Taiko's Based Contestable Rollups 101 with CEO Daniel Wang

Transcript

welcome back to the rollup High Roller Community as always nothing on the rollup is to be ConEd as Financial advice we are just an Educators in the defi space toss us a quick like give us a share a repost send it to a friend who may want to watch this and finally give us a sub or a follow it means the world to help support credibly neutral educational content in def5 we're here to educate Empower and enrich you and let's jump in [Music]

o [Music] [Applause] [Music] what's going on guys welcome back to the rollup um in our previous episode with Brett we talked about uh base boost and rollup uh Brett is the CEO of Tao um as part of our CK evm for dummy series but specifically with Tao feel like Tao is pioneering a lot of really cool Concepts

so or or at least piggybacking off to some really cool ideas from ethereum researchers and innovating on them today we have Daniel Daniel Wang from Tao it's a morning for him it's a nice evening for us over here and we're gonna let Daniel uh blow our minds about uh these contested contestable rollups so without further Ado thank you for being here Daniel and uh the stage is yours thank you Andy thank you for having me uh yeah so today I'm going to talk about the BCR so here at Tao we have uh two major Milestones the first milestone for our mainnet launch uh which is due soon is the the BCR uh rollup design and after that bre um is

going to launch a test net based on bbr so we can so so now I'm going to talk about BCR a little bit so this is our latest design uh for the midnight so BCR uh is a based contestable rollup so uh I'm going to talk about this uh from a design perspective um so first all we we are a base roll La so uh we are Bas up from day one um although we don't have the name uh but later Justin Drake came up with the name so now everybody talk us about you know base roll La so so Bas roll up allows uh you know Tao not to

have any sequencers at all um and and who is going to uh become the like the block Builder who is going to decide the sequencing of transactions and blocks is eventually the uh the ethereum validators so uh the the other benefit of being a based up is that you know we don't really extract uh value from uh block building we gave up this opportunity so all the existing um you MC Searchers and Builders they can come to tle and then evaluate the MV opportunities and decide whether or not they want to extract value From the Block building uh so all the M infrastructures are totally reusable we

don't have to uh reinvent the views uh base Ro La also means uh block proposing uh or sequencing is permissionless that means if you want to become a builder you know a proposer M Searcher you can just jump in without our permission we don't really have any like white list we don't even of course we don't have a black list so everything is permission l so that's why I think tyo uh wanted to be to be uh type one from day one because we feel like if we we only say we can scale ethereum by 500 times a thousand times and if we are centralized infastructure then you know

the rollup cannot claim that it has helped ethereum to scale right so I think that's our vision you know all the scalability you know is supposed to happen on top of this uh very big um assumption that you know the underlying infrastructure is is decentral and permission list so yeah that's that's about uh you know Taco being a based rollup so even real quick Daniel oh um if you want to finish that thought I have two things to uh sure gon of bring up well first I I don't know if you saw today I I know it's early still there uh succinct just launched their uh prover Network um

which is which is super interesting so I think that we'll start to see uh more um uh of these prover networks kind of come on board similar to like like the shared sequencing networks um where you know where they will try to get involved in further decentralized the approving of of um you know ZK EVMS and ZK based ropes which is which is interesting and and I I think it's cool that um you know proving is is permissionless so like something like a succinct as long as their technology is there they can come and and potentially prove on on Tao um and then on this base part I we had a question from the community that um you know I was looking to get some some

clarity on um and effectively the question is I mean assuming that we have understanding of of Base rollups um the the question with regards to um an L1 transaction needs to be done so that it matches the booster rollup State uh wouldn't that cost just as much as doing it on the L1 and I assume not so um I guess the question is if the state of the boosted rollup L2 has to be the same as the L1 or if an L1 transaction needs to be done so it matches the boosted rollup State um how is the difference of costs

