Messari Mainnet Convos: Learn more about Quantstamp and their perspective on Optimistic & ZK Rollups

Transcript

foreign let's talk about the presentation that you did yesterday because that that was awesome um and it was evident that there was a lot of research a lot of it was very thoughtful like the amount of like effort that you put into it so sure um maybe we could start about kind of like your approach to putting together this framework for assessing Roll-Ups sure yeah so let me let me give a little bit of context and then I'll tell you about how we did it and uh you know sort of what we want to do still and what uh how we think it can be helpful so um we we being Quant stamp we got a grant from the ethereum foundation to uh study the security of Roll-Ups um you

know obviously they're pretty popular these days but they're all very um custom in a lot of ways right there there's a push towards you know certain common features like evm equivalents or things like this but the reality is each Roll Up is at different stages of development and metallics have a nice post about you know some actual stage classification um but beyond even that we we'd looked at this in the past and saw that you know a lot of them have road maps which don't necessarily correspond to those those stages perhaps um and they were sort of advancing at different paces and different places and and so we thought well that's definitely going to cause some risk right like people are gonna assume something about these systems and

those assumptions are are at least occasionally going to be wrong and we wanted to figure out how we could uh help the community you know make fewer wrong assumptions and make sure that as a result you know systems that they are using are ones that they're comfortable with in terms of risk and Trust before we before we dive into like some some of the the assessments that you put together could you just kind of rattle off a few common assumptions that we tend to make as users or developers of these Roll-Ups yeah I mean like is it at all possible that my funds get locked right is it always going to be the case that there's a state proposer online to post a new State Route um is it you know the case that my

transaction will always be pushed through or will it always be pushed through in some certain time frame um these are the most common ones right and and in some cases the answer is definitely yes there are systems out there that have layer one like inboxes that the sequence are sort of supposed to pick up from after a certain time frame so you can avoid censorship in this way and in some cases that also lets you withdraw funds in some nice way but uh if you if you're new to this space you sort of think that it's like ethereum itself where you know you submit a transaction provided the gas is sort of sufficiently high it's going to go through eventually right maybe you'll get they're very quick if you pay a lot maybe it'll take a little bit longer if you are cheap and you sort of go at the

bare minimum uh but Roll-Ups you know fundamentally because of these these other components that they introduce don't have that by default you have to build that in because the liveness assumptions is that like as long as ethereum where you know the smart contracts for the Roll-Ups are running uh is running then you sort of assume that the roll-up should also be running but if the smart contract is blocked waiting for something to come from a component that's offline like a sequencer or a state proposer um that assumption is not there so you have to actually build that into the into the smart contracts uh and the role of design itself and I think a lot of people still take that for granted I think that's probably the biggest awesome okay okay so that that gives us a few of the common assumptions

um when it comes to looking at these Roll-Ups and then how to think about like okay if these assumptions are are necessary and if they are like how to make them as like safe and reliable as possible and then even if they're not necessary how to remove them those assumptions yeah completely so I mean I think a lot of the assumptions uh you know are sort of fair in the sense that the perfect roll-up design will probably have a lot of them countered right there will be things like Escape patches and there will be sort of ways to handle uh failures of these components right so you could for example have a Time mode saying if you haven't seen a block come in from the layer two in a week you know anyone gets to proposal Block in which case you know in principle assuming no Hardware

requirements are really high you can be that person and so you can sort of get your own funds at least right and you can also maybe serve the community depending on you know if there's an incentive to do that or if you want to do that uh so these These are sort of I don't know if they're required assumptions but sort of the design space requires like doesn't prohibit you from making these assumptions in the sort of Ideal space the problem is these Roll-Ups are pretty new most have you know been in development for quite some time and they're still not complete if you if you look at sort of a complete road map no no roll up out there I think is is on the website saying we're done like this is it we're not making any more upgrades you know everything is good everything

