Ethereum's Roadmap, Simplified: What Changes Next and Why

Transcript

[music] What's up everyone? Thank you for saying my handle is cute. Um, not sure I would describe it that way, but I'll take it. Good morning everyone. Uh, I know it's early. I really, really appreciate you showing up uh this early. Um, yeah, thanks for being here. Thank you to DevCon for having me. Uh, I'm Cheeky Gorilla. You can call me Cheeky. Um, I work on Protocol Guild, which is a nonprofit that helps fund Ethereum's core protocol development. Uh, but today we're going to talk about something else. We're going to go over a non-technical and hopefully layman friendly breakdown of Ethereum's core protocol roadmap.

And to ensure that you guys know what we're up against, we're going to start by showing you the road map. And there it is. It's lovely, isn't it? It's really, really a great flowchart. Um, how many of you have seen this before? Does this look reasonably familiar to a few people? Fantastic. Okay. Now, what if I were to tell you that this road map right here is actually from 6 months ago at a talk that I gave at ETH Clouch. By the way, shout out to Simona, the whole ETH clue team. They're fantastic. My question is how much how many changes would you expect there to have been to the road map in the six months since I gave my presentation at ETH Clouge. So we're going to do a little bit of

audience interaction here. You have five options to choose from. By a show of hands, raise your hand if you think there were zero changes to the road map in the last 6 months. Show those hands. I see one person at least. All right. How about 1 to 25 changes? Any takers? I see maybe 10 people. Cool. 26 to 50. I see some more hands. All right. 50 to 100. Maybe four people. Okay. Is there anyone here who thinks there were more than 100? There are some of you. Okay. Well, look, the correct answer is in fact more than 100 changes in the last 6 months.

And so here are all the changes. There's 108 in total. But to be very fair, a lot of these changes are actually purely cosmetic. Around onethird of them, nevertheless, the point does remain that there's a lot of changes. Okay, so this is the road map that I showed you, which is what I presented at Cluge. And here is the new and improved road map with those 108 changes. Yeah, still a lovely flowchart. But I think I know what you're thinking. A road map that changes that often surely is not a very serious road map. But guess what? I'm here to tell to tell you that it's totally fine.

And not only is it totally fine, but I would also say that the fluid nature of Ethereum's core protocol roadmap is absolutely a feature and not a bug. And in fact, I'd argue it's the reason that Ethereum can lay claim to a 100% uptime since its inception approximately 11 years ago. And furthermore, I would also say that when it comes to the long-term road map, specifically the endgame, that has actually been virtually unchanged for years now. And I'll get to why there's a cheeky asterisk there a little bit later. But the point that I'm trying to make is if you ever feel like Ethereum's core protocol development is hard to follow, that's because it actually really, really is. Like, yeah,

sure, it's very technical, but it also moves at an incredibly rapid rate that most reasonable people could not possibly follow while also, of course, simultaneously being a lot slower than most people want it to move. But the good news is, hey, you're here with me today, so by the end of this talk, you will be up to speed. So, let's talk about what you can expect from Ethereum's core protocol development in the short, medium, and long term. And we'll start with the short term, meaning the next six months. And that's a beautiful little thing called Glamsterdam. But it turns out it's not such a little thing. And it's also not that beautiful actually. Um it's been a major major pain in the ass. Uh we originally hoped to get this out the door in the first half of this year.

That slipped. We were hoping to ship it next month in December. Um now we're hoping January sometime. But let's talk about what was actually in the or what is in the upgrade because it's a beefy one. There are two headliners. The first is something called enshrined proposer builder separation uh or EPBS. And to put this very very simply, every 12 seconds another block full of transactions is added to the Ethereum blockchain. And it is the job of so-called validators to basically propose these blocks. But the problem is that nowadays most of these validators, they don't actually build these blocks themselves. In fact, they outsource it to these specialized third-party block

builders who know how to basically create blocks in such a way that it results in the highest payout. Um, which means ordering and selecting transactions and this payout is then shared between the validator and the block builder. So what this means is there's a financial incentive for validators to use these third party block builders. And you know in principle this is fine. We should expect individuals to act in their own economic self-interest. But the fact that this is happening offchain is not so good. Right? Ethereum should be about disintermediating trust and that's not exactly happening in this case right now. Enter EPBS which enshrines everything that I've just mentioned into the core protocol itself which obviously helps reduce trust assumptions and also

reduces some dependencies on that off protocol software. But we also get another huge benefit which is essentially if we're given the chance to redo Ethereum's plumbing in this way, there are actually huge scalability unlocks by essentially giving validators more time to do more stuff. And I'll quantify exactly what I mean by that a little bit later. But for now, let's move on to the second headliner, which is something called blockle access lists. And like EPBS, this one also optimizes the inner guts of Ethereum. Uh so today nodes, they have to process transactions each in each block one by one sequentially. Um and that's

obviously slow and it's because until a transaction runs, we don't actually know what this transaction does. We don't know what parts of the protocol it touches. And so what block level access list does is it basically removes that unknown, right? It basically makes it such that each transaction has to declare or each block rather has to declare within its transactions what will read or write um before it runs. And so with these dependencies known up front, it's basically possible for nodes to pull all of that information up front uh for all the transactions and check the non-over overlapping ones in parallel. So basically unlocking parallelization which means nodes can do their stuff faster. [snorts]

