The Road to Ethereum 2.0 - MacKenzie Sigalos, Cayman Nava, Alex Stokes and Preston Van Loon

and today what we're talking about is ethereum's migration from a proof-of-work mining model to a consensus mechanism known as proof of stake now this is an effort that is years in the making and it is set to take effect in either q2 or q3 now here to help me break down what that transition will look like our three experts i'm gonna let them introduce themselves uh go for it guys hey everyone i'm alex stokes i am with the ethereum foundation uh doing some research but also like prototyping and various development efforts to get all this research that we'll probably talk about with the history of the merge and things uh into a place where like client teams for example with cayman and preston can implement things great hey guys my name is cayman nava i'm with chain safe i'm working on a consensus client for ethereum and uh working to get the merge done this year so great and preston hey guys i'm uh preston van loon i work at prismatic labs and we're also building uh consensus layer client software for ethereum cool okay let's get right into it so a couple weeks ago the ethereum foundation said hey we're not calling it eth1 eth2 there's no ethereum 2.0 i wanted to know what is behind that rebrand and has everyone adopted that so who wants to kick us off i can give a short little overview because so yeah i mean i kind of like you hinted there's like a lot of history here there's like been a mini mini year effort in terms of research and also development to get to where we are today and i would say this like eth1 eth2 thing came from a place where you know maybe i don't know even three years ago it was like there's a constellation of things you want to do there's still like very sort of forward-looking things we want to do um and they're in various sort of degrees of sort let's say production like in terms of how developed they are going from just like random ideas to like actually things you could like put into production clients so anyway there's this bucket of things of eth2 at the time you know sharding proof of stake these things would fall into that bucket and then it just became clear over time that uh you know we would actually have this sort of merge event and then when you have that it doesn't make sense to be like oh here's eth1 here's e2 um there's one ethereum there's a consensus layer there's an execution layer they're different layers of the stack but it's all the sort of the same protocol so i think it helps people especially like people are entering this space now to not have this like very confusing like what's eth 1.0 versus e 2.0 uh yeah okay and preston yeah i think a little bit differently like uh i do agree that there isn't two you know things like e1.082.

0 but e 2.0 it's this effort to upgrade and really change the way ethereum produces blocks and and does consensus and with that i still think it's valuable to understand that this is a totally new version of ethereum that operates completely different from the way ethereum works but but alex is right we don't have eth1 clients and e2 clients and and they're not different they're actually all working together to uh operate in this system that i like to think of it as e2 i think it's you know important to communicate that it's very it's it's a very different paradigm from ethon what we have today okay so and one quick comment i will say you know the the sort of word eat to can be helpful in like technical conversations you know in terms of like referring to very specific things so like for example amongst us it like means something specific so you know there there's a place in time okay it's easier to say too it is it's catchy consensus layer is a mouthful okay so the migration from proof of work to proof of stake it's not going to happen overnight so walk us through what that transition looks like are we talking about like a couple weeks a couple months and yeah well yeah do you want to hop in uh yeah i'll i can kind of get into it um so just the whole the whole kind of idea of how how we're going to bring proof of stake to proof of to ethereum um had been a very kind of um just a hard problem and over the past year uh we got together all the all the core devs got together and we basically started hashing out exactly what this looked like and from i'd say q3 through now we've we've basically hardened kind of the idea um through kind of the prototyping stage all the way to the point where we have several test nets um where we're actually deploying and testing this process of switching um in in a real live network so what is what it's going to look like we're kind of testing that out right now and we have actual software that's that's doing that so we're feeling pretty confident about it yeah well i would say like i think we have a pretty good idea like we've had long-running dev-nets for a while uh there's different sort of sprints if you want to call them that uh we just had one for consugi is like you know labeled for it we have one now for kiln um and these basically are referring to iterations of the specs and maybe it'll be helpful just to like overview that process briefly [Music] you know because we have multiple clients on the network they all need to agree and execute the same thing the way that we coordinate that in ethereum is we have specifications so like when i keep saying specs this is what i'm referring to it's like a very uh very precise description of what the what the software should do um and that's i think like well that's definitely like a huge key piece of like you know when merge is like are those specs in a solid place and definitely in like the last couple of months they've gone to a place where like there's very minimal changes uh i pretty much consider them done and uh yeah so then from there you just want to