Taiko Based Boosted Rollups Explainer

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 what is a based rollup what is a based

boosted rollup and why should you care I don't know I think that's what we're gonna find out today I think today we're going to find those questions out yeah um and then we're gonna go through Tao's um upgrade of Base boosted robes and talk about their kind of in Innovation space there I remember I watched a presentation de about that I believe you watched one about Justin Drake yeah um so Jus Drake put out an article about based boosted rollups um so we're going to go through that article I thought he just did base rups I think base boosted rups is just Tao Tao is the only one doing based boosted rollups yeah the IDE do mind base rollups is you remove the

rollup sequencer and you just use the L1 as the use L1 block proposer as the sequencer right that's what I got from Justin Drake okay it's based rollup commits the transactions directly to the L1 they don't send one as a sequenc right there's no sequencer it's just it's just go it just goes directly to the L1 yeah so there probably no possibility of me or any kind of like sequencing optimization or yeah there's a there's definitely significant less risk of toxic me amongst other uh significant improvements m is funny because like it it is it is toxic in a certain sense of

users are the ones getting screwed yeah but kind of finds more value or like more efficient ordering me like more efficient ordering of transactions so the validators are getting you know it's not toxic to them they're extracting the value but we could pass that value back on to the users yeah it's certainly a debate yeah yeah whether it's toxic or not but there's a couple other key key uh points for for why um base rollups are better than centralized sequencer rollups okay the other one is this idea of uh and I'm I'm trying to pull it up for my notes in the presentation but the other one uh that I know off to my head is idea of uh the there's less risk to lose the network effects of your

ecosystem so um if the arbitrium sequencer goes down uh and you know for a long period of time and all that value on arbitrum is gets misaligned with the state of ethereum and it's basically a catastrophic situation where like that value becomes like like almost like worth nothing um you lose all the value on arbitr yeah or on any rollup you you lose all the network effects that you built already oh for sure um and so and so by using ethereum uh or a cement layer specifically for your sequencing you don't risk losing the network effects um that that you've already built okay so let's uh we're gonna we're

going to share the Articles uh let's pull it up while I get this information because so just interesting things here let's get this up here okay so you should see uh the Bas this is what I needed goads um so so here's the base rollup uh article by Justin Drake and this is before we get into what boosted Roll-Ups are so first we'll take the the base case just base rollups uh superpowers from L1 sequence like we were saying uh kind of like our our primary understanding here was that

transactions I is transactions the right word like essentially like uh uh whatever you call them intentions like there's basically the the transaction is submitted directly to the L1 and we're calling that based rollups because they're no they're not using a sequencer uh to submit these transactions but it is a rollup so they're still batching transactions together and then committing those to the L1 for changing the the state of the rollup yeah it's he says when the sequencing is driven by the base L1 um it's basically when that block proposed of the L1 May in collaboration with L1 Searchers right um the INB Builders right permission

permissionless include that rule that next rule of block as part of the L1 block so basically you have you have the L1 proposers um sequencing um or sorry A B is one with the next album proposal in collaboration made permissionless to include that next worldle block as part of the L1 block so basically you don't need the sequencer to uh B to to to put the transaction on L1 that's is done by the L1 proposer but the sequenc of the L2 signer would still sequence of transactions though like order them or no no I just done by all by the proposer right so all so there's basically no sequence or interaction on the L2 basis so basically like if you arbitrum optimism and whatever these aren't like

these are techn technological uh developments But ultimately it's a they're all extended to the them as like this like offchain trust layer basically a rle was effectively like you're trusting arbitrum or optimism or inser your roll up here um that they're going to do right by your transactions and properly bash and properly secure their nodes and have good Hardware um and so basically we're trying to get away from that for for these reasons or at least Justin Dres for posing this because because um centralized sequencers are kind of a problem centralized sequencers are kind of a problem um and Justin Drake here describes the Simplicity of

based rollups um as uh maximally simple and therefore inherits a lot of the not a lot of the L1 properties including liveness and decentralization so this this concept of maximally simp um I think I think is pretty important and he described Simplicity towards the bottom here uh because it's even simpler than centralized sequencing you're right it requires no sequenc or signature verification no escape hatch I thought that was an interesting part was no escape hatch when I was listening to this presentation what an escape hatches I believe s similarly to how um similarly to how data availability layers have it

