EIPs and Decentralized Rollups

Transcript

Now, please welcome our next presenter, Mr. Hari. Mr. Harun, the floor is yours. [Music] >> Hello everyone. Uh, my name is Harim. I am a smart contract lead at Offchain Labs. And today I'm going to briefly talk about some challenges of decentralized rollout. But first let me explain what is option labs like in the arbitum ecosystem. So option labs is a service provider to the arbitum foundation which act as a steward to the arbitum dow. Albert Dow govern the arbitum one and arbitum no chains and like option labs provide

technical expertise to the foundation and DAO such as like shipping upgrade to this chain when like we need to upgrade like say uh we have like a Ethereum hot focus new EIP. So as I said like or like I mean maybe a lot of people already know one is like one of the leading decentralized roll up and friends to our friends at L2 bit like we can look at this dashboard about like um some um some summary of the characteristic of one and you can see that one is actually a stage one roll up according to their definition and if you look at the pie at the right side you

can see that we have Every slice is in green color except for one slide called the exit window. So what is the exit window? Like um in simple word you can think about exit window is like a time lock for rollups like you allow user on this roll up to exit the roll up when they don't like a certain change. For example, if the owner try to run or like they try to like make a change to how the the bridge is like custoing their fun user should have like a window of time to exit the chain and and even if the sequencer is sensoring but how do you actually um guarantee they have like certain amount of time to exit like it's basically a mix of waiting and also like

some protocol level like sens censorship resistant like cook such that you make sure like if someone want to exit, they will have the window and also have the like ability to exit the chain. And why really not a stage two right now is that because like if you look at L2 based like u requirement on the right like bottom right side, you can see that they actually require a 30-day exit window if the change is not related to a bug. And we also look at the life cycle of AIP which is average improvement proposal and like due to the governance latency people have to work like we have a ch period because we are optimistic roll up there exit window requirement

there also like censorship resistant like like requirement and AIP actually will take at least 35 days from the proposal creation on chain to execution and that is not including the time of a potential column extension of two days. Like um we have an optional snapshot temperature check of one week and also like there are audit time for any code chain which is can be weeks and then of course there will be development time for any change we want to make and you can see that like it's actually like really like a slow process once your your rollup is decentralized and it's actually decentralized it because it's like take a lot of time to ship any change.

So why are we talking about L1 hard fork? What is the implication of roll up? There are usually like three types of change that can impact a roll up. Um the first kind of change is basically um L1 breaking change. Sometimes like L1 decide to ship certain breaking chains that like on on on L1 and then some example will be like 7702 which like changed the behavior on like in the past for example we can detect if you're a EA by using text origin and message standard you cannot now and then like in the upcoming forak we also have like a new change that going to introduce a transition gas limit cap. So this like some of the L1 braking change can impact roll up

because like for example the proofing system of the roll up may actually need more gas than the transition cap and that's something we have to evaluate. Um there's another cost of like difference is like L2 I have to implement new feature of L1 like basically any like Y change we have to implement on L2 and like some of the example can be also be 7702 because like like when L1 have 7702 like user would usually expect to have like 7702 across all the chain and then they can use abstraction everywhere but like because like of like the process I talk about is really usually really hard for us to ship that like when like

magnet ship the same feature and then we will user left a window that they just cannot use this on average one and the third type is basically about L3 L3 is like arbitum's term orbit chains uh arbitum term for like roll up on the roll up and then there are additional cons consideration when trying to implement EIP on this chain because of how like the parent chain is different it's not actually L1 L2. So you have to consider how L2 works when you implement those chain and there there's like additional comp capacity when we have to implement those feature.

So like to recap basically there's like three type of like challenges. First is timing like do we want to like ship before mainnet or like at the same time or after like it's really hard like actually like to ship like at at the same time as magnet because of the 35 minimum delay. Usually it's like two months to be honest because like we actually have to write code like no AI haven't replaced that yet but like um that's a thing. Um there's also like certain like um sometimes they make last minute change to how what EIP is included in the hard fork or like what is the actual implementation like for example if they figure out like they suffer a bug after the test net fork

then they have to make change or like for example famously like in the past we have like a year removed it completely from the for like the last fork and then that's if we like decide to frontr run mainet and prepare everything like in advance then we will basically have to scratch everything and then like redo and it's like the problem like what usually is that why we want to ship as close to mainet as possible is because like user expectation like in a road centric road map like user want to have the same experience across different chain likes but if like for for example if one don't have 7702 enabled in day one is really like not the best