created um okay so these are good questions first of all you know decentralizing the pra network is good I think uh like scroll like other layer tools they are trying trying to decentralize the appr network a decentralized appr network will also work with tyo although when we when we say a Bas roll up we really mean the uh the block proposal or the sequencing of transactions and and you know the blocks those part are decentralized not just the proving part right the proving part is relatively easier because you can always as a centralized the proposer you can always create simple and easy to prove blocks and then your prover will have less problems but if we are you

know since we have a decentralized proposer sorry a base rle a decentralized network allow anyone to propose then there will be all kinds of blocks some blocks are so hard to prove you know we have to solve those problems so uh so it's good to to hear that you know uh the news that you know decentralized like Prov networ is available now um on the other hand the the interaction if if we talk about booster rollup which is I'm not supposed to to to talk here because I'm focusing on the contest rollup but the booster roll up if we do a cross layer one to layer uh to Layer Two transaction and this transaction really has some uh storage

right to the L1 State and it is of course as expensive as a pure L1 transaction maybe it's more expensive uh but usually we don't really expect a lot of rights to the base layer most mostly we want to read from base layer for example if there are oracles if there are like a amm pools like a uh you know random numbers those stuff we can just you know get those data from L1 um and verify the the data all together with other later to transaction that use the data but if you want to keep writing back to L1 of course it's expensive it's supposed to be expensive otherwise wise

people will stay with uh you know transacting on layer one rather than Layer Two not for that answer thanks for the clarity now let's move on to the meat and potatoes of what makes a base contest rollup Sir Daniel quick question before we move on uh you mentioned the decentralized proposer network does that mean that Tao has proposer Builder separation well um I think we should say we inherit that from ethereum we do have you know the separation that's why we can't allow mbv uh participants to to join Tao to you know extract values but we don't

really we didn't invent anything new from uh the the you know block Builder proposer separation you know from that perspective we don't have anything new um the you you can always you can also say that Tao also use a d centralized the sequencer but these side of sequencer are actually the L1 validators so if all validators you know they start to look into L2 block building then you know we are fully decentralized as etherum right makes sense thank you hey Robbie nice to meet to here likewise okay all

right uh okay this is uh yeah it was the it was the why slide y okay cool cool so why we need to contestable roll up um well first of all uh although ZK is the future you know we all believe in ZK proofs uh in theory there are trust versy you know it's Mass so we don't have to rely on you know trust in human but the implementation is really really complicated if you look at the code base uh and and you know if the Engineers tell you there's no back you you can fully trust our implementation it'll be crazy for anyone to say yes I fully trust this code base so it will be

um it will require a lot of time to be battle tested so before that you can always say you know we are trying to make progress maybe the proof is 95% trustworthy but maybe there are like still a small you know percentage uh of proofs that are not right or invalid but it still can be verified on chain so you have to as a protocal designer we have to handle this small odds right so uh no valid proof can be uh trusted uh you know without better tested so that's what I'm what I'm seeing um and then I also mentioned previously that as a base ruup anyone

can propose a block and uh if you are malicious you can say okay I'm going to try my very best to pick um some special transactions and put all those transa together into a block and this block is super hard to prove or even impossible to prove because you can for example the proving the if you read state from um Layer Two you need a mro proof to really make sure that the data you read is uh you having the correct value if you have so many reads in one block L to block and you want to prove all those uh you know data uh read from the state all

correct then you need a lot of like mercur proof right the mercur proof is encoded into the Zer knowledge proof circuits and the circuit may just explode for this very block um is the circuit may be so large that it exceeds the uh the upper limit so that means in Tao you know if we don't handle uh this situation then it's open for for attack so to handle this the tyo protocal introduced the a actor called a guardian the guardian will say okay this proof is this this block is not able to be proved by ZK how do we do that you

know we have to just insert something into the progr say hey the correct block has is this one with don't have any uh proof that's more trustworthy and the ZK is the bad thing we have but it's not provable or in some other situation it's provable but the proof is invalid right so we have to handle those situation uh so that's why we introduced this trust uh trusted actor there and there are a lot of like if else in the code if this happens and if this happens then you know the actor can kick in to do something otherwise the actor is prohibited from doing like all kind of other things