in shape hatch for if their data isn't successfully uh if the nodes in their you know uh validator set don't successfully uh come to consensus and plac the block uh or some sort of um data proving to to the L1 then that then the L1 can basically go back and fall back back to the sequencer uh who batched those transactions and get the get the data directly onto the L1 kind of basically saying okay if then if there is no data from this data layer that's accurate and proven and verifiable then we the the code basically goes back to the uh the sequence in the batch transactions and stores the data on L1 so I think similarly what this does in skape Hatch

is basically um in essentialized sequencer model there's a roll back right where in insert roll up company here or project here should it something catastrophic happen you know like like the North Korea hackers hack all the chain they can rule it back as an es Gap patch I believe or or they can uh yeah there's some sort of mechanism where the the uh the um escape hatch is like a second level of trust yeah you can like circumvent kind of like the properties that are normally taking place on the everyday transaction float through the ruup you can you can kind of go to this escape hatch you can eject yourself from the this ongoing system and I noticed

like from what you're talking about it has a lot of properties where you could you have a little bit more customizability yeah Justin Drake is I don't know from this it sounds like he's a little bit iffy about Escape hatches um notice that non- based Roll-Ups with Escape hatches suffer degraded liveness then he goes into like some of the properties about degraded liveness um rollups with Escape hatches are liable to toxic meev yeah exactly but the short-term sequencers censorship during the timeout period yeah but without getting too deep into Escape Hatchet maybe we'll save that for another video um we'll keep going on based uh rollups here and one of the points that I found uh pretty interesting uh was the uh based rollup

naturally reuses uh Searcher Builder proposer infrastructure and now you'll notice uh there's no sequencer as part of that infrastructure so it is the Searcher I believe that is picking up transactions when users submit them and then it's Builders who are building the the batch these Builders are are they're not sequencing the batch they're just bashing them all together as a sequencer excuse me as a Searchers find them um and then the proposer proposes that block onto the L1 so yeah and there's there's the me involved which is the incentive to include these blocks from the rollups in in the transactions because um they basically will will will make money by doing so um and you can

extract somehow you can extract this will up somehow through whatever the transactions are on that rollup or just basically on like the cost of of of in in of including it we talk a lot about like value acrel and uh you know we talked about like in the modular ecosystem uh where value acrs at the at the VM layer or the da layer or the interop layer and it it would be particularly cool to dive into where value acrs in this infrastructure whether it acrs at the Searcher at the Builder the proposer the sequencer level like where does value acre depending on some of these different INF these different architextures it'll it'll kind

of like tweak and adjust the flow of that value um so the concept that I keep getting here is that there's more directness to the L1 and that has that has tradeoffs you don't get the escape hatch for the roll back and the fail safe uh opportunity but you also get the liveness of the L1 and the L1 is much more likely to stay live because it has the consensus built into it it's more likely to stay live than something like a rollup that could that could go down essentially um yeah okay yeah and I think we kind of glossed over that livess part of the top but more or less

I mean you get we we mentioned the network effects we mentioned mu and we mentioned kind of the Escape hat so those are important and then of of course like the gas penalty right so going down I think the basically Justin Drake's pretty ethereum aligned um he thinks that L1 economic alignment is very important for base robs perhaps he fears a move away from ethereum as a settlement layer um or or perhaps not so well I think he he does he does keep this relatively General to L1 he doesn't say this is just ethereum so I mean one could this is certainly this is three so stock oh it's certainly written for ethereum but one could imagine a based rollup that settles on

Solana uh perhaps not Solana maybe but uh yeah perhaps Solana but more like a like another evm chain you like an avalanche or a or what I I say if they do paraly like another evm chain could probably apply this but non evm modic chains is so for conversation um but yeah I mean going on to the next two is gas and L and economic alignment so basically see enjoy zero gas over that you don't even verify signatures um basically you can do no no token very low gas or no gas payments for uh settling blocks pretty bullish um and uh the economic alignment basically the

value flows to the Bas L1 that's what you're getting at what do the value flow this this model takes away value in my opinion from like arbitrum and optimism and whoever else not directly necessarily but it takes away value from rollups with tokens and we're trying to go that model and it brings all the value back to ethereum um what you interesting because uh there's different models for how people are doing this and then sovereignty super interesting I kind of it's kind of hard to understand this one in a sense but um they retain the option for sovereignty by delegating sign to the L1 like to me it's like well I mean I'm not sure it I mean sovereignty is basically

saying that it has the option to charge base fees um and is that essentially saying that you could have a based roll up that charges gas in its base in its in its own token doesn't have to charge e gas even though it settles to the directly to the L1 doesn't have to use eth as as gas maybe yeah and can use that's interesting all right and it's opted so that's the other thing it's like you can still choose right that's little Sovereign but to me like Sovereign is like you're not connected to etheral you're like like let like Cosmos app chains is what I think of when I think of sovereign um kind of like we were discussing the other day so it's it's different folks within the etherum Q that's this is what Justin is suggesting