build this stuff out and actually see it run and run in these testing environments which is where we're at right now and the beacon chain has been has been running since december of 2020 and so are we expecting to have like two parallel parallel chains for a while like i mean what is that what does that look like in practice how long like how many months does this take to get everybody to the new chain um yeah well so the way it works right is that we will have a proof work chain and a proof of stake chain and so we have like the main net proof of stake chain with the beacon chain and then uh they'll be well they're also proof of stake test nets there are also proof of work test nets and i guess when i when i'm talking about these testing environments is what i mean is that we can basically do some like you know dry runs ahead of actual mainnet merge uh and those will happen you know those are ongoing and they'll be happening you know with sort of more seriousness in the coming months like one to two months like soon tm well and the entire idea of the merge is that we're we've taken the beacon chain software we've taken like the eth1 software geth type software and the solution that we have is basically they're both going to be running alongside each other but kind of it talking together um so from kind of users of ethereum like not much is actually going to be changing just under the hood now geth is talking or now you know the execution layer is talking to the beacon chain and vice versa okay now e2 is coming as practically a meme at this point so when is it coming when it's ready i mean with this kind of thing like ethereum is running it's this massive ecosystem and we have to make sure it's done right so these these things take a lot of time the security we got to think about and with all these test scenarios we're trying to you know we've got the spec and we know the happy past and we even know some of the bad things that could happen and we want to make sure that when those things do happen they happen the right way but as alex said and cayman said we're really close you know we're in the test net stages now we're ironing out the bugs and we're getting really close to having it done uh how can devs prepare for it like what are the what are the main challenges you're anticipating well i think that um there's a lot of different developers in the ethereum ecosystem so if you're like a solidity developer or or someone is using dapps in the ecosystem there's not really much of a difference in in the way that you operate it's really about how infrastructure operators um participate in ethereum so if you run your own node which i encourage you to do it's really great helpful for the network uh you're gonna have to upgrade your software and then run a new piece of software so ethereum one or now the execution layer um that's what everyone's been running today and now you're going to have to start running uh the consensus layer client alongside so you'll have to run two pieces of software in parallel they'll be tightly coupled and that's the new way to do it so it really only affects infrastructure uh let's talk about the difficulty time bomb for a minute we keep pushing that out is that like the hard deadline like will that be what finally pushes everybody into this like to make the migration so it just so happens that the one we have right now is pretty uh auspiciously timed because i think most people involved you know with developing this stuff would say you know yeah like you said q3 or q2 timeline for the merge and that's when the next difficulty bomb will start going off right and it'll increase and increase uh how are we going to pull data does it change and if so how does it change yeah so um i guess like preston was saying like the the intention of the the sort of design of the merge is to like have as little change as possible so yeah if you're a developer sort of about the application layer there should be very little change if you're an infrastructure provider running nodes there will be some new things you need to do but also it shouldn't be you know it's not an ask beyond what you're really already doing um yeah and then there there's different uh i mean that's kind of a broad question but there's a lot of different ways that uh data will be available uh thanks for bearing with me through that question that was pretty broad um okay let's talk about what this enables like what's next like once this goes into effect like what is possible right so the merge definitely sets the stage for all of the future ethereum developments i'd say like right now there's definitely there's like two fires in the ethereum house one is this proof of work the proof of stake switch and everyone is focused on that right now but i think the next phase is going to be scalability and making sure that that this can scale to like global commerce anything else alex preston i i think cayman said it well i i don't have much to add to that yeah okay so when you talk about proof of stakes sometimes people say like hey you got the problem of centralization now i assume that we haven't been building a system that you know whales could take over so how do you work around that how do you brace for that any thoughts there are a number of ways um sort of one obvious option is like the existence of staking pools so what that will do is lower the barrier to entry in terms of capital requirements uh instead of needing a 32-week deposit for a validator you know you and 30 friends could put together one and some change eth and and run this thing um then yeah again this is we could have the rest we could spend the rest of the time talking about this topic in particular but um essentially you know proof of stake does not necessarily lead to increased centralization um there's a number of ways that it sort of you know works against