so um so that's one one reason the other reason is of course uh we we talk to app developers and they say okay in the future maybe they want to launch their own action and they say tao seems to be interesting we love the fact that you know Tao is based uh it's U the design is simple but you are a pure ZK rollup and I don't want to have DK proof for each and every block uh that I'm going to have for my for example game right so the cost is really a issue for massive adoption and uh then we ask ourself how can we improve the protocol such that if someone say the tyo stack can be used to

build their own layer tool they can opt out of DK they can opt in like sgx proof even something we called optimistic proof which is just no proof but a you know a a game theory based uh proving system so um after talking to those uh dep developers we figured maybe we want to improve our product design in such a way that we are not going to be the only party who are going to launch something using this technology we also allow other people to launch their layer to all layer series using Tao but config the the layer to our L read differently uh using different proof systems so I

think that's why we you know decided to go with this tier based uh you know contestable rollup design so yeah y I just have uh yeah of course um a couple comparisons in a sense to draw uh for my mental model um you know just earlier today actually we were speaking with a project that's building you know a a virtual machine which uh offers or a virtual machine which is powering um kind of like a hybrid rollup and we've seen this now twice in the last couple days where uh these these rollups uh will offer optimistic um fraud proofs for uh you know 99% of the

transactions but then if there's a fraud triggered it will go and it'll go through a ZK proof to to basically uh either confirm or deny if the fraud is is um you know is accurate or if this transaction is indeed fraudulent and so like similarly and also many differences but like the mental model that I'm drawing is like not all transactions need to be proven in this model um and I'm just gonna uh go push back here on the slide um like on it says not not all chains and apps need ZK for every block I assume this is like as you said like games and like like like areas where there's not a lot of financial like

value at stake um or uh or there's a need for very very fast execution and uh users and developers do not want to wait for proving times is that kind of like the key reasoning there yes I think that's one of the reasons that we we adopt this contestable rollup um the as I said the cost is really the biggest issue for uh massive adoption um and different apps will have different security uh needs right we cannot say you know DK Roll Up is the the best thing to offer for all kinds of apps even with say you know theum is probably the best solution for certain

apps maybe like social media you know things like that they don't really even need a roll up like valum should be just good enough right so the Tao technology this this protocol actually offers a wider range of options for different uh uh layer to or layer series but the tyo roll La itself I mean when we launch our layer to we probably will still try to uh make sure we have validity proof for each and every block and eventually we turn it into a pure DK rollup with all the blocks ZK proven so that's that's the our yeah positioning um you mentioned the other project

and you also mention this word uh you know uh what's that fault or um fraud proof fraud fraud proof yes so in our contest rollup design we really don't have any fraud proof we all have like the only proof we have is validity proof so if you try to prove something you are not trying to prove something is wrong or identifies where uh the the error has happened you always prove you know the this block is supposed to have the correct block hash or or state Roots is are these values you don't prove something wrong you always prove something is the right one is right yeah yeah yeah and then you have this

Guardian as the kind of backup which what is just like is there a selective process for this is is this Guardian like is it to be chosen by Tao or by the the a community or is it a group like what's the thought process with regards to how that works either like on a personal level or on like a smart contract level it's on the smart contract level first of all um then we really haven't decided how to choose the guardians but right now you know eventually we are going to have a dow maybe one year later after midnight we are going to have a dow the Dow will have the security Council the security Council will uh be the owner of all the

smart contracts to decide whether they want to fix things right they want to do a do a fork they want to uh upgrade the smart contract based on the do's decision Etc so the the Security Council is um a large like a multi with a lot of actors from not within Tao but also from the community like token holders um delegates Etc and the secret Council will decide who will be the Guardians the Guardians is a multi uh uh smart contract on chain and each guardi will have to run their own full node and they independently um they don't have to reach a consensus option they