Sovereign all right disadvantages no incomeon um sequencers make a ton of money right now off transactions uh from ordering and batching um just in fees as well on top of that so um Justin actually suggest something really interesting here the reason is that ruup is that the ruup landscape is plausibly winner take most and the winning ruup May Leverage The improve security decentralization Simplicity and alignment of Base rups to achieve dominance and ultimately maximize Revenue basically saying the user experience is like what wins and if you can use if you can use a based rollup uh that doesn't charge meev income you're going to deliver a lower cost more practical pragmatic solution

to the end user which in theory would mean that the user sticks around on your based rollup rather than going to another one that charges toxic M and sequences uh the user funds but and the reality is uh the disadvantage here is sure that's the like the idealistic outcome that you aim for but unless you're Mr big bags well you're kind of in trouble you gotta fund it you gotta fund it somehow that's what he saying so if you're not getting sequence of Revenue how are you going to fund all the development all the grants programs that it takes to build an ecosystem only that he calls me like SE Sur say me yeah it's funny it's like yeah well me is no longer minor extra Val yeah me

stopped being minor extractable Valley after the merge when when it miners were taken out it want to proof a state exactly now it's m all right let's not get too cut up on this and um we can kind of finalize our understanding of Base rollups constraint sequencing is interesting um basically you don't have as much flexibility makes sense um what pre-confirmation are very easy with sexualized signature so rollups have this situation from here where basically there's there's a the goal is as quick to time toi time to finality as possible uh and Tao is really focused on this which is super interesting basically uh with uh with centralized sequencers you can get pre-confirmation

in like less than a 100 milliseconds but you have to wait on the blocks of that rollup to to confirm and then you also have to wait I I think for the ethereum block which you are putting your transaction in to confirm as well which takes usually like 12 minutes and then the different times for um you know how many blocks arbiture optimism or whoever needs to uh needs to approve so basically you click a button on your rollup you get the confirm transaction but that's just a fast centralized sequence pre-confirmation that that you're getting that rollup has a short period of time maybe a minute to ad me it's don't don't quote me on that where that is that confirmed on the rollup and

then you that that that transaction gets centralized centralized sequence and Bash to ethereum which then it needs to be included in this confirmed ethereum block which also takes time so that's a disadvantage where he's saying that fast confirmation of L1 sequencing is an open problem right well well it's he's also saying that you don't get fast pre-confirmation in based rollups because they don't have the centralized sequencer to provide that fast pre-confirmation which is a problem because the reason why to ropes because they're fast cheap first fast second I'm very close I mean you got to go centralized if you want fast and cheap right there is the TR along yeah providing arbitr style first come first serf sequencing is unclear with L1 sequency um first come first served is

like basically the transaction ordering mechanism that the arbiture style sequencer uses yeah it's tough because like how you know you're relying on proposers and Builders and things what what I would like to uh understand though I mean these are all like the the naming is what it is but um scroll down I'd like to hear what like see what some of these people said all right so Bas robs capture their own M could enshrine an option mechanism inside the L1 contract okay so basically it's like yeah I mean you kind of have this relayer network in that case uh and uh I believe me I don't know if it's me boost but there's kind of like this auction mechanism that allows s uh that

allows um uh proposers right to bid on the blocks you exactly and and this guy's saying in that second paragraph even that you could have a mechanism with with a small number of of batches can be canceled or replaced by grabbing the contract with a higher payout so kind of similat how on ethereum after e559 there was that that the tip yeah it was like a minor tip first and then it was the merge then it was a right a a validator and the tip is to bribe someone whether it's the Searcher to pick up your transaction right exactly so he's saying that he's saying that um let me screw it down a little bit more yeah let's see what we got

rollit roll kit cool I think one roll up per L1 block cannot be guaranteed by proser Builder separation requires to modify the L1 consensus this is something that we kind of glossed over which was that uh vitalic had talked a little bit about this in 2021 yeah total Anarchy multiple Ro blocks submitted at the same time causing wasting he now proposing bu allows for tightly regimented base sequence but here's here's the thing and I know this is one rollup block per one L1 BL it is now understood that proposer Builder separation allows for tightly regimented based sequen at most one roll up block per L1 block oh that's

