L2 Panel | ETHTaipei 2024

hello everyone welcome to l two panel uh I'm Queenie I'm from I'm token one of the earliest uh ethereum wallets in the space so super excited to be here moderating this panel so first of all I'd love to invite every panelist to introduce them themselves briefly and uh tell us uh what the recent progress that excites you the most about your project uh let's start with theal V that's relevant to l2s I guess has to be the Denon hard Fork so as of about a week ago you know we have the blobs and you know this has increased scalability for a bunch of lar 2s and there's a lot of work in the pipeline to keep increasing that scalability further and making sure that L2 is can continue to have very low fees as this ecosystem expands for the next few days or for the next few years so I think that's amazing cool uh so from my end again same excitement about Denon update uh you know if you've been following the crypto Twitter uh it actually did a lot for a whole bunch of networks that you know already uh adopted it the transaction fees they actually went really really down specifically for circuit uh next week we'll be launching some Builder uh incentivization program so that's the next thing that I'm excited hello I'm Carl contributor to the optimism Collective at op labs and I am also extremely excited about 4 for four um I think the other things that I would be excited for are I don't know all of the sweet contributions to the op stack that I'm seeing today so that's exciting Jord valina I'm the technical lead at the polygon zkm and mainly what we are building in polygon is uh well um ZK roll up and the agregation layer to connect many layers uh in a cross chain let's say cross chain compossibility that's very much what the next challenge that we have right now in polygon uh hello everyone so this is Peter from Mees or medis uh depend on how you want to call it uh so the uh most exciting thing in addition to Duncan upgrades which is uh a definitely good news for all l2s out there uh for Mees we uh recently deployed uh a infrastructure or a foundation for decentralized sequencer so we are super excited about it uh and look forward to the next upgrade cool thank you so I think everyone is excited about IP 4844 here so but I I've seen like some of you have adopted 444 but some haven't and some of you like chose different da rather than like Le uh e l one so I'd love to hear your ideas about your overall strategy to make your ler to even more affordable going to start the you on um so 444 right now uh gives a reduction of 16x uh reduction of cost just out of the box and the network is only so the blobs are only 10% so they are using 10% of the capacity there are some rollups still that they need to be uh included but I think here there is even a lot of margin to increase the the Proto L sharding maybe two more blobs in the future and of course we have full L sharding we should give uh I mean kind a good thing so that's that's I mean that's cool because I mean the one of the problems of scaling was that availability and looks like at least we're in the right path uh at this point and that's why this uh um is is is really I mean this step is a really important it's really important one yeah I will plus one that for sure I do think that the underappreciated part of 4 for four is that it introduces a new type of block space for ethereum L1 and that just sets us up for you know progressively improving um the protocol and you know making adding more blobs with you know dank sharding or whatever so super exciting um and we have been anticipating this for a very long time designed the you know Bedrock system around utilizing 444 as efficiently as possible and just posting all of our data directly to blobs call it a day and then kind of to touch on the the like alternative uh data availability providers um the way that we think about it is that we want to make sure that people have the option to choose but then also make sure that we are properly educating folks with the fact that you know uh posting your data to L1 is always going to be more secure than posting it to an alternative data availability provider if your bridge is also on you know on L1 um because you're you're going to rely on L1 security no matter what and so you're introducing an extra uh security property but for a lot of applications that's a great thing and you know the the cost in security Trad off is well worth it so we want to make sure we support everything and uh yeah and scale to to Infinity so uh I will be plus two on that or maybe plus three on that um you know circuit is a zero natural up that uh scales ethereum right as such I strongly believe that we should be posting data on ethereum if you choose an alternate uh data availability layer you essentially you know reducing your security guarantees and you're increasing your trust requirements you are trusting that this data availability layer will be available forever and forever if your goal is to scale ethereum well you know your data should go away only if ethereum itself went away meaning that nobody cares anymore right which is unlikely to happen for alternate data availability Solutions who knows you know they are not you know sustained by the proof of time right uh so from this perspective plus two plus three you know great then is all right so uh I have a different take uh because we did choose a different uh da layer uh for I think it's for a good reason because uh to us uh we think that uh Skilling of course is very important but for this industry to grow adoption is even further even more important uh so we tried a different layer uh of Da I reduce the transaction cost to a a very low level uh and try to incentivize different applications to try new things on L2 uh with different degree of successes but we keep trying because how we think adoption I think is even more important than the Skilling yeah and and I mean maybe V can also talk to this as well but just to kind of plus one a little bit what you were saying there there are going to be limits to the total amount of data availability on L1 and those limits are not going to be terabytes per second more you know I don't know the latest uh you know estimates so we're going to need alternative data availability