independently correct the the wrong onchain and if the aggregated uh approval onchain is greater than you know like for example two3 of the the Guardians then um the previous proof is over overr U by the guardian uh to make sure the network is uh on the red path on the right had the red state but the guard improving should be really rare I mean it should happen really because we have this Tao token based Bond designs we have the like the validity bond which will be burned if the proof is um repr proven to be uh

incorrect or we have the contestable uh contestation Bond so if you contest a proof and it end up that the the original proof uh is correct and you are wrong then your uh uh contestation bond is burned so all those things will prevent people from like uh you know spamming the network unless they want to burn money right otherwise they yeah so yeah So eventually I have to say eventually after uh the best tier or the highest tier Che proof that we are going to have is is solid is uh better tested and we feel like it's

really bug free then all the Guardians should be gone there we should remove the guarding prover because it is really just for the training wheel so after like a year or two and maybe a few more years I don't know when the DK is really trustworthy then we are going to remove the guard I I think the ZK can can be considered trustworthy Maybe maybe just two or three years later because right now we we try to use DK VM like Risk zero DK VM uh to generate the DK proof uh to avoid writing circuits because writing circuits is uh air prone it's really hard hard to say you know

the the code base is is bug free and really hard to test this uh you know uh TK VM uh what I what we have to do is just write a small program the cast program which is in Rust is is super easy to be to be tested yeah I appreciate that U explanation um whenever I see things uh in the in the ethereum space specifically with rollups um with regards to like backups or fallbacks or trusted I always like to pick and PR to really understand uh the smart contract uh programmability and um and yeah I I I

appreciate your answer and I I also find it quite interesting that um the that that ZK based validity periods need time to be deemed mature um something that yeah it's completely new to me all right that's all my questions here I think I think uh the necessity of having guard improvers on Tao is because we are base roll up so if if we're we were not then we say okay there's the bug we found a uh invalid proof somehow got a verified onchain and the chain is going to be broken let's stop the chain now so we can make the decision we are a centralized sequence we can just stop

chain and you know you know for for maybe five hours again right and then start it uh restart the chain but we we cannot we cannot uh sh down the chain um easily to be honest even if we have the the admin control we cannot really just start down the chin quickly yep yeah so the benefit we have kind of like mention the benefit so first of all from the as engineer you feel like oh the the contestable design is good because we don't have ifls everywhere in the code to handle the the guard improver it's just another tier um so it's good so uh the contestable ruup

also features the the bonds actually we have three kinds of bonds uh I how men the two of them the other one is called L bound so basically if you say I have worked with a prover of chain and this Pro is supposed to submit a proof within like 15 minutes but he didn't then the bond his laves bond is going to be burnt something like that so uh so that's uh part of the design um what else let me see yeah so as I said you know uh it's cheaper because uh most proofs initially will not be DK and DK will only kick in when it's needed uh so uh I

hope I think I think people will challenge us when we launch midnight you know uh other rollup may say hey you're not fully ZK um I used to be a really a ZK maximum list I would say you know I think everything should be ZK uh but uh I become like more realistic after after talking to uh app developers so I now I can tell those developers hey you know Tao is not fully ZK but we are much cheaper than others and um you should just try to deploy your your app here and give it a try and the I think one not so obvious benefit of this

uh contestable roll up is that you can always change your configuration a long the way you don't have to read Genesis you don't have to have a hot Fork you can just use one base layer transaction or layer one transaction to change your chain from a 100% optimistic to 100% DK with one single transaction and then you are full DK roll up so it really doesn't matter whether um it's initial F or not as as long as DK is getting cheaper and you know more trustworthy buree then we will increase the percentage of ZK proes and then eventually we will become fully DK and and remove the uh the guard

improver um yeah I think uh the the downside of course is that contest rule up is more complicated in terms of smart contract design uh than the most straightforward design right the most straightforward design is you know the code should be really simple uh but uh I think the benefit uh overweight the the complexity of having this design and we have run the contestable rollup for it works well enough although we still have like minor issues to to fix but in in general I think we are happy with the the current code so the the first one will be uh BCR for the main