that especially in ethereum um one thing for example is that the return you get on your sort of you know your stake as a validator is constant regardless of your size another one is that we have incentives and sort of you know designs in the protocol that actually uh incentivize uh anti-correlation meaning like there's actually there are incentives in the protocol such that you shouldn't centralize in various ways which again that's a very big word um but yeah so bottom line is it's uh when people say that it's pretty much fun so why did it take so long to happen like was it harder than anticipated was i mean did challenges come up along the way ooh so a big thing is that ethereum you know like a core value of the ecosystem is like working in public and open source and all this and so what that means is that like you like anyone can like literally see quote how the sausage is made like it's all transparent on the internet like and that's the thing where i think people have just been following us so long like there's so much excitement around what we're doing with ethereum and like the potential that it's like very exciting and engaging and then you like dig in and you're like well wait okay there's like so much you know it seems like noise and like part of that is just like you know how we aggregate is again as a community like we do have this ethos of decentralization and so sometimes that can look chaotic um but again like we do it for a reason the reason is that you know it leads to like a more robust community over time any other thoughts on that yeah research just takes time you know and and like alex is saying the it's all open so more having more contributors saying we get better opinions and many times over the over the course of the years of developing this software we've had to almost start over from scratch at prismatic labs we were you know one of the first people to get into it and we have i think rewritten our whole code base at least three times and that's just sort of the way it goes with this kind of technology where once it's deployed out there it's really difficult to change and it's really difficult to ask operators to upgrade their software so we try the best we can to get it right the first time of course you know we want to upgrade and change things as things uh as time goes on but it's it's research and it's really important to get it right yeah now there were incremental steps taken in like the last nine months i think about the london hard fork in august altair in october are there other like smaller incremental updates that'll happen before the bigger transition takes effect or anything to look out for yeah i don't see any i mean definitely everyone right now is very focused on the merge um you know for some reason the merge wouldn't happen but for the difficulty bomb starts going off then like we'd have to consider a fork to adjust that uh but yeah i think already we're considering you know what happens after the merge and planning for that fork we could name shanghai you know vitalik tweeted a couple weeks ago that we should be careful about bridging between chains we saw a wormhole happen like do you think our feature is multi-chain are any thoughts on that so i think uh the current road map for ethereum we we're going to be seeing a lot of competition and collaboration between uh a lot of l2s on ethereum and so in some way that's you know you can think of them as their own separate chains so i think in that sense there's definitely uh it's going to be interesting to see a lot of the experimentation around that come into the real world so with arbitrarily you know optimism and even really kind of experimental optimistic roll-ups like fuel labs and um it'll be really interesting to see right and so you know like this is all just trade-offs and so there's always going to be applications where it'll make sense to consider having like a separate chain right like you can be on ethereum where it's almost general purpose and like all the activity is or you could be on a different chain and then you know you lose sort of this network effect but what you gain is like the ability to like precisely sculpt your protocol to whatever your application is um you know i think there will always be like mini chains like there's no reason there wouldn't be but uh a big question will be security uh a big question you said is security exactly yeah well yeah i think you did you mentioned the normal hack yeah yeah yeah so i mean that's a great example where like there was this bridge between salon and ethereum and like there's a bug um part of this is that it's super early and just terms of the whole space right like there will be bugs the software will be hardened over time but it's still kind of the wild west um and yeah because of the way these things work like if you have sort of the security of the proof of stake chain with the beacon chain right that doesn't necessarily transfer over to a separate domain you know vis-a-vis some other chain uh and so until we sort of like harden these bridges uh it's yeah it's it's always going to be uh there'll always be some risk um so you you know you see articles talking about the so-called the the you know the rival blockchains that are perhaps like faster or cheaper there are obviously trade-offs that come with that so do you have any thoughts on like a solana as a competitor and how the migration might change that dynamic or what does that express i don't know i mean i have plenty of thoughts um i don't know so like one thing here is like when people ask me these sorts of questions i think they view this as almost like they view blockchains as like products that are competing in almost like the web two cents if we'll pull that terminology in um and that's not what we're doing here like these are protocols uh and that actually means something like