Getting to Lean Ethereum

Transcript

[music] everyone and thank you for being here. Uh my name is Leandro and I'm here with Ibam. We're both from Lambda class and among many other things we have been working uh building execution client and more recently in Lambda which just got their first release for the VCON chain. Uh so today we're going what we wanted to present is like um our point of view and our approach on how we are seeing things into getting to linear right and a little bit more on of what is linearium because well it's something that

is being spoken in in in many places and we're talking about postquantum cryptography and we're talking about the cap consensus and well this whole thing h is part of of what init initially was kind of the idea of the bean chain and and let's start here. So I want to talk to how we get to the becom to lin right and Lynn is not not me is the name of the chain that took my name. I don't know why but initially I I think um the idea of the being chain was introduced I think that two years ago almost two years ago in in Devcon in Bangkok in yeah in Bangkok it was in Deon by by by

Sashin Drake he introduced this idea of of the of the beam chain but uh then it changed the name because I'm back and back and forth in in legal terms but the idea of a be chain was okay this is the evolution of ethereum this is the next step of ethereum this beam chain well that's then that beam chain was called lchain but between two years ago and today many things actually happened AI happened quantum computers are a threat that that is getting much more stronger and and well yesterday even Justin talked about this alien maths that are also a threat. So with all this

I want to present a little bit where do we have the time? Okay, there a little bit what's the the the idea behind um LIN. So now we are in in BCOM right and in BCOM what do we have? We have BLS signatures for the validators. We have ECDSA signatures for the users. On the consensus side, we're using Gasper, which is like for choice and finality that are coupled and we're having a finality of around two epochs and plus a couple of more minutes average that taking us to 16 minutes and all the states of the validators are tracked in the protocol. Right? I'm sure most of you are aware of all of this because you

working this is a protocol track and probably most of you know more about this than myself. But so what is Lynn? Lynn introduces this changes in the crypto and the consensus of Ethereum. introducing postquantum state signatures, a new consensus algorithm trying to achieve finality in seconds or at least in less than a minute. Uh and well most of this is a student research and then we have this extremely lean which is also lean. So the the final state of the lean at some point was

talked about this idea of extremely lean. I separate these two ideas because well that is in more research state than than the lean itself. And the difference between lean and extremely lean is basically who keeps uh the the state of of the of the validator and in the extremely lean chain idea. Each validator keeps their own state. So in in the chain you have much less data to to sort right and and they only upload to the to the chain a proof of of it.

So the first step kind of changes the crypton and the consensus in both layers and the step two is who carries the state right. So with this um with this idea of of of lynching uh there is an important thing which is okay quantum risk right the most important thing is is quantum risk but it's not the only the only thing we we can just change the signatures that we have right now in in the beacon chain and and we are quantum safe we can say that Justin just in introduced yesterday a couple of things and we have been having meetings with the protocol with the with the core devs about this that can be done. But this idea is not just

changing the signatures is more like okay let's pay technical depth that we are having from from all times. Okay, let let's change all all the all the consensus the way the consensus is designed right now is based on okay what we have at the moment or where we're coming from and the idea of ling was changing this in this uh what what I'm I'm saying about okay we have new cryptography and new consensus this is not uh definitive but still in research But it's almost the there is some consensus about this but it's not confirmed yet. It's still

exploratory. But the idea is to have XMss for validators. XMss are are are hash bait. So it can be easily aggregated. And then for the for the user for the execution layer we are exploring the uses of linfix. I I think that was yesterday there were many talks about this. Uh both of them are postquantum safe. Both of them are hashbased and both of them work very well with aggregation stark. These two and also lindda that I think that kev yesterday mentioned something about that uh are being uh proved but this linm that is being uh also developed and also being spec. So

we we will have like this BM one minimal CKVM that aggregates every commitment for XMSS for sphinx and for Linda and for the capital consensus the biggest idea is to separate fork choice for finality votes right so right now we have one single vote that we're voting for the two things at the time and that is getting us a finality of two epochs and we have the possibility to separate these two things in in in two different layers changing also the algorithms what what is being explored is goldilocks which is like goldfish plus locks which give us some limeiness uh and for finality some simplex I'm not

going into details of this because I'm not the expert as well of those details but but some experiments has been done around this and and the results are quite good all of this is been also formally verified. I will talk I will mention that a little bit later. And for the validators when I say this extremely ling chain which is steam research. Well each validator the idea is that they keep their own state. So we will reduce a lot how much goes on chain right and the idea is to use a stark as aggregated signatures. I think Tony mentioned not not this but well there is something with A288 that that also does some some

aggregation not for the not for the state but for the signatures everything is like going through stark aggregation. So [snorts] uh there is some uh source about Vitalic talking about extremely lane uh from July this year. [sighs and gasps] So lynching is kind of a destination and it's being worked like in in two different parts that well Ian now we'll talk about that a little bit. Come on, we'll talk about this. But the idea is we have consensus, we have ctography, we are talking, we are building the specs. Not we, this couldn't have been possible without all

the researchers that are working on this. We we have been building a client. We have been helping them. But actually all the hard work is being done by the researchers and we're very happy about that. And the and one core idea that I'm missing is that all of this is being formally verified. Most of this is being formally verified. Limbbian codebase is written in Rust and is being formally verified in four. And the idea is to write that 100%. Having said all this, let's talk about these two roads that can be done.