night okay I of does the contestable nature of these rollups add security Beyond uh the regular based rollups because you have uh you have actors that are able to dispute the proofs of certain transactions um yeah you can you can see it that way I think it's not wrong um the one one small thing I haven't mentioned in the slides is that the contestable rollup really applies a something we call a cool down window for

validity proof um the mo the most straightforward DK roll up we just say okay if you have a verifiable uh DK proof it's final right because it's ZK so it's trusted it's final but we say ZK are not fully trusters let's get give it like two hours 3 hours so that if nobody challenge this DK proof then it's final so from that perspective you know we always allow valid to be valid proof to be challenged and overrating with a higher tier proof so I think that did that does add some security to the to the you know asset on on this layer two

but the the guard improver from one from you know someone will argue you know like if the guard improvers are trusted then of course it's it will add security but if the guard improvers collectively are you know they try to do something evil then this whole chain is not trust trusted at all right so it has like two sides you know but I hope that uh even if we we find the guard PRS doing something wrong I think the uh you know the fallback is to use the security cancel to you know stop the chain you

know reor the chain you know make it right but it doesn't uh it still have some some some harm that can be done by the guard improver if they are evil but if we look at all the centralized uh sequencers you know all the layer tools they all has this uh multi as their admin as their like a um upgradeability controller right um so I think it's really just which the the question we can ask ourself is can we make sure this guardan PRS they are trusted they are trustworthy before a um ZK proof or TK prover is more

trustworthy and available in like two years if we have that we can say say Okay Within These two years let's just trust a guard improver because they are stakeholders uh and then we kick them out and then we rely on ZK right so yeah it's I think Andy and Robbie I think the thing we want to tell our developers or token holders is that hey we are sorry you know the ZK is buggy you know the the hiker got the money away we don't have anything but don't blame us it's the DK proof DK isue right if we use risk zero can we say Hey you know that team is the the one that uses blame we don't want to do that I think we should

try our very best to do anything that we can at our disposal to secure user asset uh so that's that's the really important thing that uh you know we we consider yeah I mean I I view it as ZK is the future and it's also at the same time immature in current state in a sense if you want to put it that way I'm sure ZK cryptographers would scoff at me for for for saying these things like a like a un unknowledgeable ple but the end of the day like bringing these middle grounds to progress forward with this like secure um way of of of growth similar to like when protocols put caps

on their tvl it's like a it's like a it's like a slow and steady growth method to kind of bridge the Gap and like test and prod um while keeping the you know the users and and stakeholders and community members and everybody else as secure as possible so onwards and upwards sir yeah true and if someone say you know I I really hope that too doesn't have this security or this guardan PRS they can just Fork our chain Fork our code base remove the Guardians use whatever uh you know Val proof system we are using and launch another chain if users really Buy in that design I think

uh let the user decide right um I I hope ZK can be uh mature enough that we can remove the guard improver very soon but um I I I don't think that will happen within like two years especially given that we are type one dkm so ZK is it's easier um I mean it's easier to to debug to fix C CK e choose if the application is simple like with for example L my previous project you know it's not a evm it's it only empowers a few scenarios like offers a few functionalities so it's much easier to to say hey this ZK proof is trust wory because it's so

simple right but with a programmable you know type one ZK evm is so hard to to say that uh I I don't even trust myself if I say that I hear you yeah I hear you yeah so this is this is so for our tle midnight so initially we'll probably go with optimistic plus sgx PS evm and then guardian and later we are going to add uh the risk zero based uh so this R IC not risk I hope it's really risk zero though yeah right yeah so so we are going to