will always be what you want it to be they're all usually missing some level of functionality right if you look at you know some of the optimistic ones they're not they don't all have fraud proof systems if you look at the ZK ones they're you know better at proving every sort of block but they're not necessarily having Escape hatches right so it's very very uh rare that you find one with a lot of them so like I think there are a couple out there that are pretty complete certainly if you look at l2b there's a it's a great website by the way uh we'll link it yeah um they you know they have this pie chart which is where we started this framework so to go back to the framework for just a second and I think there's only a couple on there that are fully

green which means they're sort of like guarded against all of these assumptions and so you know that's that's the ideal World everything's always green there but as I try to explain some of the framework issues that's not totally sufficient that's not to say that lgbt's done a bad job there's just a lot that these systems can do and not all of them are sort of big security risk some are very minor and some of them are assumptions that aren't common and if they're not common assumptions of course it doesn't make sense to always present that information in a very clear way there's a lot more to these Roll-Ups than you can present in a pie chart yeah for sure for sure so you know when we when to go back to the framework then when we started looking at it we wanted to see exactly what l2b

was missing because you know people would say well why are why is someone building a new roll up if you know these ones already exist and starting to ask the questions of why you know there's obviously reasons for some people to do it you know like they just want their own chain or they want money or whatever but in other cases it's because you know they think they can do better and why do they think they could do better well because some Roll Up is missing some functionality and feature or you know they're they're they're not being fully decentralized so they don't have a plan to be decentralized or things like this and and so we started looking at those differences and specifically the security assumptions right it was sort of right around the time where I think uh on one of the chains one of the ZK

chains someone had made a contract that or deployed a contract that it was supposed to work just out of the box and the gas limit had changed on it I can't remember if it was I think it was ZK sync I don't know if I should name the album and offend them but uh like I'm not trying to like throw them under the bus or anything like that it's just it happened and and they sort of needed the project who who deployed the contract ended up getting some funds locked in it and and sort of Zeke I think needed to step in if it was EK I think I apologize if not I can send it to you later with the specific instance um they had to step in and sort of like recover the funds in part they had to give a hand in doing that and for big

projects and early stage Roll-Ups with a vested interest of you know progressing to be really mature systems that is totally fair but if you could imagine you know you don't have the cloud or you didn't raise millions of dollars you only raised maybe a couple thousand bucks on your project but maybe that's more than than you've seen in a long time or you know that's all you need for your project uh to get off the ground maybe someone wouldn't have been so accommodating right you'd like to think ZK sink or whoever was you know doing that would still step in but if they were too busy or they didn't care like you would have been sort of out of luck you would have had to do something different and and you know you probably would have lost some users from it and and that's not good because now these

these Roll-Ups have some control over not just like their own security implications but also you know they're affecting projects launching on them and so these questions specifically ones that that weren't present on this pie chart like is something evm compatible is this gas pricing a little bit different are there other things like operational concerns so you know ZK sync they did a great job and they responded to that but you know in a world where we have 20 or 30 or more layer twos which is you know from the looks of it gonna be happening since we're already in probably 20 on lgb.com yeah um you know not all of them are going to have a good BD track record none of them are going to be able to to provide that support they're going to be stuck uh dealing with other bigger clients who

they care about more or just you know overloaded with other things or they just won't care some people like we saw really bad you know things in the space before we saw people Rob people all the time we saw nfts that were just in it for the quick cash grab I wouldn't be surprised if there's going to be some cash grabby type uh Roll-Ups out there right I hope that's not the case I hope everyone builds one with a really good you know intent and reason because these are a more complicated so it would be kind of impressive if someone did a long scam like that but on the other hand you know because they're they're so useful like we don't want to sour this I'm not saying enough tea is more useful but you know flipping jpegs got boring at some point and I you know I won't share too