Other features in Glamsterdam include some general repricings. So for example, that includes raising the cost of uh creating and reading permanent data. Uh these are things that have basically been underpriced relative to the amount of uh the cost and work involved. Uh but not all the prices are going up. Um sending ETH to an existing address will get cheaper up to 71%. Um and for developers, they will get bigger contracts. uh exchanges and bridges, they will get a uh enshrined tool to track ETH deposits without relying on custom tooling. And for validators, they get faster withdrawals, while slashed validators uh they will stop being picked to propose blocks, which means in a mass slashing event, the uh there

shouldn't be as many gaps in the chain. Now, going back to why this upgrade has been delayed. So, what happened is a few weeks back, an audit found some pretty scary stuff. And that penguin belies how scary it was. Um, basically remember those repricings we talked about in the last slide? Well, it turns out uh because of one of them, ETH can actually get stuck in some smart contracts on the app layer and in the worst case can actually get lost. Um, to be clear, nobody's funds are at risk today. But what it means is that post Glamsterdam, there's a chance that some contracts on mainets so on the app layer um that were working fine pre-G Glamsterdam no longer work so well after Glamsterdam. And so

that's obviously a problem. And so the upgrade is effectively paused while we all take a step back and say, "Hey, how many contracts are affected so that we can figure out what exactly to do about this?" Now, this is obviously a very very unfortunate situation. This should not happen. Um, but it's also a great example of what makes Ethereum's core protocol development so incredibly unique, right? Because Ethereum doesn't ship first, patch later, right? There's too much value at stake. Uh, to quote a core dev, security trumps timeliness, which means do the right thing, not the fast thing, which is slower, which can be frustrating. The trade-off is 100% uptime. So, yeah, that was a lot.

Glamspam TLDDR is very simply it is a scaling upgrade. It will allow us to increase Ethereum's GLA gas gas limit from 60 million today and more than three exit to 200 million. And so this will result in lower transaction fees until demand catches up. And this is really the whole point of Glamsterdam, right? This this enabling this kind of scale in a way that is safe rather than reckless. >> [snorts] >> So yeah, it's a huge and complex upgrade um but with a great payoff and hopefully we'll see it shipped there in January next year. So that's Glamsterdam and that's the the short term. Now let's move right along to the medium term. So the next 6 to 12 months which is a

little thing called Higota which is shaping up to be an awesome awesome upgrade. Um it is still in the planning stages. So we don't know everything that it will have but what we do know uh are the headliners and whereas Glamsterdam focused exclusively on scaling or more or less uh Hagod focuses much more on censorship resistance and privacy while also kickstarting Ethereum's postquantum migration. So let's talk about the headliners. The first one is fork choice enforced inclusion lists. Yes, that's a mouthful. The the two last words there are the hint for what this is. Inclusion lists. And it's basically a list of valid transactions that the next block has to include. And if the next block doesn't include them, well then it gets

rejected. And so what it essentially does is it introduces an economic cost for censorship. And so if you ever dabbled with tornado cash and uh you saw that some of your transactions were not being picked up right away, um Fossil prevents this. So, Fossil is very cippher punk. The second headliner is something called frame transactions, also known as frames plus when you include all the different features that kind of fall under the frames umbrella. Um, and it's a solution to something Ethereum has been chasing for a very, very long time, which is native account abstraction. Um, native account abstraction has been tried many times before. Um, you may have heard of 4337 or 7702, which both enabled account abstraction, but weren't really the

optimal end states that we were looking for. Um, frames is this optimal end state, right? It gives us everything onchain with no external dependencies. It's incredibly flexible. Uh, an account can be, you know, a multi-IG, a pass key. You can even choose to upgrade it to to use a quantum safe signature. Uh, and things like gas sponsorship, they're obviously supercharged. So, you could live an onchain life with USDC only. But make no mistake, there's still a place for ETH because all transactions still require ETH. So, just because you're using USDC, that's just being swapped in the background to ETH, which is invisible to you. So, uh, abstraction there. These kind of things also mean that privacy protocols like, uh, rail gun and privacy pools, they no longer

need so-called relayers. So basically, you can unshield funds into a fresh address uh without requiring a relayer. So that's really, really cool. Frames is overall incredible. It's awesome. It's really exciting. It's going to supercharge Ethereum's UX for sure. Um while also making privacy protocols a firstass citizen. So those are the two headliners, but what's this? A third headliner? Yes, I did say there are only two and this is not in any way confirmed, but I predict that there might be a third headliner soon. And even if it's not a headliner, uh it's still worth having its own slide for um which is postquantum registrations.