add this uh more ZK proof there and get rid of the the PS ZK ebm eventually because it's um the code base is too too complicated but I don't want to say that the code is uh is not usable with ZK VM we can still use some existing PS codebase to improve the performance so still a lot of credit to to the pic group but it's we just don't want to to use uh uh the whole solution as it is and the the end game of course is uh dkbm uh no circuits no custom circuits only um Logics in uh Russ language or some other language that are

easy to to write and debug I think this is the last slide for uh for PCR so if you have other other questions you can go ahead yeah um so uh with regards to to um the I think was spefically on this slide um so I know that um we're looking for you know communities like Wang mainnet all these things but the the reality is once main that goes live we have TKO as like the as governance and kind of you know the overall token for Tao we've we've seen the utility here

for vality bonds conation bonds and liveness bonds um and these are all effectively around like kind of like security and livess of just kind of keeping the U you know the the role of going effectively in a decentralized manner and keeping the incentives aligned um you know I'm I'm curious if you kind of thought deeper about what each of these are how you guys kind of came up with these incentives like how you guys are looking to align these I'm just generally curious like now that we understand what a base contestable rollup is is there any other

game theory that you that you're thinking about or incentive alignment structures that you're thinking about when it comes to the community to make sure that that this runs smooth so as as a base R La we need the token uh to facilitate the the the network from day one so once we launch the network the token will be live um initially because we are the only party who has token so um you won't see a lot of like proposers approvers kick in to help us to you know secur the network we are the only party and then we are going to uh give some token away to uh people who have helped us on

testnet um they will use those tokens to uh to to you know help us to to become our proposer approvers after a couple of months if the network is stable and we feel like hey we are not going to throw it away it's the one that really is going to to be to be uh you know maintained for the next coming few years then we'll say okay now let's do air job to to make sure we have a certain uh you know percentage of tokens to all the uh participants of our test Nets or uh to those people who wrote code and you know we use their code or even to

like ethereum contributors uh so we have a long list there so after a couple of months I think uh this percentage of token will be distributed it will be liquid and people want to join tle as proposer approvers they can purchase token there and then uh uh to earn money to to earn the L2 transaction fees tapes on Tao so initially after the first uh I think during the first six months also after midnight Tao will be the utility token uh the Dow will kick in uh after the airdrop after the network is truly stable and we feel like you know uh we

need to make better decisions uh with the community um through the the Dow uh you know proposals uh eventually I think we we try to become you know like how how ethereum operat the the organization we don't want to be a company uh we are still a technically we are still a for-profit company now but we are going to you know um get rid of this company and you know build a foundation with the community and eventually all the decision making will be through the uh the foundation the Dow um so maybe six months or one year later uh Tao token will be the governor's

token with your ti um and it sounds like the token and the well the Dow goes live after mainnet so for mainnet initially these Guardians will be uh at the highest proof tier and if the Dow launches after main net then it's not a method of Dow governance that is electing these Guardians so what is the what's the the process you have in mind for selecting these Guardians um so right now uh it's mostly Tao's Engineers because they need to

maintain their own machine they need to be able to run uh proposers uh sorry not proposer their their guard improver software a full node L1 and L2 full nodes Etc so it's still a bit of technical they need to uh respond to um some notification for example if something is wrong if for example two out of seven Guardians are trying to overwrite a a proof or block then the other guard imprs will be notified and then start to investigate whether there's a software issue or there is truly a this is a legit override something like that so it's

it's uh highly technical for the first uh few months after that it also involves like you have to put your own money there like your own eer private key in the on the machine to to send transactions uh so because of the nature of the you know the uh the guard improver initially I think most people are Engineers uh and then after the Dow is available then um we have two options first the first option is that the do directly select who the guard improver will be uh we have some concern there because uh some people are you know they try to offer Service uh to to to become a guardian but they may not have the

knowledge and time and the other approach or other option is that we propose a set of uh candidates and then let the community to vote like who they they want to uh have as a guardian um yeah so so that's something on my mind but I we we still have not yet decided like how exactly those uh provs will be selected yep that makes sense like I can ask the question for you like you guys if you guys can will you guys be committed to like become a guard improver if we offer some compensation for the hardware you know