many opinions on that uh adding more block space so we can scale ethereum is less boring right maybe maybe it's boring at some point uh from sort of the business point of view but technically it's still very interesting and technically still very necessary so even if it is boring well it's necessary so you've sort of got to do it right yeah and when it when it comes to I I think that uh is probably a word of advice to developers who are building at the app layer uh which is choose your roll up wisely yeah um yeah yeah so I mean the whole point of the framework and and I'll send you the link to the framework so you can link it too uh the whole point of the framework is so that people can and make these decisions and right now it probably needs a little bit more work we

need to make it more interactive and a little bit less dry because it is fairly long and it's it's not something you can complete in you know three minutes it's going to take you some time to really evaluate this but what we found was at you know without even this dry lengthy thing people had no real idea of why to choose one roll-up or another they'd often go where tvl is locked up and that that's not necessarily a bad thing um but that pvl should also be there for the right reason and you know for all you know the tvl is not there because they did their due diligence they just you know aped into it or whatever and we wanted to provide a way to say you know this one seems like it's probably not going to screw you and you know if it isn't going to screw you

here's the ways it can kind of do that right because we're a security company and you know ideally we would review every one of these Roll-Ups in depth and say oh this this is a bug please go fix it you know or tell me on the road map when you're gonna solve it but in there's a lot of Roll-Ups and they're really slow like not they're not slow to operate they're they're slow to review they're big complex systems so we can't do that and so if we can make this open so that other people can do it uh you know maybe it's not as good as having an expert review it but it's something and and if that just protects even you know a handful of users it was well worth the time and effort and the money to do it okay I I have

a couple more questions on the framework and one specific question about a security assumption that that you've mentioned sure and we'll we'll switch gears into some higher level Concepts so the specific question about one of the security assumptions is an escape hatch I've heard you mentioned I heard you mentioned last night uh and then you know in this conversation now so what is an escape patch and why is it so important so yeah an escape patch is usually a functionality built into the layer one smart contract of the layer two com uh system itself that allows you to withdraw funds or you know other assets that are important even if the sequencer estate proposer goes down so usually what happens is you know the roll-up is functioning for some period of time you bridge some some things into the canonical Bridge everything works out but if something goes goes bad with

that roll up you still want to be able to get your funds out of this and when you say get funds out and withdraw that means basically get to a different safer chain no I mean it could be a migration to another state but a chain but usually it's drawing back to the layer one like the layer uh on which the roll up has been built and for Roll-Ups that are built from ethereum Centric architecture that would be back that'd be exactly okay yeah so I mean I we've done we've published one paper on on these these ideal properties and you know it might be nice to actually escape from say optimism to another optimism chain if something were to go wrong um because you know it might be more efficient well we would also what we're scared about this is after off topic of

the specific definition for what we're sort of scared about is if someone needs to use them thankfully no one has really needed to use this and it's it's sort of a feature that if everything goes right no one needs to use it's almost a waste of engineering time almost if you know if everything else is good but it's it's you should have one I will definitely recommend always you know having one if it's not I think they are difficult to build and that uh but we are also like I'm personally scared of of the situation where someone actually needs to use one and all of a sudden you know everyone is trying to get off the chain at the same time you're going to get this congestion on layer one it'd be a gas War to get your funds out and you know this doesn't seem all that impossible

impossible because you know if you actually see a layer two with with massive adoption and then if it if it did God forbid crash some people would want their funds are really quickly for obvious reasons they just like to have their own money because things could maybe go worse so yeah that's the functionality yeah and we we don't I mean we like to think of things as as self-custodial or non-custodial and when funds can get locked up in a chain and we don't have control to move our funds anymore it's almost like this this chain is custodying or funds but the funds have to live it's this like almost weird social

like uh connection that that we have where like money can't really live in a silo because then it's not used to exchange for anything it's like it's Gotta kind of have this like social connection to other money for it to be useful yeah so we I think I think on Twitter there's been a couple conversations where you know people said well the layer 2 Chain itself has actually its own blockchain and it doesn't need to be sort of tied to the layer one and so the Escape patch maybe isn't strictly necessary uh and I think technically that might be true but if you have an under collateralized chain because you don't have the the cash available because it's locked and you sort of socially can sense your way like made a consensus to switch what the train is you know maybe technically that was feasible but socially you know if I