experience for user and then um that can make like like the experience worse and this going to be even harder when the exit window is set to seven day to fulfill a stage two requirement because like when I say 35 day is actually like built around a 10day exit window like like considerate like assumption but if you want to make it 30 day it's going to be like at least 60day kind of like scale we're talking about and then it's make going to make the thing even harder. So where do we go from here right like I mean as a one and other roll up progress and want to be more decentralized there like they want to

like move away from like having a onchain dow voting on things but like then having another multiscake to execute that's like the easy way of doing decentralized stuff but we all know that is not that decentralized right because like the that those multis usually don't have a very like obligation like at least they can kind of do whatever they want like and you want to do something more close to what the AB Dow is doing which is like allowing the DA to have complete control of everything on on um on chain and then um there's also a security console which can make change like immediately but like we usually like basically like

security console is built for like security issue like like working shipping and mainet like like EIP IP is usually not a security issue. So like we usually don't use the security console for this kind of stuff unless it's for example the type of thing I talk about a L1 breaking change that may actually impact how the L2 like security. So how do we solve this? Um I feel like I don't really have a good answer to be honest like decentralized is really hard like we are still early we are still figuring out like Dow is one of the biggest D right now and then but like there are still a lot of like issue we have discovered like in the recent time like how how hard is to run a D to make sure

everything is decentralized also but also functional but I feel like there's like a few area that we might be able to improve like one being the VIP process like we should like try to avoid breaking change like that's like I don't think we like usually like try to breaking change but like sometime like this this thing happen and then like we should like have better like tracking on like what like the IP might break a certain protocol or not. Um we should also aim to see if we can have like more clear signal on when is the hard actually finalized it to avoid any last minute like change. um DOF probably one

of the good example uh but like you know Ethereum itself is also decentralized so like it's also something that really hard to figure out and for the Dow side I think there's probably something we can do from the how u to improve the D governance process although I currently don't see it like something that can be changed materially because like a lot of like delay is basically just giving people time to vote and like like exit and you you you don't really want to like say hey I'm just going to this going to give this multiscake the ability to like ship whatever change they want like this also not very decentralized sounding so maybe not something what we want to do um user education is probably one of the thing

we can do like maybe we should let user know that L2 actually because of the security properties they should let behind L1 like they maybe they should not expect L2 to always have the same features set immediately ely as L1 ship them. But that's also hard like I mean even if you educate them them they like the UX is still bad. So like probably not the best solution. I hope like someone like maybe we can also think about this because like I feel like usually like these um like we want everything to be decentralized but we don't usually think about too much about how these thing operate once they are decentralized it like we have like a long way like to go before we can make

like um L2 rollups like fully decentralized it optimistic or not I feel like this like a problem like for for it's actually not a problem just for roll up. It's probably all for all other deps too like because like if you're decentralizer you will come at a time where like you like you will need a lot of time to ship upgrade and then it's very hard to like anticipate like when what is the IP is going to incur and then how can you time your upgrade with those. So thank you like um yeah thank you very muchent thank you for waiting now resume the

presentation now please welcome our next presenter Mr. The Suzuki [Music] hi hello everyone ZMZ uh thank you for coming today and for listening to my talk. So hopefully like you know this talk will be or one of the unique talk you will listen through this aircom and also like I I just wanted to say like thank you to like Eric organizing this is a really great event

and thank you so much for inviting me to the speaker. So there I I will start like about like my introduction first introduction. So I'm Udai Suzuki. I'm co-founder of Fractton Ventures which is not a VC or this is a incubator based in Japan other my background is a record startup incubator since 2017 18 or since then or I saw first generation of blockchain startup everyone called blockchain startup at the time. So the I supported record blockchain startup first generation or in 2017.

So I since then like you know I was like a big fan of like I I'm a big fan of blockchain technology in the also like a decentralized infrastructure technologies and uh so like I introduced so many like blockchain first generation of blockchain startup founder Japanese founders to VCs investors in Japan market but anyone everyone couldn't understand this potential one of the guy could realize try to realize that such such a kind of similar application of record or or chain analysis at the time is it 2017 or maybe you guys know chain analysis or offer recently signed with FBI or any big American agency using

this technology as well but you know there so many record BC couldn't understand this point you know why someone why you where is a user who would use this technology so you have a blockchain you know is a scam. So it was a it was my beginning story you know. So since then I have been starting to oh we have to create really strong ecosystem for Ethereum in Japan. I just started like a founder community first in Tokyo and for all of like blockchain and especially for Ethereum protocol founders I collected in there and like do a like a drinking or networking event together and for knowing each other. So this is my starting ry. So fractton or