uh the transaction fees I mean I you guys can come back to me later off offchain sorry offline so that we chain of course offline so yeah we try to improve the process but we yeah we we need some like suggestions and feedback yeah it's sounds very very cool yeah I mean don't de Pride just just very genuinely curious think um and also similarly um kind of moving from the Guardians to to the pr Network as well like I mentioned uh succinct is going to come on um and they just announced they proving networking you know we have proving tyo blocks here is

is permissionless I'm curious like what like what would a process look like if if I had like a Prov Network um like first like what even is a pro Network like like if if I'm like I'm Andy and I have a Prov Network like what does it look like Hardware software and then how do I come and engage Tao and like what are the economic incentives things so that's that's a really a good question so Tao protocol doesn't enforce a rule between um the approver and the proposer so the pro the guardian sorry the the the block proposer they need to find out they need to find a prover we

call a assigned approver before a block is proposed so let's say if you want to propose a block in Tao you can just work with this Guardian PR Network and figure out a way to send a deal with them either onchain or offchain so if you don't know this uh this um approver you can just uh use a Smart contract to sign a deal and then uh we have like plug in you know we cut hook so you can write a hook for your own proposal uh to validate the deal from between your proposer and the one of the provs in the pr network uh so it's really open up to

you to decide how to interact with the the other party and the and the tle protocol doesn't really care because that's that hook is a smart contract that you can uh customize yeah okay interesting and then the the uh like are you guys going to make it are you guys thinking about like getting like is is is there a difference if if you have like you know 10 people or 10 enenes with provs helping to prove luck or 100 or a thousand like does it is it really something that is like important or like how are you guys thinking about this this like like I understand the separation

which what just uh um explained but what I'm trying to get at is like about like this this like ability for there to be you know dozens and dozens of people helping to prove the Tao Network yeah it would be helpful if there are like some uh redundancy there so if one pro is often there will be backups other people to kick in but I also don't feel like there if there's a redundency there won't be a lot of Hardwares just waiting there to be the backup because you know is wasting their resources so I I feel like um going forward if ZK proof is is the the cost is going down to like $1 like 50 cents

USD you know uh proving blocks won't make a fortune it will make money but it won't like a mining a blockchain like a Bitcoin block blocks right it's not that way because eventually all the provs no matter those you know there are active provs or backups all the provs their cost is compensated by the L2 transaction fees right and if we say we have scaled ethereum by a thousand times in terms of fees then you can divide you can calculate how much fees can be generated from L2 transactions and those fees are the only fees that

will for for the long run you know incentivize all the PRS all together so it won't be a lot of money there uh so I we I think I did a calculation like a few months ago because other people feel like DK proving is the new mining right I say no it's not it's not on the same level because what DK roll up trying to do is to minimize the proving fee So eventually the proving fee should be so small that only like for Tao maybe just 10 mer will be enough to to prove every blocks right so if you count all the layer tools together maybe a Sal mercine is good enough so so you don't really

have to have a lot but you just need to figure out you know quickly if these machines are down you can quickly find out other machines like like the cloud service right so I I yeah I don't think uh uh you know the DK DK mining pool for later to will be such a no I'm not going to say some people yeah well we can draw our own conclusions but it was just really curious just with regards to kind of like the the broader idea behind these prer networks so um I think that's all the questions that I have um and um I I

appreciate you coming on and uh sharing the model behind base contested rups with us um truly something that I I think is just so cool to see these Innovations continue to happen um and you know this idea behind not needing a sequencer at least in the current state amplifies the level of of of ethereum alignment with regards to decentralization um uh immutability um you know and many other properties which we strive to see in in modern-day blockchain so um very excited for maina and to see these things come to light um and yeah I appreciate your time today thank you Andy and uh Robbie

for having me really nice uh you know talking to you guys again yes sir keep building sir thanks yeah thank you guys yeah