don't have the same eth it's a different copy of eth it's not very useful to me right now maybe there's a world where that is useful um I would guess most of the time that's not the case I think this was a conversation by uh John Charbonneau and Calvin Victor versus sort of like togirl from scroll where you know the I think the scroll side was bridges are essential and critical and the optimism as well it's just not it's just another chain and there happens to be a really nice Bridge that's very useful but the bridge is not critical uh but I would say the bridge is somewhat critical because you know that escape functionality is what makes it more than just a little bit of a bridge it's sort of a bridge with this extra guarantee that if this specific Bridge stops operating in this way you can still get

something out if this other functionality exists which you know in the ideal situation it does so I think you know the social consensus is very true technically but you know most of the time you're going to want to not have to get Society to re-agree on everything because that's that's pretty tough yeah so well yeah there's a lot of coordination that goes into consensus um okay so zooming out a little bit and then thinking about like from what I've seen there's two uh there's two popular mental models when we when we think about kind of like this this map of networks that are that are out there

um and the escape hatch I think points a little bit more towards the layered approach where you have the layer ones the layer twos they roll up you can sequentially imagine this paves away for layer threes fours fives all rolling up onto the the lower numbered uh uh chains that are that are kind of closer to the the layer one essentially all roll up and and flow uh down the stack yeah and then there's the modular approach um and this is kind of like not a layered hierarchical approach it's more just like these Roll-Ups operating in space um how do we reconcile these two mental

models that's a good question you know I I've spent some time thinking about the more modular approach um but I guess I'm more of a classical roll-up person where I get this this layered model of things on top that's not to say like other types of of uh Roll-Ups like Sovereign Roll-Ups or trying Roll-Ups or whatever you sort of want to consider as a more pick and choose type approach aren't useful or helpful but I I don't know that they need to be strictly disjoint in the mental model either right you still could envision a roll-up with different layers except now it's sort of like your layer one is fragmented because maybe you know you're settling on ethereum but your data availability is going somewhere else that that still sort of implies that there's some level hierarchy some level

of hierarchy to Me Maybe it's not quite clean maybe it's more of a sort of a acyclic graph that sort of works out I don't think you should probably be a circles there it's probably going to go bad so I think I think there's some way you could actually you know get that hierarchy to be kind of consistent but it is probably less clean so you know I I don't think it's it's super problematic and I'm I'm not giving a very precise answer because I'd have to go like really Define these terms and figure out what they would what you know we're talking about in terms of modularity and things like this um but I think for now we're sort of fortunate that we don't need to be super concerned about it because you know layer twos are just coming online like

they're pretty new in in the grand scheme of things uh and layer threes are still very much new right I don't know if anyone actively using a layer of three I could be wrong I mean uh but it it seems like people are just getting layer twos to a point where everyone's really comfortable using them and until that block space you know gets full or until you have very specific app needs layer threes and app chains you know are our engineering for the sake of engineering and I'm not strictly against that it's just that there costs so much to develop that you know maybe maybe they're not worth doing right now and I think we've seen that like again unless you know of any layer threes I don't I've seen I've seen their projects take this app chain

approach um I'll name drop a few maybe we'll scrap these out of the written piece later but zkx is a Perpetual decks on Stark net um they they announced their app chain in the last couple weeks um they have there's um SX Network which is a decentralized Sportsbook uh rolling up onto polygon like polygon sidechain um they consider themselves a layer three um so we're seeing kind of like this app chain thesis and the the piece of context I want to bring into this and then and then definitely Keen to hear your thoughts is that um [Music] this app chain thesis may end up

