# Coordinating Ethereum | Mario Havel (Berlin Ethereum Day, June 2026)

- Speakers: [Mario Havel](https://streameth.org/speakers/mario-havel)
- Channel: [Berlin Ethereum Meetup](https://streameth.org/berlin-ethereum-meetup)
- Date: 2026-09-09
- Duration: 20:17
- Watch: https://streameth.org/watch/yt-N-33APpGPio
- YouTube: https://www.youtube.com/watch?v=N-33APpGPio

## Description

Building a large scale FOSS protocol with zero downtime by Mario Havel (Ethereum Foundation)

The Berlin Ethereum Day was a one-day event held on June 15, 2026 during the Berlin Blockchain Week.

The full-day program brought together speakers from the Ethereum Foundation and the broader FOSS, privacy, and security ecosystems to explore the future of Ethereum and self-sovereign technologies - from technical direction and core values to the challenges and opportunities ahead.

Future Meetups and Events: https://www.meetup.com/berlin-ethereum-meetup/

More information on the speakers and the agenda: https://berlinethereumday.com/

## Transcript

So, yeah, today we are going to be coordinating Ethereum. So, Party, thank you so much for the great uh roadmap talk, and you you could see the crazy long-term roadmap there with bunch of items going through the years. So, I'm going to talk about more uh what do we actually need to do to get there, or how do we do all of this core development to be able to uh produce all this stuff at a scale and with zero downtime and completely decentralized way, hopefully. Uh so, uh I'm going to be quick. Don't have as much time, and there's a lot of a lot of things to cover, but uh just a bit of a background why we even do this, or why we have these properties that we have uh in the network itself and the way that we develop it, we coordinate development. Uh cuz everything comes from the background of the Cypherpunk dream of digital cash, right? Like some um uh some of these early attempts in '90s, uh this community trying to build actually something internet-native, something that cannot be controlled by a single entity. And uh uh it did not not to be that easy. I mean, until Bitcoin and the Nakamoto consensus, and uh Ethereum is taking this, I believe, into a completely different level uh by having not just a digital cash, but completely open and permissionless platform for all kinds of use cases, all kinds of uh uh uh contracts and programs that people can run. So, the idea is to uh build a network that is not biased towards a single use case, but has uh more uh neutral uh open uh nature to it. So, uh we, as a core developers, we don't dictate what is the use case. We try to make it uh build the basic building blocks, the primitives that people can use to then create their own standards, and so on. So, that's the way that we build Ethereum to enable more freedom, I believe, to give people uh more power uh to be, you know, their own bank, not to be worried about other banks, to be able to, whatever, uh, coordinate, uh, collaborate, uh, much more efficiently, uh, without some third parties. And, uh, that means that we need to do that on the level of development of the protocol itself. So, um, it's, um, it's not always that easy because we make the process a bit hard for us. There is no single team that would just decide like when when they say, "Oh, core developers, Ethereum core developers." That is just these three guys and Party telling them what to do, right? Like, that's how you imagine the team, but it's much more complex. Uh, there is not a single team of Ethereum core developers. It's, uh, uh, many independent teams and, all together, uh, more than 200 people, I would say, collaborating on, uh, pushing these Ethereum upgrades and maintaining the protocol. So, these are all, uh, different teams from different organizations. Some of them come from Ethereum Foundation, but most of them are actually outside from different companies, different organizations running these teams, running this development. And, um, there is no top-down control. There is no single entity that would dictate, uh, uh, what needs to be done. So, no single point of failures. Um, um, the, uh, uh, the idea is to have everything in public, basically. Again, more fast native, completely open, completely centralized. So, everything that's happening, again, like this this crazy stuff that Party was showing, there is like, uh, so many ideas what to develop and how to do it. The process is completely open in every step, in the way that the proposal is being born, it's being implemented, it's being tested. Everything is happening completely openly. There is nothing, or nothing critical, uh, behind the behind the closed doors. Uh, everything is accessible to all of us. And, um, uh, this needs to be done, of course, with uh the utmost uh precision and security and very good testing process in mind because well, the security is critical. We are playing with real money here with like a lot of money, actually. The network is running like, you know, billions, hundreds of billions of dollars. Uh less these days, but still it's a lot. And uh uh the um uh yeah, the uh the most important thing is like to keep it running. Uh to have zero downtime. We see blockchains that just get stuck at some point, right? We don't want to be that. Uh we want to provide the maximum uptime and security for our users uh on the protocol level, of course, like uh the You can see many DeFi hacks and so on, right? Because some user opted in to some contract, some protocol, and it's a responsibility of that DeFi developer, that contract to make it secure. Our responsibility as core developers is to make everything secure, right? The whole platform. So, it's a it's a big task. Um and of course, we need to do all of this in the limitations of blockchain itself. Blockchain is kind of inefficient, right? It's like uh it's like a very slow database. It's uh just it's just annoying how many scalability problems we have and so on because we adhere to these fundamental um limitations, these um these features to that actually keep the network decentralized. We don't sacrifice the lowest common denominator computer. We always uh want to let everyone to be able to verify everything that's happening in the network, and that fundamental limits us. Uh but it allows us much more the actual neutrality and decentralization. And this results in many different solutions. Uh Ethereum spans, uh you know, networking, databases, uh virtual machines, this uh cryptography, different domains of the computer