Okay. Hello everybody. Does is the mic on? Yeah. Okay. So we are not doing a just one hard fork that includes all the lean features. So now we have two roads that have to meet and we'll talk about how they are meeting today for in one hand we have the PQ interrup green field consensus open road map. This this was Justin's initiate initiative to bring all the lean Ethereum features at one fork and there were several clients that joined this effort. We have rim seam qin lantern

grandin gene and ours if lambda there were the the efforts were done in several devets. Now these efforts are kind of post uh because the road map is changing but we got to do several devets. Each one of them focus on a single feature. Um yeah and we got the last devet was devet 5. We didn't get to devet six but we are getting it in the current uh bacon chain road map in the next few years.

So this was the stroll map a couple of days ago but it has now changed. The the idea of the old road map was to have all lean Ethereum features up until uh 2029 but we are now uh making things ship faster. So HAR will be our last forks with without lin features and on AR we'll start having them. We are doing linvm and lin sphinx transactions on IAR and we are doing linux MSS at the stations lindda sampling and lin pcs on jar

and uh by ear we'll have lin ethereum complete and now lean will talk about the dilemma. >> Thanks. Okay. Well, and Ivan just mentioned that okay, we have these two parts and at at the first it was kind of a dilemma. How do we how do we get to Lin Ethereum? It was these two ideas. Okay, we can do one big fork. H or we can just go like okay fork to fork upgrading and well what are the difference between this? If we go with this green field idea, we work in a blank page. It's very

easy uh to do. We we did several dev nets, one month for demnet. We have the specs out of lank. We had no migration to think about. It was very easy. What is not easy is is okay when then we have all the clients that are not there. So how do how do we measure those those words right? Erh so with the sum up idea it was going to be a longer path uh finishing in in 2029 probably with upgrades uh in forks and each forks had only one headliner per layer. So it was going to be a longer path but everything like lean towards the idea of

okay let's do the I don't know the the most safe thing where clients already are where people already know and with the process that we already know that is okay let's upgrade as regularly in the forks so we are doing incremental uh wins And what what I wanted to show here is okay the things that were were tested in the different devets of of this PQ inter group the PQ interg group I I don't know if I mention it but it was like mostly led by Justin Emil and Thomas that did a a great job over there and I and I and I truly hope that we'll continue doing a few tests

over there. I think they're still worth to to explore in this green field a couple of ideas but many many ideas from the dev nets ended up in okay a mainet entry in the stroll map in a different uh fork. So now most of them with the new stro design are are being like merged into into mainet before jar or up to jar [sighs] and this har fork is introduced as the last normal fork right why why the the last normal fork work because it is the the last one that doesn't have any lean [clears throat] or postquantum related thing and it's

the one that will prepare us to be quantum safe, ready to begin upgrading and to begin uh migrating whatever signature we need to migrate in the upcoming forks. Right? So the idea is with frames we will make crypto swappable. H and well we have this I have this in red. We had the idea also here that didn't succeed to to include optional proof already in HR. I believe that's going to be in IAR or or even later. I don't know. I'm not sure but that's our point of view. So in HR we will put everything that shortens this gap between uh where we are and quantum

safety. what we done right now very quickly I have time but very quickly we have built in Landa Landa was was born in in the lin uh side of of the world but with all the experience and with experience that we have in in the in the execution client we decided to go also to the to the beacon side of the world we have the first release out there we already transition to glamor in in sepoleia and the idea is to have a client uh to have one clients in in the two fronts with modules and can swap. So we can test like okay if we have still something that we want to try on this PQ

interal we can use the client and the same client will be used as well for testing things in the current um upgrades of of the network. So we we can easily swap modules. So we we can do that and also with the idea to integrate this into so is the execution. So we will end up also having the execution thing without even the engine API. It will run faster and and so on. But yeah, so this this was linerium. This was I believe and and with conversations that I'm having with with many people, it's kind of okay that they they were not aware of

this as a whole thing, right? It was okay. We have the postquantum menas. We have this idea of the coupling consensus but but as this PQ interrupt was interrupt group was an initiative that that was separate. Okay. No one has like all the information we wanting to introduce this because we believe it was a great effort. And with all that effort, we we are much more ready now to decide what to pick and what to prioritize to to accelerate the postquantum leap that we we have to to get in into Ethereum.