providing revenue and value because you are able to drive revenue from Mev and the ordering mechanism yeah um but that takes a that their startup costs associated with that so um so yeah I mean I think I think you raised a good distinction there which is like the app chain as a specialized type of layer three which I think makes a lot of sense because you probably occasionally want really fine grain control right if you're running a decks or a sports book these these types of ordering things probably matter a lot to you and you want to make sure they're done in exactly this way and very quick and users can can respond as quickly as possible to it like there's a little

latency in these systems that's what you want um in part and uh for the app chains this makes a lot of sense right because then you can optimize that specific functionality whether that's on the technical level of performance like if you choose the right Layer Two to build on uh or the sort of ux of like well look we have our own wallet because this is what we need or something like this and you have total control and total customization so you can optimize the technical land or the user experience and or both and and that would be something that's going to probably make a lot of sense for a lot of domains like even gaming I think is going to come right you're probably going to have gaming options because you want to have your game be as performant I think you know as it can be and to do

that you want some level of its own chain but you also want you know a settlement layer that is universal and well trusted and so a nice layer too is definitely going to be there um the general Layer Three is I'm I'm less sold on right now as a need but yeah so the app James will come up and you're right there there is startup cost I mean I think there is uh there's a lot of effort to make that work but if you have a good enough use case it's worth doing and and with the right underlying layer too like a fully developed ZK Layer Two you could definitely get good performance and good finality time like withdrawal finality so that even if you have users who are who are not expected to be isolated on

your app you can still get them to bridge through ethereum relatively obviously not necessarily instantly if they have to go all the way sort of down to the layers and back up but you know you could get this to work quite well so the app chains make sense and there is going to be a cost but I think for a lot of people it will be useful what I'm not sort of I don't know if I've said this already now not really sold on the layer Series in general like just putting like another start in it on a stark net like that's cool but sort of why I I don't quite see it just yet uh it's nice that it's technically feasible I think that's really really cool but uh these other use cases yeah you're right I think that's where it'll it'll come

through where you have very clear incentives and you want control and we're going to see a lot of those but then I I almost wouldn't consider them layer of Threes because to me it's sort of a real layer is a layer that you can build on top of and so unless your app chain is also building a framework of some sort of programming language like you can't build on top of it the layer ends there it's the end use so which case it's almost indistinguishable if you squint your eyes you know from a from just a smart contract app it just happens to have more octane components which you know complicate the Engineering Process and and hopefully for the right reasons that's either incentives for the project or usability and you know good experiences for the

end users yeah very cool um yeah I think this has already expanded my horizons uh so anything in general that that like I don't really have any other specific questions um so just kind of like based on your experience diving really deep into this subject what are what are you know anything else that uh you think is relevant like for this conversation uh yeah let me think about it for just a second here yeah uh just in general really just any any thoughts you feel looping in yeah so I think I think one thing that I think will be very interesting to see is some layer twos

who sort of deviate from the norm in terms of the CVM equivalence in order to get better support for these things like app games right like layer twos that support programming languages that aren't necessarily solidity for example we'll do a really interesting job of this I say interesting because it could be very good it could be very bad but I think it's it's most interesting not because of its technical sort of challenges I'm sure there are many um anything like there are projects doing this and cartesians doing like Linux on on the blockchain and things like this uh the real interesting part there is because they actually give an opportunity to onboard a whole bunch of like webview developers right there's so many developers working on the Linux base and and I'm sure ethereum's you know captured some of them but there's going

to be a lot more if you can capture even more and it doesn't have to be just Linux it you know it can be what other whatever other language so for me my personal thesis is a little bit of like let's move away from evm equivalence for the right reasons when the time is right and for some change that's now if they're willing to take that risk and for other Chains It's probably not the time and for some change it'll never be the time that's fine um but that will make it really interesting and kind of scary because those are going to have a whole bunch of different security assumptions and implications like you know people are going to assume that their programming language Works may be exactly the same way and maybe the semantics will change uh maybe they won't you know it'll depend on the exact