interesting yeah because if you could submit more multiple rollup blocks in to the L1 in theory you could have Atomic composability as long as those those blocks are able to kind of like match orders with one another think this is the opportunity with shared sequencing but that's a tangent for another time um the the multiple blocks submitted at the same time would pose an issue to the L1 because maybe not all transactions that get picked up in that are going to get picked up in that one L1 block so one of the Roll blocks is going to get is going to get thrown out canceled have to get resubmitted longer transaction time and wasted gas yeah it's a problem but but now proposer Builder separation allows

for tightly regimented base wasted Z all right so now we've got a pretty fair understanding of Bas uh I I I I can't be bothered to dive into snarks um cool okay there's hu given his thought two cents I guess get quite intuitive solution nice yeah this Anarchy part and of course that's the part where vitalic ways in that's where things get particularly hairy so we'll uh maybe we'll dive into that uh a little bit de on it on another time all right so uh let's dive into Tao now now we have a really good understanding of what a Bas is uh it's

equally crucial to understand uh what a tao based boo is so yeah yeah so we so we clarified and just to put it simply the based part of a rollup is going to uh is going to mean that the uh transactions are submitted directly to the L1 there's no sequencer as part of the proposer Builder uh Searcher infrastructure no sequencer doesn't appear anywhere in there uh so batches are submitted directly to the L1 to be included in blocks um okay so that's a base rollup now let's learn about how we can take that based rollup and boost it what is a based booster rollup new major milestone in Tao's road

map um based and boosted uh so we're going to explore this new rollup design um that comes very close to being a native ethereum scaling solution call it a based boosted rollup uh we already learned about based rollups and how they are arguably the most effective decentralized ethereum aligned way of running a rollup uh by delegating the sequencing job to the L1 it inherits the l1's decentralization Simplicity and liveness we spoke about uh essentially all of those properties um so skipping through some of the uh based parts so now for the more technical folk here's a one sentence

technical summary of what booster rollups do booster rollups shard the execution of transactions and the storage oh boy let's dive in go ahead so yeah so Shard the execution of transactions and the storage technically decent teally decentralize it the execution of the storage okay Bas robes is a black bear booster r as a white bear Bas it's 2023 we're all going to come together and create a panda Bas booster rollup bvr baby all right so uh any okay so any any rollup uh whether optimistic or ZK can become boosted so

let me just give a stabit to what I think this is first B my conversations with the team I'm pretty sure what they're trying to do is uh you can you can deploy your your dap on l2s without needing to actually go through the full deployment process huge developer easability thing developer convenience uh major unlock yeah so so basically if you want to achieve native ethereum scaling so let's say you have a Dap on L1 you can launch off l2s as a base boost draw so um let's di into so it base Bo ropes retain the advantage of the base rope and adds direct ethereum scaling um basically by allowing L1 valid proposed blocks for the whole boosted Network you're scaling ether out

of the box so again like you're you're using that L1 consensus to to validate the the l2s um MH and let's have a look at the diagram here so you can deploy your dap on L1 once yet execute it on all of l2s because the smart contract logic lives on the L1 it shards the execution and the storage ah the difference between parallel and Shard yeah because parallel parallelization is a is a pretty popular concept now it is par parallel execution or

parallelization kind of comes more at the VM layer I think where you can have multiple sets of transactions happening at the same time which don't which don't uh affect the same fee Market uh so you can have instead of a global fee Market what you get is more of like a local fee Market where you can have an mft mid happening on one platform uh charging fees but not affecting the price of gas in Defi and the same execution environment whereas sharding I think is more like like uh kind of like a scaling mechanism and I honestly think of rollups as sharting where you got ethereum as like this Hub and you have all these different Beach to rollups as shards of ethereum that's not that's like the original vision of sharding for where they split up the actual L1 that's not it's like the original Vision but

that's how I think of it it's like splitting up the L2 I think rups are effectively the Practical Vision right they just kind of like took it off of the base yeah one and they put the sharding at at a at a level uh away from the consensus exactly exactly but they would use a the consensus right right uh but they're not breaking up the consensus into different like little cities where they have their own like you know validator consensus mechanisms right EX in those no no it's all these booster robes I think that's I think that's important otherwise you would end up with uh like an easily corruptible or at least a less expensive Network to corrupt if you break up the consensus this way you maintain like a very high cost of corruption so it's it's more difficult to corrupt but you still get

the scalability by kind of off loading all the execution onto the rollups and then splitting those ups are needed to scale the the way I was imagin originally I legit think that bbr's are effectively sharding but in the Practical form like rollups as they exist today with C sequencers are like the form that we have economically which makes sense to Shard ethereum based rollups take it to another level as far as security and uh livess and decentralization of of of roll now bbr is take it to a whole other level because it's like okay now we have base ropes but now we're also going to implement like metallics original vision of like sharding the ethereum main chain