for you know Mass adoption but yeah can I just add because I I was uh casually chatting with my col if we start to adopt 4844 like can start competing with optimism Arbitron all that and uh where we can start competing in the fee market and bring all the transaction cost up again so it was an interesting interesting thought process yeah just give some numbers um we were already so before pre EIP 4844 more or less were like 10x cheaper that L1 uh with the post rollups just out of the box is 16x and with that a compression we expect with a we full State thata compression we expect a 7x on top of that so that's like three orders of magnitude and we are talking here about transactions that the the current cost is going to be 1 cent of course here of course a market and we need to see applications and so on how this evolves but this is really looks really promising until the point that uh many cdk I mean many many side chains that they were thinking maybe to use other chains or just using validium they are really considering to to just being normal rollups full security you just put the that on chain that's that's really I think that's the the real path maybe not all the applications but I would say that most most of them uh at least they need to consider really hard to really be a normal roll up and forget about security issues agree so maybe the hybrid option is the best yeah exactly con that's Howa we converge yes do you have any comments yeah hi so I'm from ethereum which is a layer to in in denial of Bitcoin which is we because it has proof of work um but yeah I think uh the right way to think about this is basically looking what is the level of security that your application needs and I think there's like different data availability strategies across the Spectrum so one of the things that I've been a fan of uh really trying hard to revive recently is plasma and like basically the whole class of data availability strategies where data is stored off Chain by default but is brought on chain in the event of any kind of dispute or agreement and that's a strategy that can actually get like ensure the safety of a like it's it's extra armor that you can add to offchain solutions that can ensure the safety of a very large number of of assets so I think there's a lot of valuable ideas all across the Spectrum you know basically makes it's all about you know like are you is your L2 that's like heavy L2 for know like large monetary transactions in like people's account data and defy or you're running a game and if you're running a game you know you probably don't need like maximum security for the game necessarily cool yeah so I I think we should all agree that a blob is amazing right so in order to have more adoption of blops I think there are some discussions about like increasing the uh gas limit and also increase the C dat cost so are you in favor of such a proposal yeah I'm so the increase the call data cost I mean for people who don't know I there's an EIP for this I I forget the exact number it's in the 7,000 but one of our researchers Tony vad has been doing a lot of excellent research in like figuring out the consequences of this the basic idea is to put a cost floor on transactions which is about likely to be either 48 or 68 per nonzero byte and the idea of this is that for transactions that have data and that have a large amount of execution they would go over the FL already and so almost all applications would not be affected but a limited number of applications that use a large amount of call data and only call data would see increases in costs and the goal of this is basically that the core developers are worried about increasing the gas limit or increasing the block counts because the theoretical maximum size of a block is pretty large it's a PO post compression about 1.8 megabytes and what for what this e does is it basically decreases that size to the point where it's somewhere between um know 500 like 750 kilobytes without increasing the cost of almost any application the only thing that really gets affected are like there's Starks that might go up in cost by about 40% and then there's like potentially Merle Branch verification that might go up a bit right but but even for Starks like yeah this would get more compensated for many times over because if you can bounce the execution cost or if you can bounce the execution block size that opens room to increase the maximum counts of flabs that can go into a block which then decreases the cost of inside um so that is that's one of the approaches that we're looking at for short-term scalability and that would make it much safer to both increase the blob accounts Target and also potentially increase the amount of execution gas in a block as well which is definitely something that there's ongoing discussion over whether that's a good idea to do so sound exciting any comments yeah I have one I think that uh you know this approach it's actually really important I think that we are right now or you know pre8 4844 we were using call data for something that call data was not meant for right okay I love doing that yes Carl so uh you know once now we have blobs well let's use those blobs and you know this repricing it in some sense is a deterrent for using the call data for the right purpose right which is good there is one technical thing that I think is important to mention is that blobs are fixed sites and they also have a I mean they have a field Elements which have a bite that is difficult to use I mean you need to use bit a bit and you know things uh but um and that means that for maybe for applic so if you have an application that's data intensive you probably already have an incentive and you already go there and but if you are if you are using part of that maybe paying the full blob uh is not uh the best idea so I would be a little bit more prudent before increasing the the price of the C battle yeah that makes sense maybe blob aggregation anyway but yes no some protocols I mean I think that here we need to create some protocols for reusing blobs for different applications a lot of space here to to work yeah exactly this is great great collaboration um very excited we need more blobs give me all the blobs um and yes we have been absolutely