implementations um but that will be a very interesting day when when we have you know ethereum conferences and not everyone is necessarily a solidative development that's not that's not true now right you have some Viper developers you have some other language developers like people who just go right into half and low angles low level stuff but it'll be really interesting when you when you see ton more developers who are building in different languages and and then you'll get to see a lot of really cool Tech and stuff you'll have like new support for crypto Primitives you don't currently have and so things will get more efficient things will get cool uh there'll probably be a bunch more hacks that's you know that's the reality like when you when you build so many things and there's so much money on it people are gonna find you know ways to exploit

bugs uh that's scary obviously I don't want that to happen uh but the reality is I think it's gonna happen um and is that the reality of uh like building aside from evm equivalents um like if you're if you're building something unique uh that's not based on this pre-existing framework of the evm and it's more likely that you're taking uh security experiments almost I mean almost I I mean yeah experiment might be a little bit too strong like I don't think anyone's out there really like rooting for people to lose it just you know sometimes things don't work out the way you wanted them to uh or you forget to write your code properly in some cases you know it doesn't break up it is kind of an experiment I think that you

know the word might be a little harsh but yeah uh but it's for the right reason right it's again it's not it's not a literal experiment so you're not it's not an ethical violation to sort of do that but it yeah uh but maybe not right you could also Envision a world where it's simpler right this is sort of what the app games are doing but like if you if you really hated create two or create you know you could cut that from the evm and then it's an experiment but one with very little risk because you sort of knew what that train looked like it was out there before these op codes were introduced uh or at least create two so yeah it is a risk but there's also a lot of reward and upside right if everything goes right so it's it's one that some people should

take and I I honestly think with layer twos people should build a little quicker than ethereum right ethereum needs to be really secure ethereum's slow there's a lot of money on it like it's slow in the sense of building not not necessarily in latency I mean it is probably that too but uh yeah yeah uh you know ethereum has to move slow so that they don't screw it up because there's too much at stake on a single layer two you probably shouldn't push an update you know every week or every month but you can certainly move more quickly than you know one hard one sort of Geth client update per year or two years I I actually imagine ethereum does more than that but you know most of them are very minor changes but you could probably make major changes on a roll-up

somewhere quicker especially while you have a low user base uh and if you have a good community that understands that what you're doing is something that is kind of experimental and very much you know not yet battle tested because they're trying to do something like this new innovation uh they might not even mine now that Community probably then won't bring billions of dollars in if they were rational because you know you don't want to lose that but and it's almost slow them down yeah it was on there the more money they would bring out of the system it might be worse um but you know the you as a single app chain or a layer two sorry even really uh you can move fast enough here so you so I would like to see that I would like to see someone building an L2 that sort

of story is the evm in weird ways that we haven't thought about like almost random iterations on like gas code oh but just to see what breaks that would be a useless chain absolutely useless but it'd be interesting to see if you could do it technically and then really like get the community familiar with this notion of you don't necessarily need a roll of this evm equivalent although you're comfortable with one and you like it in the future it might not look like that so that thought experiment would be probably a waste of resources but maybe a fun hackathon project just to see like what would you redeploy and what would break and you can sort of make it a sort of a white hat sort of sandbox where you can play around with oh would this contact have been hacked

if this situation was around uh because then you can use that to guide future ethereum up so you can say definitely don't do this because look I now have data that says this is gonna go bad if you do this um are projects staying away from taking those uh Leaps and Bounds in their development process because they lose composability with the rest of the evm yeah I think so uh and and the composability is a big big feature it it's it's not something you should totally discard but um when you start to think about composability which sort of requires transmission through a bridge anyway so it's just specifically like app chains uh who sort of like to talk to another app we'll have to