this is like a Tokyo based incubator and it's not only for Japanese founders or we are providing incubation cohort and as a cohort style to our more wary people as like as like Asian founders too. So the fractton has a three things we are doing. First one is a fractton incubation which is a flagship incubation model. So that we are providing cohort style means six months or three months. Cohort style record acceleration incubation model or plus educational content or plus inter or super or or interesting or innovators like mentoring opportunities or we are providing through this program. So fractum enterprise teams. So like

recently we created enterprise teams for or supporting more enterprise. So how our enterprise or big corporate especially in Japan too or can understand the value of ETH. So there is a two meaning of like Ethereum the Ethereum is a world computer that it can covered by fractal incubation through this inc protocol incubation. So there is a ETH is money ETH money is covered by this fractal enterprise. So this is the things and also like we are organizer co-organizer of east Tokyo and we are organizer of DA Tokyo.

Yeah there hopefully like so many guys like already attended to this year's D East Tokyo as well. So it could be success like successfully we finish record this event. Hopefully like we will re back to record Tokyo again and uh hopefully record Eric will keep providing our event in Osaka as a like any collaboration that we we'd like to see and we are are hosting like a monthly meetup in Tokyo and like if you come back to Tokyo please say hi to Fractton team. So if you want to talk about something, if you want to share your research, please use our meetup opportunities. This is for yours. It's not my handle. So this is for yours. So fract is doing recruit entering is doing

eMM incubation is Asian protocol funders and the incubator over 40 protocols is a target is very simple after hackathon and two angel and pre round. So there nobody this is super tough hard this timing for cryp founders you know anyone will not support normally you know you are just challenger oh you are doing things okay like good luck that's it so who will support strongly for this artist founders this is the things I will skip it the cohort five will start very soon we are we are expanded like a deadline of this application if you are building really new protocol and the super in you are in

super early stage please apply to fractton website so please find like our Twitter or x record or our handle is we can dow it so please search we can dow it so there yeah I would skip it so as a back to the main topic so we are eam everyone is our etherans so we love why we love Ethereum I think there is a three reason I I personally too the first one is a decentralization decentralization has a lots of meaning there node client re decentralization geographically re decentralization and or any other record you know or any other or any specific or name of

organization couldn't provide all the record you know central record or top down redressive model in the market or something like this so the and uncensorship. So, so we are we need this point as well to because like ERM is for everyone. I posted this this word like yesterday EM is for everyone means like you know any single person like or we will we should not to shut out them from network. It means like you know the someone carefully AML okay like you can use GKP for checking AML right. So like Ethereum itself they shouldn't do like you know

shut out someone from the network and the last one is a positive subcoordination. So oh sorry so maybe like my internet my connect is not good. Okay so the positive sub connection coordination is like a most important things to do it. So the everyone is like standing on open source based culture. The open source based culture means every protocol can forkable fork forkable or to from source code and you can use use the code. So the record and as a tokconomics too the if token incentive model is a bringing to new value and outside of the community to into the new community or to own community the community will grow. So

there everyone is just a comm talking about like we have a community we have a community there are you is this real human is this is a program you know so many bots maybe you guys are hold like over 20 wallets or something in person right so like so like more important things is that like how we can create like a more positive some loop or looping or through the protocol design through the token token design as well. So big question what are we missing in the Ethereum ecosystem? When I say Ethereum ecosystem it's compar or aligned to like Ethereum L1 plus L2 as well. I can describe as well and so next question is Erem ecosystem means it call

EBM ecosystem. I personally think sorry for any other EBM guys like you know I personally think EVM is not a ecosystem it's just there is a just a solidity developer developer can write solidity for virtual machine can run using solidity language this is just a EBM EBM is completely open source everyone can use it so EBM if fork from like a Ethern is sorry Ethern to any other chain is it can network it's it doesn't mean to actual ecosystem I'm personally thinking so so this is my imaging so there is a etherm l1 there's em l2s and the specific side

chain when I say specific side chain impersonally I think like a nosis polygon they are in so this is my imagination so everyone can easily to say we are in like a EBM ecosystem together right bro no no this is not an ecosystem Yeah. So this is real ecosystem Ethereum ecosystem. So as a question is that like you know where are developers in the market. So the maybe like just research has made by electric capital there they are reporting like every year for where is the developer are in in the market. So there maybe like uh Solana other chain

or internet computer all interesting appto base oh bas is ecosystem eum ecosystem as well so near okay so I don't know the I think this is just like a developer number so it's not it doesn't mean to number but like also like where is the developers if developers keep developing some solid or using sority for EBMS in other chain outside of Ethereum ecosystem. I want to say like please come back to here how we can do it.

