and for me is the ability as you said for one rollup to have access to the state of another rollup to either read the state or to write the state uh but this is not done explicitly in the sense that there isn't like one transaction that will um you know it can't have rollup a within its

own rollup a transaction read and write state to to rollup b instead you have this notion of a super transaction which is the concatenation of subtransaction and each subtransaction is specific to one one single rollup and uh the synchronous composability is basically simulated or emulated or given to you by the the shared sequencer um because they have Monopoly rights to sequence on all the rollups and so they can make sure that these subtransaction within a super transaction Are all uh as they as they should be now in terms of reading the state of another rollup this is done trustless using SNS so when you have real time settlement you can pass the

state route plus a stock proving that it's correct um now in terms of I think the second question which is like are we going for just simple asset transfers or are we going for something much more expressive like we're going for like the the thing that's Maxim expressive so you can do arbitrary calls from from arbitrary contracts between uh between arbitrary rollups so it's it's a little bit like today where you could have any contract called Unis swap or a and have like an immediate and synchronous uh uh response uh but all of that is is simulated economically um and so users have to pay

for the overhead of synchr composability which you know could involve uh adding locks so if you have a super transaction over rollups A and B um that that is blocking both rollups simultaneously while the user is going to have to pay gas on both roll apps uh while that super transaction is being executed even though internally to Super rollup there's only one transaction that's executed at the time and so that that's going to lead to a market uh where basically you have the the sequences selling uh synchronous composable locking mechanisms and you have arbitragers and users and whoever it is who will be buying those virtual

locks yeah I mean just one more thing I want to uh also like say that like there it's not necessary that we have like this super transaction kind of design necessary like um um from my perspective there is like this stman construction I call it intermittent Fusion like say all the rollups agree on some very simple VM and this VM is some extension to their like VM rules or some subset of them VM rules all of them agree with and every 10 blocks or every 100 block or every once in a while that somebody is able to pay off all these rollups this small VM comes into play and manipulates the estate which you know some values

written some and they all finalize on this you know intermittent fuse infused like a small VM like I'm not saying that this is a perfect design but I'm just saying you could have like a design without super transactions per se that does this I guess um yeah I mean I think this is uh what the asset verifier is all about you have like a minimal VM uh which only handles uh the asset deposits I think that's the ZK sync idea um like one of the downsides of like sharing VMS now suddenly you you share bugs you share secur security assumptions you might um and so really what you want is have very very clean boundaries between the

rollups so that each rollup has different security assumptions and has its own independent governance and whatnot exent perfect uh I think we're at the hour so the public call is finishing now but we have a half an hour of a after um so I guess I'll open it up to the floor to uh whoever wants to to bring a topic up