talk to a layer too well you can have Layer Two composabilities that always have to go through layer one in which case you know you can still get composability it just gets more technically challenging to sort of conform the call and during the bridge to a composable thing and you could have a different form of compost ability which might be more valuable on your local L2 that's a lot of engineering work that piques my curiosity because like we haven't seen the interop and the composability across these app chains no we haven't and for you know they're they're good reasons it's technically challenging you know the things are not Atomic so there's a lot of like things that can sort of break in this process whether you know the fun just aren't high enough to relay the message or the message was malformed or you know some

other reason which could just be delay right you know it takes time to get things across chains but that doesn't mean in the future you know it wouldn't be so bad if it took 20 minutes to get a cross chain call between two Roll-Ups which seems like a high amount of time to be honest and they're sort of Ideal World that still could be a lot quicker than you know waiting for my bank to do it so even though it's slow and air quotes here it's actually pretty fast right so for the right reasons we can do this but uh I think everyone got used to like oh I'm just gonna do it right on Atomic transaction and good I mean that does help with a lot of things you can do probably better like defyer because everything's more Atomic you don't have to worry about things but

when we move past things that are so time sensitive to to other forms of apps like if you imagined a social network like popping a like on someone's the post doesn't require instant latency like you'd like to see it come up there but for the sort of useful case of the app if Facebook has a use it is you know not important that that happened very quickly right it happened eventually and when we start to see apps that influencers would uh Venture they would probably push back they were no they were definitely pushed back yeah absolutely they want to see it they want that dopamine right now it'll come it'll just come later um when we start to see slower apps more useful apps not not that the car stops I use just like more of them right literally more of them uh I think we'll

start seeing a better need for that and you know maybe there'll be other trade-offs right maybe those transactions will somehow be cheaper and so maybe people will will start to question if they need it right away and maybe I don't know I'm just I'm spitballing I have no idea if this is technically feasible so you know like if you're reading this and I'm saying something wrong just you know let me know like I'm happy to discuss it I could be very much wrong um but it's interesting to see where the future will go and unless we we start questioning about why this is or is not desirable or useful um we shouldn't discard it and and we should keep it open because for all we know the next killer app is going to be one where you know maybe you need some sort of roll up to roll up cross-chain

communication and and there's a good reason for that and you'll either have to deal with this or you'll want to deal with this for some reason um so you know I I look forward to that future that's technically interesting and talenting I'm a technical person I like you know seeing where these challenges are and some of them will probably work out and and most of them probably won't work out and I'll probably look like an idiot if I read this in five years but I I'm happier to look like an idiot now and be proven wrong than to sort of you know not take any risks and and sort of live in my cushy little world I you know I'd rather someone say no you can't do it here's why because then I at least learned something and and hopefully the community learns something right even doing those those

even attempting those hard things and figuring out that you were wrong is still valuable yeah that's still some level of science yeah so it's I the risk of overextending this conversation hopefully this can provide a nice end cap and maybe be the the predecessor for our next conversation sure um like we're starting to get into I think we've laid a nice uh topology for these Roll-Ups and and talked a lot about kind of the differences and why someone may build a certain a roll up a certain way and with that with that now kind of being out on the table we're starting to get into interoper interop and composability across these Roll-Ups so

I wanna kind of like summarize like my understanding of how we've arrived at this point and then and then let you just Venture into your imagination sure so when we first started to get uh alt layer ones and started to even break into Layer Two space we started to get into um cross chain communication and early on this this start this began with okay we have a roll up here don't even call we we can call it Roll-Ups or we can just call it chains we have a chain here we have another chain here and the earliest most primitive forms of cross-chain composability and interop were okay take what happened on chain a pull that off chain and then go put it

back on chain B now go into the layer two space what that looks like in practice is take what happens on Layer Two a pull it back onto the layer one and then put that from the layer one onto layer 2B now once when we get into like roll ups and kind of like even more of these Networks like do we ever get like a like a direct path I I hope so I I don't think I have an answer for you that is definitely yes I've actually talked to some people and they're like oh we can definitely do that and there's this very weird cryptic way and I was never convinced that they knew what they were talking about but you know I I think people are excited to try it and I think there are some cases