abusing call data it was multiple years ago we were doing the math we're like wait if we just submit everything through call data will get most of the security benefits of ethereum but not pay as much money and yes that time is over it's time to uh use the real thing the Real Deal use some blobs nice we should be confident that we will have much much lower fee on their too in the near future so yeah so uh I have a long list of questions so second topic I want to talk about L to fragmentation because fragmentation is there which actually hinders the network effect of the whole ethereum ecosystem so I think there are some uh recent proposals like base sequencing like shared sequencing aggregation layer like uh aiming to solve or partly solve this problem so I'd love to hear from you guys about your opinions on uh shared sequencing and based sequency or even more recently based Shar sequencing so what are the trade-offs and the what's the direction you're heading who want to go first if you I can start um the idea of the liquidity is about friction I mean if you are in the same chain you have let's say no friction you are already in the same chain if you are in the same aggregation layer in the same constellation you will have a bit of friction but probably you're going to be fast if you want to cross uh I mean cross constellation communication you Pro you will need you probably need to go through uh layer one and I mean layer one is not the fastest Network I mean the finality time there is uh like to to I mean to EPO it takes some time if you want to be S but you can do it I mean it's maybe about minutes uh we are talking and there is even another is like cross L1 and here is a lot of work in the bridging in back to pack so this like probably going to be even higher friction uh for this all this is are starting I mean all this is very much powered by the ZK Technologies is what's enabling and we will see what's this friction at the end of the systems yeah and and by the way definitely shout out to polygon for being very inclusive and having uh you know in you know including all of the different chains and I strongly agree we do need need to include them the one thing that I think uh maybe uh you know to also mention is having a kind of like Universal way to address all of the different chains I really like the idea being able to send and receive information from and to all of these chains of course proofs are a very important part of that but also having a kind of you know thankfully we have chain ID infrastructure but it will be nice to like build more tools that kind of like abstract away the underlying systems that you're working on and you can now address all of the different chains and you know build applications that kind of bridge bridge multiple ecosystems um so we're you know we'll slowly get there but to set expectations I think it's going to be a brutal a brutal at least six to 12 months uh in terms of fragmentation but we will get there we will get there so uh I might have a little bit different opinion here to be a little bit controversial so uh look where we started we started with single nonfragmented chain and then we LED competing you know as technologists as innovators we you know love doing that and we broke it and we fragmented it so let's not create a super chain shirt sequencing whatever we want to call it yeah this is going to exist for a little bit and then we're going to break it again because somebody else is going to come and say hey I have a better super chain you know I have some Innovation that I want to put out there so let me start comp competing so you know unfortunately the world is not black and white right it's always going to be some sort of a sine wave sometimes we will be closer to you know having homogeneous environment sometimes we will be fragmented uh I think that this General problem is a problem but I think that it needs to be you know addressed from the interoperability and cross gen communication perspective you know yeah I think from my view like shared sequencing is like honestly bottom priority as far as improving Crossen interoperability is concerned I think I mean to name a few higher priorities one is just uh like making getting to the point where we have more and better cross chain wallets like just like wallets that more like make the cross chain interface look much more natural and like I've been using Rabbi and Rabbi does like a great hu on that front and I know other wallets are improving the uux on that side as well um another one is improving cross chain um token swap infrastructure and I think particularly we want to see a yeah a cross chain at least one cross chain token swap instance token swap Protocol no liquidity provider assisted that is like actually neutral and like non-dependent on centralized actors to at least the same extent that Unis swap is right like you should be able to build a cross application and be confident it'll still work in 2027 and you know Unis swap does that you know do some of these you know like new fangled you know like janky cross Bridges you know do not necessarily but you know we need to get there um a third is uh like coming I think agreeing on a Clear Vision of how um accounts cross chain are going to work and one of the key questions I think what the ecosystem just needs to conclusively settle on is I you know are we even trying to give people the same address across chain or are we not and like for the people I've talked to I think I increasingly the answer is to lean toward not doing it the basic reason being that like will just always be chains that have different enough technology that you will not be able to get the an account at the same address that has the same functionality but then if people have different address on different chains then you know like we might us while adopt you know for example ERC 3770 and put the chain you know like name into the address so that you don't accidentally send people coins on the wrong chain and like if we have that then a lot of daps also need to upgrade to not assume any people have the same address everywhere um which is I think like also improving and once we have that then things will get better um another example of uh just like forms of interoperability that