into 64 shards no we're just going to do it in this capacity but it's effectively the same thing where you can launch your dap on L1 and all the shards that are base boosted lups can can uh can uh execute and use it as was normal but you get the L1 validation and then that's how it scales and the L1 sequencing which is a particularly uh uh unique part um their Bas rollups right they they so they are are using the L1 for sequencing and they're all able all of the base boosted rollups are able to use the L1 for sequencing and that has the benefit of atomic cross rollup transactions and again they like I was

listening to uh to a panel with Charles from espresso about shared sequencing but the concept of atomic uh composability is consistent across shared sequencing and and based sequencing at the L1 because at the L1 layer uh all the based rollups are essentially using the same shared sequencer it's the L1 that is providing the the shared sequencing yes so so with shared sequencing whether it's at the L1 level or or another level you're able to get Atomic composability and that's where like just you can do so many you can do tons of cool things you can get greater price efficiency you're able toer is this is even better than shared sequencing because in shared sequencing you can get Atomic inclusion of

transactions in blocks but that doesn't necessarily equate to Atomic Atomic execution because there's often you know plus or minus x amount of seconds milliseconds or time difference between blocks or between the the way that these transactions are sequenced in the in in in in the block so you know you share sequencing gets you really freaking close it gets you Atomic inclusion of into a block but it does not give you Atomic execution because Atomic execution is not the same thing as inclusion in the block this is giving you Atomic inclusion and execution because it all happens at the L1 layer however again you could argue that like well is it really because yeah sure they can be included in one block but not executed the same they can be executed

in that same block but they're wed in a different way and so you know again Atomic Atomic comp possibility is a tough one because inclusion is not the s's execution yeah yeah I mean there's there's tons to uh to unpack when it comes to Atomic composability um whether that means you get inclusion and execution one or the other um no one's really like perfected that yet none of the shared sequencers um even even base boosted rollets are still are those of not even out right or I don't use a base R maybe there is I'd be confused doubt it so what what does it all mean let's let's roll it up um the bbr approach the Bas boosted

rollup approach means that um rollups no longer need to worry about fragmentation and jumping from one L2 to another uh favorite daps because they're they're deployed at the L1 lay are able to exist across all the L2 layers um this is the sharding mechanism that that uh Tao enables um so Tao is inherently like a base boosted rollup Network where you can deploy uh and then you can deploy to the L1 and then all of the rollups the base boosted rollups that are part of the Tao Network um they all are composable with one another yeah yep and this can be implemented across many different um ecosystems as well from my

perspective for uh for devs you can just take the vvr design um and Deploy on all l2s um without need to like it says redeploy new code you just deploy your app once on L1 and you're your dab is automatically scaled across all boosted l2s now and in the future so like the situation where you know you're a Dap Builder and you're you're trying to consider where to build your dap well you for sure are going to build it on Main net because that you know that's where a lot of transactions have that's what a lot of value is now we build on Main that tyos coming in and saying okay all boost your R will have access to your dab thing about it and that kind of can dictate your strategy as to how

you're thinking about Deploy on different network will come in and they will boost your based rollup they'll boost your rollup as long as it's based no no no they will Tao will boost the your dap on L1 like they will allow it to be on any based uh based ruup that exists up perhaps they could also boost a previously base role but I I want to say I'm I'm not sure I want to say you have to launch it as base boosted role in in from the get not okay but perhaps we but only title could take optimism and like turn it into base boost I see so boost so Tao allows them to boost their daps without the need to redeploy to L all l2s instead just deploy your dap once on L1 and you're done your dap

is automatically scaled across all boosted L2 is deployed now and in the future uh so when building so yeah it's almost like they're like they're building more or less like a I'm not even sure contestation means like um uh contestation is is about bbrr is based off like just based contestable World mechanism I suppose so um they're going to implement in their main net and then they're going to do an upgrade the protocol or launch a separate L2 separate from the zkm as a major Milestone and then I think they're just gonna spin up like a a cdk probably or a or a Rd a bbrd kit based boosted will development kit and uh and see happens from there so epic yeah that's Bas rups

well we're gonna have tons more Tao content coming out and and it's still so early there's still developing these tools so this is still more of like a conceptual conversation for now and just stay tuned for all the content that we have coming out as it becomes more practical to implement on chain yeah Che thanks guys