where uh you know maybe maybe that will be very easy and maybe that'll be very hard right so if you can imagine an app chain like that has a game for example uh maybe you could migrate the state of that game to another player or two fairly quickly without going down to the layer one uh in some some very limited settings probably centralized settings right now but maybe that's that's one case where it's pretty clear uh because I say game State because game State might not be collateralized right so when it's collateralized it seems like it's almost necessary to go down probably necessary because you sort of want the same consensus on what that asset is and if one key one chain is just minting it without talking to the collateral itself maybe that's a problem but uh for these other applications for

collaterization is is less of an issue or non-existent yeah maybe you can do you know Layer Two to Layer Two direct transactions in some meaningful way I don't know what they look like like I said I've never seen anyone do it successfully I hope there is research going on on that and I think that would be very interesting if you could do that so I hope so at this point is it appropriate to reduce um to the composer to the consensus like it seems easy enough it makes sense to if you have something on base you have another Network on the op stack you can somewhat uh uh like

collate like the the state based on whatever the op stack says like you can you can make sense of the bass and you can make sense of this other network by interpreting like the state of op the op and the optimistic Network um so now I actually don't even know if optimism has its own consensus but at a certain point it what I'm what I'm gathering is that we can almost like reduce back to the the consensus that those roll-up networks share and based on what the consensus of that consensus layer says then we can make sense of like the state of these other roll-ups so I'm gonna say something I could be very

wrong on so you know include the disclaimer if you if you put the text in I think that's what the whole idea of the super chain is supposed to be is the Optimus The fact will share a sequencer and you can sort of Bridge into a single contract and sort of specify which destination change so if they're sharing the sequencer you can say well I'm gonna put my stuff on base I'm gonna put on the um optimism mainnet uh and then you can sort of rely on the sequencer itself to be that that interop between those two l2s directly now if I am right about that I think it's still forthcoming in terms of the technical level I don't I don't think that functionality is implemented yet or I'm just wrong and I've misinterpreted the super chain and it won't be coming because I'm wrong uh but I think that's

that's the idea so that would be also interesting and and for the consensus part right optimism is currently centralized they don't have a decentralized sequencer so um it's pretty if that if that is the assumption that I'm I'm correct about this understanding it that would actually be pretty easy to do because the sequencer could just say a cross chain thing and then it sort of moves the funds but because they had the same Vault on L1 it doesn't need to sort of go back and talk to L1 maybe maybe it needs to write some little data somewhere but I think that's the intent so I you know I can get back to you and confirm that that understanding or someone from Austin Business would be like no that's he's wrong he should not be doing that I'm saying that but that would be that would be cool like if you

could do that type of shared system and then I imagine you can sort of if that assumption that's correct I imagine you could ramp it up to like 11 by using decentralized sequencers that sort of talk to many networks at once because I I like I like the open stack but you know maybe a decentralized system it can do that even better because you can do more by decentralizing like you know maybe your version of the sequencer does something different and so you you can do that for more chains because you can add end points or something or something like this I I'd have to look at the specifics but I think what you're saying is possible and intended but not yet I could be wrong okay um

yeah I'm left with so many more questions than I started with this is probably a good place to leave off sure yeah I mean I I don't know how much time we were I was promised or you were promised but I came in with no expectations really we're at 40 minutes sure we could probably transform this and post the entire thing on Spotify at this point yeah and like make it into a full podcast um but it is it is almost like like reassuring and encouraging to think about like the value of ethereum it really lies in its consensus um because through this consensus we're we're able to make sense of the state of all these other uh networks and Roll-Ups that are built from the ethereum-centric

consent like architecture which is really like the ethereum consensus I'm coming to find out I'm coming to realize absolutely it's an exciting time to build yeah cool