are easier than we think right is like a lot of people want you know like chain abstraction like cross chain operations but like 90% of the time people want something simple right so like for example use case I had I am on like you know mainnet and optimism in arbitrum and I needed to send someone someone coins they were accept coins on base I should be able to S them coins on base with a single transaction right now that reques two transactions step one the bridge step two you know extending over the coins well like real realistically the bridge should just include as a you know mandatory option which address you're sending to and then if you do that then you're supporting cross chain um you know accounts that have different address on different chains and you're also supporting s other people so now you can already send someone coins to an address on a different chain and this is something that doesn't even require a separate app this something that should be integrated in basic wallets right then going further um you know within the context of smart contract accounts key changes key recoveries you know account upating like do you do that by sending a message on every L2 simultaneously or do we do the P store rollup thing or do we first do the second and then explore or or the first and explore moving to the second over time and that's uh something that a lot of people wants to do right so these are all things that like make things easier from the perspective of just like real ux problems that a lot of people have and like basically none of them actually depend on shared sequencing right so I think yeah the way I I think about this is like focus on the stuff that's like lower up up the risk ladder and higher up the there's a big pressing need for it ladder and then you just like move up over time right and I think it's a importance to like there is a lwh hanging fruit and we have to really improve on the low hanging fruit where the solution to that low hanging fruit does require like less fancy Tech and less hundred million doll startups but you know like more social coordination and like actually solve that problem first and then you know keep going I think that this is a you know really good take uh when I look at these problems I always uh look at them from the perspective of my grandma the synonym of my you know Grandma is a lazy user right for me as a lazy user or my grandma the goal is to use some block application you know be whatever Unis Swap and such right but the goal for the lazy user is to not worry about whether this Unis swap is residing on that chain or that chain or that chain and how the you know funds uh float over right the lazy user wants to have funds on blockchain and you know use this funds on blockchain with the respective application and be abstracted away from everything that happens under the hood so all the bridging everything else and we I I do think that even though right now there are a lot of problems we are entering a kind of new paradigm or a paradigm shift of this kind of chain abstraction and you know higher level uh way of interacting with multiple different chains and so even though while the user experience is kind of bad you know it will be brutal however it is also an amazing opportunity for incremental progress across the stack so that there's tons of things that we can do right now to improve this experience unify The Experience across multiple chains and you know we'll wake up in a little bit of time with all this development progress and you know our grandmothers will finally be on chain no no no you're supposed to like put like insided MPC so that like the private wait we're not talking about mind uploading you part of my grand letter yeah I want to go on and I'm I'm NBC you know I'm kind of down anyway yeah we're talking about grandmothers but I remember the day I I'm talking that remember the day that my grandmother start using teletext I don't know if you know but ugly interface probably the worst thing of that just because he wanted to see the lottery the need but I mean and he learned to to all these bottoms I don't even know how that work and and and they learn to to go to the number the PCH and and jump and navigate and and was my grandmother the need I mean here is we we are thinking but the need and this is very much for early adoption but the need is more stronger that the that the that the easiness if you need something you are gonna you are going to get it so are you saying that the key to getting my grandma on chain is tell her about I tell you that if my Grandma has a need for going if your Grandma knows will if your Grandma knows she can make $100 a day on blockchain she oh God you cannot imagine what grandmas can do all right you're right I've been bearish Grandma I need to all right I'm not grandmother but I'm also you know lazy yeah by the way rabi is amazing but I'm toen is also working on it okay so let me let me know when you have a wallet that has a just immediately you know has the same accounts on a mobile and a desktop and Sports I you know like smart contract and recovery and all that stuff I think even even rabie has not gotten to desktop and mobile support or like they have both but you know not linked but no this is the reason why this is actually important is because uh especially like there in the context of like a lot of these apps trying to appeal to mainstream users um like including farcaster including a bunch of them like some of them are starting to do like key signing in and like Key Management in the application which is like a bit bad because the more that happens the less the wallet can actually protect the user security but a lot of the time the reason why they're doing it is because it's like a standard expected property of every application that it works on desktop and mobile and you have the same account in the same content um but like the only way to not require individual applications to all do that themselves is for you know what like w it's to be the thing that that is where the user is and like is able you know do the the authentication piece everywhere so the closer we get to that also the better nice we're on it yeah just uh going back to the sh sequencers just I want to express that I'm a little