So how can we EBM developer return to the Ethereum? This is a program. This is a question for you all. So why don't they build on Ethereum? Please think about like one minute like why what do you think why don't they build on Ethereum maybe other chain has lots of grant opportunities printing money printing money something I don't know so and the partnership offers we can easily to say co-arketing together bro okay okay it's good so the my interest is that like you know if EBM developer means like a solid developers

rethinking about the what is the meaning of Ethereum development what is the meaning what is the value of like development for Ethereum if seriously they will start to rethink again I think they will go back my answer is that no place to get serious mentorship for super audi protocol founders the reason is that so the point is that you know so many audi developers looking for opportunity as grant as a partnership.

The reason is that like you know you don't know any new protocol names the new protocol comes in the market there there's a really serious founder they are developing over 20 years writing a code this is crazy this is good but also even the situation you don't know the name of the protocol right you will not use the protocol so if they say you know we partnership with bra foundation ABC foundation ABC foundation ABC braos something like this so the record so we got a record grant and or from like this hackathon this finalist something like this. Okay. So like you know or finally like a foundation approved your name and also like there maybe like maybe will maybe someone will get interest to use

this application this protocol as well. So the my why I say this answer is a my opinion is that there as a incubator EM incubator we always thinking about where is a good place to like get provide a mentorship for super early stage founders. So like I I am believing like all of attendees in here is like a maybe like a maybe like a half a percent is a speaker in the aircom over half a percent I think so. So there now you are speaker on Edcom but like a you were like a timing you were timing to like a just a super beginning

phase of the protocol developer protocol founder right so at the time who supported you so how so stage protocol founder need opportunities need opportunity means like are if new protocol creation activity comes from hackathon. They have to start to thinking about like is this a good idea? Should I spend my life for keep building this protocol? Something like this. Okay. You know, you have to share this idea to like a very good mentors you know. So record I think these guys record you know these attendees are really very good where if

we you can you have a motivation to do a mentorship for Ethereum ecosystem from tomorrow maybe new founder we will see more new founders and they need I say like they need mentors mentors is not just like angel investors but also like are more collaborate together and share their networking resources trust resources and I know him guy the the guy understood at this point you have to talk with him like that so someone should like introduce or in to or bring new founder to core deeply ecosystem prayers and start to doing a chat the type of opportunity is really really

important for everyone and the and the last point is they need a grant or angel investment too so there it's it's depends on the But you know like you don't need a money you have a money and you can keep developing this is the best but also like realistically you know there is a timing to like you know we are we can easily to face financial program we have an idea we have a motivation we have a good team as including developers but we don't have a money okay you have to raise raise the money so the so one question big question especially for early stage funding or fast funding is that like you know

the investor audi investor investor too can easily to ask the who or which ecosystem which player approved your name in the ecosystem so you know like it's just bit of hard it's just I think this is a just like a chicken and egg program you know someone should approve like I'm supporting this protocol but also like before coming to this information so like investor cannot say yes like okay I am personally like like you but like personally like you means like you know I can't invest like a to your protocol directly like on that so what we are doing in the space so fractum benches is a continuing to

provide 1% of token offer to incubate protocols so we are not investors we will not invest we haven't invested but we just offer report like a 1% of future token from each of incubated protocols. Someone say oh you guys doing like a ban three work in the space kind of kind of yeah but okay why I saying why I why I'm believing this model is good is that so other other incubator or accelerator in the market in web3 in c of crypto they are are offering around like a four to six% token to each of projects maybe you guys can understand four to six% is

quite quite big number you know maybe like a roughly 20% of equity or something the similar meaning it's too risky to do this and more important things is that why I offering why we are offering 1% of token to each of project the reason is that the we are not an investor we we want to take we want to get more position as as a side of like a more co-ounding member rebel for you guys to keep supporting your project. This is a sense.

So we started to thinking about this model. We have been providing this model since 2021. Over four years we already incubated over 40 protocol 45 protocol in the market or mainly from comes from Japan but not only from Japan. So recently we started to new dowo or alliance the initiative it's called east reactor east reactor is like a new dowo formation a dowo dorum incubator collective for etherum why it needs because like a preach on like a east reactor or eth without e reactor or east reactor on x there is a

there is a x account only 100 members following right now it's super beginning timing so so why it needs fractton offering this model so ino in in in many days and also the like you know we are also thinking about like you know how we can do more global rebel same action in the market so reactor has recently started with these communities They're from Japan or Fractton will join in this D. Suchi or blockchain rout which is a behind of like East Betam they will join and narrative from South Korea will join East Sudony has joined

and also like Isa Cinco de Mayo the which is like Mexican community they has joined and like is Panama community is is called east cano has joined and the napoli east which is Italian community has joined and the isabare which is a servanian community it has joined too. So what will will we do? This is like a E is reactor is like a global D formation through by our E is react umbrella powered by incubator collectives especially for Ethereum aligned incubator collectives.