Okay. And this was the the newest change to the road map which was uh from November 2nd. Uh which is why I had to panic update all my slides. But anyways, um, if you saw Justin Drake's presentation at the opening ceremony, he talked about mission 2028, which was basically we need to accelerate the postquantum road map. We need to do it before the end of 2028 um, in response to AIdriven advancements obviously. And the first step there is to create a process for Ethereum's validators of which there are uh you know just over 800,000 today I think um for them to create their postquantum keys. And so we don't want validators to rush this. We want them to sit down fresh pot of

coffee you know take out their hardware wallets really look at the transaction make sure they know what they're signing. We don't want them to make mistakes. So the idea is we give them this ability from Hagod and then eventually later on make moving to those postquantum keys mandatory but not yet. And all and if you're a if you're a user who wants to do the same thing for your wallet um this is not PQ registrations you can actually do that with frames and then for all other PQ migrations those will happen after Higota. So yeah as you can see obviously Hagod is still early scoping stage so we don't know everything that will go in it. Um, but the stuff that we do know that's going to go in it is going to make it already an incredible fork uh or upgrade. Uh, there's currently 29 things

that are being included uh that are being considered to be included. Um, and to be honest uh probably a lot of those are going to get rejected just so that we can get Hagod out as fast as possible. And so the TLDDR is censorship resistant and privacy focused upgrade. Fossil get makes transactions unstoppable. frames gives us native account abstraction and postquantum readiness for wallets and PQ registration gives the same for validators. Uh as for when to expect it, the target is August 2027. Very ambitious. Let's see if we can do it. Now, let's move on to the long term.

The upgrades after Higod. Well, they don't have names yet, right? Um but we do have some idea of what they could include. And here's a list. But to be honest with you, for the purpose of this presentation, I don't think it's terribly productive to go over these things because a lot of them are still very much in the research stage. And the truth is that means that they're bound to change, right? So instead, I think we should take a step back and look at what is the endgame for Ethereum's core protocol development. Basically, what is it building towards? And that is fast layer 1, gigas layer 1, terra gas layer 2, postquantum layer 1, and private layer 1. What does all that mean? It means transactions being included and

confirmed within seconds. It means 10,000 transactions per second uh on mainet and 10 million transactions per second on layer 2s. It means going postquantum by the end of 2028. And it means privacy as a first class citizen. And how do we do all this? Well, there's many different ways, but in a nutshell, a cool thing called snarkification, which essentially uh instead of validators recomputing all transactions and everyone, you know, re-executing transactions and all this stuff, we get rid of all that and we say, "Okay, look, here's a proof that the block is valid." And it is very, very easy to verify that that proof is valid. You could do it from your node at home. You can even do it from a smartwatch. Um, and so snarkification uses the magic of these so-called ZK proofs to give us crazy

amounts of scale to have very very large blocks that can still be very easily verified. But what's not mentioned here is what comes after all that, which is maintenance mode, which is basically the idea that no longer do we need to ship all these incredible upgrades because it's kind of good enough. It's reached escape velocity. Um, but of course, Ethereum is built to last for centuries, so a degree of maintenance will always be required. Hence maintenance mode. So yeah, pretty exciting stuff, right? Um perhaps you're thinking, you know, six months from now, Cheeky will be back at EClusion 2027 talking about all the changes that have been made to the road map since today.

And yeah, you're partially right. I will be um there will be all sorts of new implementation details and yes, timelines to discuss. But here's the thing. It isn't Ethereum's endgame that's changing. And as proof, here's a slide from my presentation at ETH Prague from May of last year, right? So, a year and a half ago, showing the slide that I had on Ethereum's long-term core protocol development. And so, we can actually compare them side by side. And so, except for one thing, which I'll touch on in a second, these two slides are virtually identical, right?

Snarkification is the word that I use prominently in 2015, but let's be honest, no one knows what snarkification means. It's like one of these classic Ethereum gibberish words. And so I tried to anchor on some more simple words in this year's presentation. So what's the thing that's different? Maybe you spotted it. It's privacy. And including it as a first class citizen is a really, really cool change to the road map. I love this. This is why they call me cheeky gorilla. Um, but unfortunately talking about privacy is out of scope for this presentation. However, there are a ton of presentations here at DevCon that do talk about privacy. So definitely go check those out if you're interested.