bit skeptical on the sh sequencers um I mean synchronous things in general they are hard to work and when there is two parties that they can do a lot of things to do I mean it's a good thing to have I agree I think it's a good thing but uh I would I mean I would priorize first uh uh asynchronous as synchronous calls for example or a synchronous uh messaging system I think this is the first it's a I think it's an easier step I remember I mean I remember vitalic at the 2015 or 2016 talking about a synchronous calls in some some calls I think maybe it's time to recover all those slides and and and start working on on on this on that part I don't know what you think about that I mean I definitely agree it's much easier to build a sing syn synchronous system on top of an asynchronous uh compute layer um but you know we and yeah we'll we'll have it it's just one thing after another right shared sequencing full synchrony it's G to come just takes more work how about polygons aggregation layer just work together with super chair like you guys can you know the two ecosystem can talk to each other first oh yeah yeah yeah I mean here is via layer one for sure here for example one of the things that we are start working I mean we have a lot of work to do this is in something in the Q but it's um what we call um massive uh brid I mean M massive deposits and massive exits so the IDE is that you can uh between rollups with a single mean with one transaction every block or every 10 blocks whatever it's just doing all the consolidation uh between uh two rollups or even can be a smart contract with Ops can be and so here at the end idea is having a special chain where you can it's some kind of a proxy chain you can set everything there and then you can do a massive exit to another chain and and the same the same way for importing so there are interesting things to do for this uh cross chain uh I mean cross cross constellation or cross aggregation layer uh protocols but we are very early stage I would say totally totally and I think here's something you know we might say oh they should work together they should you know collaborate we are working together and we are collaborating and that is the beauty of Open Source polygon has been pushing forward the ZK space for years making tons of advancements which we have benefited from you know ZK proving the op stack and you know various other things there's a huge amount of like cross-pollination in the industry and so you know we don't need flashy headlines necessarily to collaborate and be a part of the you know broader thing and important I mean I think that having many aggregation layers in in in in ethereum uh at least at the current stage is very important to learn from one each other I mean we if if there is only one aggregation layer you can do things wrong I mean you don't have uh any comp competition but it's competition in the in the good way I mean it's like you need to okay message pass passing messages we need to do with a Mer tree we need to do with s we can do it optimistically there are many ways to do different things so having different teams working in different uh I mean different approach this is I mean this is a way to speed up a lot the the Improvement and I don't know in five years 10 years well we will see what we have there but let's think for for for the next five years yes okay sounds good any other comments on this topic yeah they're sitting Bes beside each other you know they're each working on some way of aggregating the fragmentation right and they are going to compete so this is what I'm talking about it's always going to be fragmented more or less but it's going to be fragmented but we can build some we can get the low hanging fruits right yeah but I mean for in the the the the rollups can be essentially very different I mean they can have different base tokens some of them they can have centralite sequencers centralized sequencers some of them they have different data availability uh strategies and I mean the world the world is diverse so allowing uh different uh chains that are some sort autonomous and they have their own thing but at the same time they are connected uh with the world this is what allow uh I mean this is the real scaling uh what happened the real weight of scaling is a more parall scaling talk about it's like a processor we who have many cores at some point there is a limit that the core cannot go faster okay so a blockchain maybe it's going to be a limit there but if you can have as many cores as you want or many processors connected one each other this is what actually gives you this parallel scaling and well I think this is necessary so this of course this has the cost of fragmentation but we have some solutions to not to remove but maybe but probably to improve a lot this uh this friction between uh chains but the only thing I want to add is uh uh this open source culture in this industry and uh because op op open source also open source also open source we all benefit other by by like collaborating IND direct directly through the source codes and and I know I'm not going to name but there's some projects deci to go close I mean if you are in the space if you are in the I mean if you in the space we are building trustless syst how are you going to build the TR systems if it's not open it's really a community effort it's industrywide effort to to push this technology going forward so defitely call to make sure that hey let open source yes let's go that's the true interoperability license okay I got to finish my question L you know so I want to talk you know touch a quick one about uh upgradeability I think it's question related to you know uh safety liveness and decentralization of the system but uh that deserves more disc discussion because there's no perfect solution yet but op team I think you guys introduced the decentralized security ccil to govern upgrades but are there any more like potential solutions that you can share with us anyone stage to or bust okay so you know of course we are going to make the upgrade process as decentralized and secure as possible Security Council is a big part of that having a citizen's house is also a really big part of that you know building out identity systems so it's not just