pretty pretty distinct and different um i yeah that's all i'll say for that i mean generally i want to view you know things as positive somewhere i can and like to extend that we can collaborate with like other ecosystems that's great um for example there's a number of like research that we've done between ethereum and like falcoin uh and and [Music] others i'm blanking on but basically uh yeah i think there's a lot of uh we all realize that you know and broadly in crypto we're all working towards like shared shared goals so it makes sense to collaborate let's talk about the environmental impact of proof of stake versus proof of work yeah i mean proof of work is really expensive computational right what the way it works is that we have machines all around the world that are computing hashes and trying to find this solve this puzzle basically and it's really energy inexpensive so when switching to proof of stake where it's just a fraction of the amount of energy you know almost to the degree you could run it on like a cell phone or something of that caliber uh it's it's going to have an enormous impact on reduction of the environmental impact that ethereum has right you're going from data centers full of machines to a raspberry pi basically anything else you wanna i don't think everything to add i mean definitely yeah like uh it's just sort of like a you know paradigm shift to use that phrase going from paperwork to proof of stake in terms of energy consumption and um yeah i mean i think definitely broadly with the ethereum community we all think this is like an important thing to pursue so it definitely motivates uh motivates the merge and you know would buy us us towards doing something sooner or later as as we're doing i mean presumably by like denver 2023 this will gone through will be smooth sailing so what is the hottest topic going to be next year at the same conference with respect to the protocol probably sharding uh this is like the next big upgrade we'd have planned so uh yeah there there's sort of this trifecta of things we can think about in terms of improving the ethereum protocol uh security sustainability and scalability and the move to proof of stake gives us kind of the first two and then uh the next big one is like scaling the actual chain so you want to you want to have sort of larger capacity at the base layer and charting is sort of this high-level bucket you know sort of group of techniques to achieve this hot topic 2023 yeah i mean sharding would definitely be the big topic and then adding more layer twos like so the next level of scaling for ethereum is in layer two uh a lot of the people on the research side and deep in the protocol we've kind of agreed that ethereum as we know it today is going to be more of a settlement layer and all of the user activities can be moving to layer twos so with design patterns like sharding we have more capacity for even more layer two solutions so i think that's what we're really going to be talking about and thinking about it is how how we're scaling with layer two next year cayman anything else will just add in uh yeah one thing about um that we haven't really touched on the merge is uh kind of the monetary side which i think is really interesting um just from thinking about like the financial bandwidth that ethereum allows right now um so uh we're actually going to be cutting issuance uh when we when we do the merge so we're really overpaying miners right now or overpaying for security we're paying miners and when we get rid of these miners we're going to be reducing issuance um kind of almost like a like a bitcoin half inning of sorts but kind of like three happenings at once like a triple half inning and so that's um that's just gonna be really interesting thing to see play out like i don't know what that what that's gonna look like but um something people should know yeah with that ethereum actually becomes deflationary on issuance so it's really great for the protocol we've been issuing overpaying for security for so long uh and now with these new mechanisms we'll see deflationary so there will be we'll see at some point i think next year will or after the merge rather we'll know what the true max cap of ethereum will be because at this time it's unbounded but after the merge we'll know have you guys been talking also mev one more thing we'll be talking about meb wait stay again we'll all be talking about mev um are you have you guys talked to any of the ethereum proof of work miners like are they just all starting to migrate to mining other things are they getting into the proof of stake model like what what are they doing yeah i mean it's interesting right because basically the proposition is hey you've had this like line of revenue for so many years and we're gonna suddenly get rid of it um that being said like this is this is not a surprise maybe it's taken far longer than anyone wanted but um you know from the very early days like this has been a core part of the ethereum vision to have for this this you know move to proof of stake so i think to that extent it like de-risk a lot in terms of like miners being upset or something um a lot of them are like yeah like we'll mine other things get rid of our mining hardware figure something out in the meantime they've already started spinning up staking arms and uh yeah they're essentially prepared for their transition okay uh we have like a couple minutes left are we taking audience questions or do you guys up yeah there is a mic if you have anything please come up all right we got somebody running to the mic okay is this on okay so my question is about client diversity in ethereum 2.0 so for example i've been running my prison node and i recently switched to nimbus so sorry preston so i think we many of us believe that client diversity is it's one of the core strengths you know of a