Um, but going back to [snorts] the road map again, apart from explicitly mentioning privacy, the endgame has not changed. So even though yes, at the start of this presentation I was making all these jokes about all the changes that are being made to the road map and 108 haha. Uh but the truth is all those things still bring us down the same path, right? We know what we want to do. We know how we need to get there. But the actual implementation details of these things, yeah, they're still being worked out because we are not taking any risks. We are prioritizing security every single step of the way. Now, if that sounds fun to you, well then please

check out epf.wiki wiki where you can learn how to become a core dev uh and get part of become part of this madness. And if you don't want to be part of the madness, but you want to help fund it, well then check out protocolguild.org, which is, as I mentioned at the start, an entity that helps provide funding directly to Ethereum's talented core protocol contributors. And that's all that I have for you today. Thank you again so much for coming. >> [applause] >> All right. Let's hear it one more time for Cheeky Gorilla.

All right. Fantastic. All right. We've got quite a number of questions coming in. Thank you very much to everybody who's submitting both online and in person. Let's tackle a few of them together. Okay. First one, what is the road map that will never change? >> The road map that will never change probably hopefully the values because at the end of the day it is the values that drive the road map, right? uh we're not we're not where we're trying to go is based on this idea of decentralization um the fact that like you can run a node at home all these really really good things and and that's those are really the drivers and so as we encounter new forms of whatever mathematics or technology generally that enable those things there might be pivot to achieve

those values um so I think maybe that's my not real answer to this question um but hopefully the values will never change because Ethereum as it is today is not the technology it is the values. >> Exactly. Values are important just like you said I think in your closing words can't compromise on security. Right. Exactly. So the only constant in life is change after all. Let's go on to another question. I think right at the top how many people are involved in maintaining and building Ethereum's core protocol. Interesting question. >> Yeah. So as it turns out at Protocol Guild we are focused uh exactly on curating the list of Ethereum's core protocol contributors. uh we have this registry of individuals that we update

every quarter and today there are 172 these are people obviously from all over the world all of whom are actively involved in building and maintaining Ethereum as we know it today um and if you think that oh are these like all the people from the EF no EF is definitely a a part of that um but there are like I don't know 16 different independent entities who are all involved in building Ethereum and this is purely thanks to its multiclient philosophy >> yeah fantastic great answer thank you for asking as as well. Let's go on to the next one. Okay, this one looks like there might be a little bit of frustration attached to this answer, but why why is everything under the sun packed into frames? Why is authentication so tightly coupled with

account abstraction? How do you want to address this? >> So, I believe that light client, if I'm not mistaken, has a presentation on this later today. Um, and as an author of EIP, I think you should ask him. Um, my response to this would be it it doesn't it doesn't pack everything under the sun. The great thing about frames specifically is that it's kind of written in such a way or developed in such a way that it gives you this flexibility so that we don't have to uh, you know, set up guard rails for like this is how you use your account and like there's nothing else you can do. It gives accounts let's call it escape velocity so that it's flexible enough in the future to uh be able to use different kind of cryptographic

signatures including postquantum one and all these kind of things so that it's not the core devs deciding how to use the wallet but frames giving everyone the ability to decide how to use it themselves. >> Right. Great question, great answer as well. Let's keep going. We've got about two more minutes for questions. Where does formal verification reside in this road map? >> Great question. So formal verification um is a through line through everything I would say. Um formal verification is still very very nent. Um it will ultimately result in some massive changes to Ethereum. I mentioned the multiclient philosophy. If formal if formal verification lives up to its hype, there could be a word where multiclient world where multiclient

philosophy doesn't make any sense anymore and we can just have one formally verified exclusion client and consensus client. Um but again it's still very re very very nent. Obviously AI advances are allowing us to supercharge all these things. Um but it's it's an enabler of the very long-term road map. Um but again who knows what long-term means with the here in the age of AI >> in the age of AI one day is considered long-term I think. Yeah. Changes happening all the time. Let's do a few more. I think we've got still got about one minute on the clock. What's the role of rollups in the endgame?

>> Great question. Yeah. Roll-ups are part of the um Terra gas layer 2, right? So they are they are absolutely part of the endgame road map. Um we love roll-ups. The roll-up road map is not dead. Uh roll-ups is how Ethereum can distinguish itself uh and become a Swiss army knife compared to any other blockchain. Right? If you like Ethereum because of its decentralization, um well, you can use Ethereum. But if you are say an Amazon uh and you want a more permissioned blockchain, well then create a layer 2 and you can have that right. So basically layer 2s are the answer answer for if there's something that is not a good fit on mainet then create a layer two for it. Nice. I like

that. Rollups make Ethereum into a Swiss army knife. I am [snorts] going to quote you on that. The last 10 seconds why Glamsterdam release has been delayed a bit. We were expecting Q3 Q4. Last question for the day. uh because of some unexpected bugs. Um I covered this. I had a slide specifically for this. So go to YouTube, scrub back, you'll find the answer. [laughter] That's right, ladies and gentle [music]