token holders that are you know driving upgrades all of those things are super important for decentralized governance without question however it is incredibly important arguably the most important thing to make it so that even in the case of a governance failure there are recourse for users to is you know exit the system in order to do that it must be the case that cryptographically we are preserving the right to exit for users how do we do that well we do it by initiating upgrades on chain with a delay and that delay must be long enough that so that users can exit the chain how do we do that that is easier said than done the best way that we know how is with a multi- prooof multic client ecosystem this is how ethereum is secured this is how layer twos need to be secured as well and so you know we're building out uh multiple clients multiple proof systems we're actually you know the the thing that we released recently was uh a you know fallproof system written in go we have a parallel implementation that's work in progress in Rust and you know various different uh zkps and all these things so we'll need many proofs again it's an open source Community uh uh work you know initiative to achieve stage two for the ecosystem and yeah prevent bad upgrades yeah I think it's important to think about the concept of a security Council and the concept of governance protocol upgrades separately because these are two different tools that do two very different things Security Council is a small group of people that with a high percentage threshold to perform any action that is there to do emergency actions specifically in the case of bugs so in a world where there were no bugs um which who knows AI formal verification technology May well get us there in a couple of decades there will be no need for security councils and even stage two basically says that a security Council only has any power at all in the specific case where the software provably disagrees with itself including you know disagreements between different groving systems governance on the other hand is there to do um upgrades and with governance one of the uh most important checks and imbalances is to have a delay that is long enough for not just for like Apple users to be able to outrun a or for the for the governance to not be able to outrun the fraud proof system but also for users and applications they'll actually be able to make an orderly exit that is you know seamless for them right and so and that's important both to minimize costs in the case of an actual governance failure and also to remove the incentive for governance to attempt to do any crazy things in the first place right otherwise you know you might imagine a governance basically saying you know like hey know like you know we're the citizens give us an extra 5% of the op or else um you know we're going to do an upgrade to to do that for you and you know sure you can uh you know users can escape but like are you guys really going to try to like do all the work of coordinating moving somewhere else right and so that in order to prevent those kinds of situations you know the the cost of collective exit for users and applications does needs to be minimized and then in addition to that you know governance for upgrading later TOS has the same kinds of challenges as governance in general um so you know there is just making sure that the governance actually is high quality making sure that it's taking into account some like diverse perspectives the having both house and the citizens house as optimism does I think is an absolutely great experiment I've been saying for a long time we need to really move Beyond this idea of token based governance as the only Paradigm um there's you know also challenges around the the communication structure around the governance like what kinds of forums exist what are the places people talk about things that I think are as important as the final governance gadu itself so you know these are things that are all important to think about and keep improving yeah and I will just say this reminded me and all I'm only saying it's a little bit Savage I'm only saying this because the room seems so like Pro open source here I feel so safe um but part of being able to exit the system is being able to migrate to a new place that you can go and so that's why you know we have Fork ability in the optimism Constitution to make sure that all of the code is open source and forkable a real bad thing that might that could potentially happen is you get to stage two you know you you technically can like exit the system but you can't actually redeploy your application or for the chain or whatever and you're stuck and so what are you really going to do you you're you're you're kind of you're kind yeah I think we need I think we needed clients developed by independent teams to perform what things like exting yes yes exactly agreed how do you like the uh L1 soft idea what is the L1 soft Fork idea K can you help me to explain it no the V does not like the L1 software sorry this is just a random spicy topic of uh basically giving L1 the ability through a soft Fork to essentially withdraw the tokens from an L2 bridge and so this is essentially like giving almost a a power that is too tempting to the L1 validators because it ca may you know do things like cause consensus and stability but anyway V you you know you I mean I think the challenge is basically that L1 governance is like historically this you know very intentionally minimalistic thing and it performs best on know very very technical topics and you know it tries to touch on topics that affect I know people's Financial adds know only very occasionally and uh like basically saying that the role of ethereum governance is to be like emergency backup police of you know what governance mistakes that happen in the ethereum ecosystem like really expands its capability in a way that also increases a lot of outside actors incentives to try to manipulate and attack it so you know this is the reason why I'm wary of such things okay thank you guys we're a little bit overtime so finally I I haven't finished my list okay thank you all for sharing and thank you all for joining us I hope you all enjoyed the discussion thank you thank you [Applause]