So the meaning of this record there is a lots of record hackathon in globally. So you can you can jump on to the new hackathon new hackathon new hackathon as a mentor ja something like this. So behind of the hackathon there's so many record regional record reader Ethereum reader player in the market we want to on board them into the one place to work as like a DAO member DA formation together to do incubation and offer 1% of token in regional market everywhere. So there technically like if we on board if we reach the on board every global country into is reactor is reactor brand can get like a global access through

this global is reactor brand name this is a dao so so this is the formation of east reactor actually do so regional incubator joint in this dowo the regional incubators means like you know behind of the is hackathon or is incubator they are already ing and run on incubation program together and globally. So the through this is reactor names is reactor brand. Okay. So we we can run is reactor or makish cam or incubation program like this and the request like incubator teams to allocate 1% token to do it and one of the goodest motivation for us incubators

is that we can share more idea more regional opportunities collaborate discussing with incubators circles this is a one super unique opportunity to do it normally Japanese incubator like us don't have a motivation to share what's going on Japan. So what type of protocol we should incubate from Japan with international people or any other ideas ideas normally we will not have a we don't have a motivation to share the idea to any other countries incubators sometime like we can take like you know it's not an enemy but like you know it's like just a bit of like business it's not doesn't mean to like business

synergy normally but like can create more aligned incubtors corrective success this is focusing on. So this is a things of like EAC as model. So first as I mentioned earlier so incubators they're normally as as IM incubators we are always supporting really very early stage founders they started from hackathon they try to hackathon so many times and starting a build record teams and just finalizing the ideas and also record here recent or you know research everything and doing for like a real finalizing think like of

this protocol architecture. Okay, this is good. So maybe these information like who is our underrated talent in the each of regional market could capture we capture by like a regional incubators right. So Zer if you are living in your country you know everything this country normally around your community and who we will who will be joined who who are recently interest in this market this protocol this type of idea like that like that imagine like you know the like if someone or if new market comes in the market like a uh like that if there's an idea like

New Oracle mechanism is need of for the market any specific reason. So okay this is good okay so like through this ecto brand we are starting like researching on everything who guys are like act already started to like you know research on new architecture model or new value model of oracle mechanism or there there's a lots of oracle in the market or including chain link too. So why we need new type of oracle is there is this a critical reason someone should create a new protocol or not and if it's it's true okay you know what type of architecture is the best so e reactor is

a hub for everyone as a correct more ideas protocol ideas into the e reactor brand and also like we can share ideas with like a collaborative arigned incubators and story through the incubators we can easily to find really good talent in regional market maybe some case the founder comes from like your Japan and the Korea and Mkisho and the co-ounding together for one protocol I think this is the best so reactor can work for more providing more increase or coordination opportunities for in the Ethereum ecosystem especially for Ethereum incubation phase And also like as as I mentioned like is

director will do our cohort style like incubation programs. So normally incubation program are is just a bit of hard to like you know are are providing actual beneficial or benefit to mentor or guest speaker each of programs. Normally everyone everyone doing just a super voluntary work or just accessing to like RD protocols are only the reason they are doing but we have to create more actual like incentivized model through this action too and if really good mentor support deeply to really specific founder if founder will success in the

future the mentor will record it something calculate something as a record his mentorship It was really very good in early stage something like this how we can capture how we can track on this. This is the things and the left one is a record grow them as a record regional reading protocols. I think as I personally think is there depends on the market but like especially in Japan too the market has been recently changed a little bit. It means so the puzzle of timing there I think like ecosystem is like a puzzle piece of puzzle. So there so many lots of piece of puzzle are already in in the puzzle. So there is

like a big lending protocol there is like a you know stable coin protocol there is a you know the risk takingaking protocol like this. So if protocol is already enough someone should aggregate protocols into the one application and sip as like one product. Why I saying like is that like a protocol is more ossbased like open source based smart contract based one and the product means you know like more providing record the things on it like this.

So and the last message from isctor. So easterector is like this things. So we have support for ridge founders and if you are running like your regional community please let me know or we'd like to on board you to retor. So what let's work together. I think this is a new chapter for EM incubation. Thank you for listening. [Music] Thank you very much.