There are still many many things that need to be decided. I don't have answers to many of the things that are are being discussed. Uh most of these are is still in research. Uh the other well recently we saw um great talks about the the cop consensus and and with insight with information with data. They are great. You should you should take a look into that. We are seeing great topics about the cryptography. You should go and take a look into that. H cryptography is getting much more better, much more optimized. So I'm optimistic about the idea of having the change of signatures

in Gasper. So I I do believe that that will be possible because great people are fighting for it. So we will do it. But what there are many many open questions that were here and with this uh no with all of this I I think that that's most of it. These are our socials if you want to find we'll be around here if you want to discuss a little bit more about what are we doing at what and what have we done with in in this lean initiative and what thank you.

All right, let's give our speaker another big round of applause. Lean and Ivan. All right, let's get both of them on the stage together. Hi guys. All right, thank you again for sharing a little bit more about the journey towards a lean Ethereum. Yeah, a bit of a punny thing, huh? Lean Ethereum shared by Lean. Okay, fantastic. Ladies and gentlemen, if you could do me one big favor, submit your questions. We got a few minutes for question and answer. If you already see a great question, please upvote it for me. Thank you so much. You can scan that or you can access it from your app. Okay, let's look at the highest question that got the most votes so far. Let's turn our attention to the screen. How should Ethereum L2s prepare for a lean Ethereum without major architectural

changes? Okay, either one of you can take it. Yeah. >> Okay, it's a good question. Uh I see Ethereum tools as as users, right? Uh so at some point they will need to to upgrade the way they are doing things. Uh it's inevitable, but I don't think that that is a big architectural change. Uh it won't be so big. they will have to start working with this frame transaction and and change the signature >> but not much. >> It won't be big but ideally they should start as soon as they can. So as soon as the protocol or the schema is defined uh probably they will be rich and and ask to okay start upgrading this it won't be

big but they should do it. >> Sure. Yeah. Upgrades and changes are inevitable right? They always happen. So make sure that you do it quick. Okay. Next question. Also the highest voted question so far. Let's look at it together. How likely are lean protocols especially PQ features will be realistically adopted/accepted by Ethereum core developers and when? Okay. >> Okay. Uh well I believe Justin was very clear yesterday talking about the the mission 2028 uh that is already um that already has a great consensus among all the core developers. So they already on board of it and we are working for it.

>> That's good. I think everybody's on board together. We're all working together as one to continue improving our Ethereum ecosystem. Let's go to the next question at the top. Okay. How will Ethereum coordinate the transition? Oh, from beacon to lean consensus without disrupting existing valid validators and L2s. Great question. >> Okay, great question. And I I don't even have the answers. I believe this is one of the things to be defined. How do we migrate this? Where in some incent economic incentives where there'll be for for for validators, stakers and so on.

There be there will be some uh disruption I think but uh not not so heavy. The transition is not coordinated yet. It's still to be defined >> soon to be finalized. Right. I remember when we were going through the beacon chain upgrades. Everybody was scared. They were worried the whole of Ethereum is going to break. But as you know, the developers did a great job. They prepared for it. So, we trust that this time as well, they'll do the same. Let's go to our next question at the top. What does lean ETH mean for homestakers and validators and our setup? Okay, that's a good question. For home stakers and validators, what does it mean to have a lean Ethereum?

>> Um, to be honest, I I I don't have exactly the data to to answer, but that's fine. >> There shouldn't be much difference on on what you have right now. Uh on the test that we have been doing there is basically no no change. Uh so you won't have any issue like >> yeah no worries >> or hardware issues or or networking issues won't be any. >> In fact uh our next session right after this uh the next speaker Ben is actually going to be sharing a little bit more about this topic right about bringing Ethereum >> yeah yeah you stay for that session we're all going to enjoy listening to the answers from his next session. Okay let's continue. We got time for a little bit more questions. Let's come on. We

can finish it up. Is the blockchain trlemma ah being maintained with lean chain especially consensus versus decentralization? What do you guys think? >> Yeah. Yeah. Definitely. >> Definitely. Okay. Short and sweet. I love it. Cool. Last question of the day. Let's see if we can finish all of it up together. What does lean layer client mean in terms of infrastructure for hosting full node and validator clients? Would it need a more powerful machine and storage? Okay, very good question. >> Okay, it's a question is similar maybe to the home sakers from a different point of view but but similar questions.

Uh I again I I don't have exactly the the data for this but it shouldn't be much much [clears throat] change. >> Yeah, >> signatur are bigger but with all the improvements that are being made in the limbm there shouldn't be much much change of it. The idea is to have more validators. The idea is to have a change on on how the committees. We have new new kind of of committees that that are being created but machines and storage will kind of remain similar. Uh >> all right. So some changes but hopefully not too major.

>> Yeah. Yeah. But I'm not the right person to to answer this. >> Don't worry. I'm sure one of our speakers today is definitely going to expand on these topics as well. After all, we're here discussing all the topics related to the core protocol. Ladies and gentlemen, they managed to finish all their questions today. Please give our speakers another big big round of [music]