theorem of the bitcoin chain so my question is how do you think about this off-chain beacon the chain governance you know how can we develop this kind of mechanisms to prevent maybe a consensus bug what would it happen if there would be a consensus bug on prism for example you know so i guess that's my question client diversity how do we see more balancing between clients you know like controlling the bitcoin chain and talks of thoughts of uh bank charting as well if you have any so on the client diversity thing i think education is really going to be very important like letting people know what the stakes are when they join a client that has a super majority and like what would actually happen in that case so i think i think people will kind of naturally just you know spread out to these different clients i think each client kind of has their own niche uh and their own and their own value to to give and for us i mean chain safe is not going to be it's not going to stop supporting lodestar anytime soon so you know you can try this out we're kind of kind of the new newcomers on the block so okay so prismatic labs has the prism client and and the data is showing that prism is a super majority uh client and from prismatic lab side we always wanted people to use our client but never to this degree that we agree this is dangerous and with client diversity i think this is like mostly a social communication and what we need to communicate to people uh is like what are the barriers to switching and i think a lot of people are realizing the risk of switching is is inherent and they have to understand that the risk of not switching may be greater right so prism does a great job of stopping the bugs but we're human right so it could happen and understanding the consequences of a critical bug in prism should be enough for for the average user to say okay well i'm going to switch to nimbus or lodestar or lighthouse or techu any of these other great options i'll go ahead and tell you that prism it operates just as great as any of these other clients there is no best client you can switch it's easy to do and it it's really important for client diversity um i think as far as on the technical side this is really difficult to enforce we we don't know who the actors are we don't know what software they're running and it's it's arguably impossible to do that in in a trustless way so it has to be through the social communication layer so if you hear it now and you're running prism please look at looking at other clients okay next audience question we've got four minutes we'll see if we can get through all three go hi thank you for taking questions and this panel has been fantastic so thank you i'm interested in staking pools i think that lots of folks are looking for liquid staking options and the natural kind of first step would might be an exchange but a lot of us are concerned that maybe the exchanges will accumulate a lot of power if that's the case that people stick to exchanges what in your mind does a strong staking pool look like that uh lets people participate in liquid staking but kind of maintains the security and the vision of ethereum yeah i mean i think there's definitely a lot of interest in uh more decentralized taking pools right you know especially ones where you retain custody so not going to like coinbase or kraken or something like this um you know some big players right now there's lydo which uh is uh oh yeah there's lydo and then also like rock up will be another one um yeah it's certainly important uh there's a ton of custody risk if you like go to one of these centralized exchanges um i will say there's also a research effort in terms of uh what we'll call distributed validators where you basically can like uh maybe this is overloaded but you can basically shard validators across like multiple actors uh and this is another way to think about pooling um that research is like coming to production still but uh hopefully it unlocks more of sort of this design space thank you come on up three minutes all right uh i'll keep it short and sweet uh what's happening with ewasm i hope that's relevant for this discussion i'm curious you said what's happening with ewasm yeah so uh everyone's using the evm i don't know uh there was a lot of research that went into ewasm and essentially what happened was uh we saw that it i i mean yeah essentially it doesn't gain us that much beyond like just investing in evm i mean at this point there's so much tooling built on top of the evm and even like last year or two right we now have other chains that are also adopting evm so it's like becoming this like very enshrined thing um that's not to say that we shouldn't have experimentation with different virtual machines right and this is again where like rollups come into play and it's very exciting to see different extrusion models there uh like uh arbor trim for example is like moving to webassembly as like their uh you know their their l2vm um so yeah the future is layered too okay last question so as we move to as we go from you know pow to more of a pos obviously there's a lot of compute need you mentioned data centers now as it becomes more distributed is there's still a need for compute at the edge or does that go to like a person's home on the node or like how does what is the distributed compute need as we move forward i know that there's already kind of a culture of people homestaking so if you know on ethereum there's a each staker uh subreddit um they've got they basically try to get people to just you know be comfortable enough and be educated enough to be able to run their own note at home and it's definitely possible you can do it on a kind of a normal consumer computer thank you three so much for your insight on what we can look forward to this year thank you everyone for your questions and your time see ya