science, uh which also makes it interesting, but well, challenging to coordinate experts from many different areas and find creative solutions in all these different different that need to come together. So, that's like overview of the process. Maybe to look at it a bit more specifically, uh the uh in practice, the Ethereum protocol is like uh coordination of all these different entities. Like, it's not a single one, of course. So, there are uh there are some EF researchers that is probably like the frontier of the research. Most of the these ideas that Pari was describing in the future of Ethereum, the road map, most of it, not all of it, but big part of it comes from the EF research. So, Ethereum Foundation is not uh employing uh many of the actual developers, but mostly these people who are like thinking about the future, right? Uh but also some security and testing people that are creating initiatives and harnesses to uh test the uh all of these upgrades. So, Pari again, he was one of these people uh doing the actual uh testing and the infrastructure behind uh testnets and being able to run this. Uh there is a community input uh to this process. Everybody can come and voice like what they need for their sort of use case, right? And uh for the core developers to consider it. Um um This everything is uh in public thanks to uh some standard processes and the EIP process that we have. Uh so, there's also like a community of EIP editors and Ethereum cat herders and uh this sort of community managing the the bureaucracy around it. And uh and finally, the I would say the beating heart is actually the execution and consensus clients, the actual implementation. So, when I ask you guys, "What is Ethereum?" uh you know, you might say like, "Oh, it's Ethereum is a network. It's the future." Whatever. I think the most tangible part of Ethereum are the execution and consensus clients, the nodes that you run, the actual software that you do to do Ethereum, right? That's how you do Ethereum, EL and CL. And uh uh there is many of them. The uh what do we achieve with this approach of having different independent teams is client diversity. We don't rely rely on a single client, on a single implementation, on a single team, or a single person. Not single point of failure. Uh there is a diversity. So, um the Ethereum comes in two parts. It's actually Ethereum is sort of like a Frankenstein of two networks connected. I'm not sure if you know about that. Well, that's how it works. There's a consensus layer and an execution layer. The execution layer is implemented in these and even more clients in get, Nethermind, Besu. So, there uh they're different teams. The first one is from Ethereum Foundation. The other ones are completely independent from, again, different teams, different organizations. Um and they are implemented in Go, in .NET, in Java, in C++, in Rust, right? So, many different languages. So, again, if somebody comes and wants to wants to contribute to Ethereum, uh no matter their background, they can be expert in different languages, they can come. This is Prism, Lighthouse, Teku, Nimbus, and Lodestar. Again, uh consensus clients. This is the consensus layer, the other side of Ethereum. So, we have like five plus some experimental EL client plus uh five plus some experimental CL clients. So, that's Go, Rust, Java, Nim. Uh there's also they also have actually EL client and TypeScript client with Zig uh being uh being used now as well. So, like again, many different languages. So, if there is, for example, uh bug in one of the libraries, it doesn't affect all of the clients because it's a Go library, not the Rust library, and so on. So, this diversity enables us to have more talent, more developers from uh different backgrounds. It enables us to have And none of these clients are run by Ethereum Foundation. Again, it's completely in the ecosystem. It's something that's uh uh that's out there in uh in the wild, basically. But, it means that all of these clients need to be able to talk to each other and uh and come um come in agreement, consensus what to what to ship. And that happens in this roughly sketched process. Everything starts with idea, right? All the ideas for new features, for improvements to Ethereum start as some research or something that comes up that needs to be done. The research is the crucial part where you need to analyze this and benchmark it and prototype it and figure out whether it actually work or what is the best way to to make it work and based on that you specify it. So the EIPs actually serve partially as as like some initial specs and that becomes part of the specification of the whole network. Based on that it gets implemented in all these different clients, right? Across across like I don't know 10 different pieces of of code bases need to be touched or more to be able to implement it. Then it needs to be tested. That's one of the most and very crucial part because having different clients in different languages to talk to each other perfectly, agree on everything on the consensus in in milliseconds, making sure that there is nothing going wrong. That's that's very crucial because otherwise network would go belly up. And finally, that's how we get to mainnet or you don't because you get rejected and you need to go back. So like the Ethereum process, the development process as as much as open and nice it is, sometimes you need to take a step back. Sometimes you realize in your research phase that the idea is not as good. Sometimes you realize that the specs are not going to work that way and sometimes your idea gets rejected completely and you need to start all the way from the beginning. If you guys heard about EOF for example or a bunch of things that have been worked on for years and then rejected, it happens as well. But it's important because this work also shows us maybe, you know, the paths which are not the way, not just the way, but also exploring different things. So yeah, that's pretty much how it works in practice. Like I could talk about like, oh, you draft an EIP, you go to the ACD, and I'm going to get to that, but like uh again, this whole process is open, and you can follow it by yourself. You can start following it today. Actually, maybe there is a call today, as well. Uh because uh all of these coordination calls of developers, all the discussions are happening in public. Uh and uh probably the best way to follow it is forecast. So, on the forecast.org, you can see uh all the current hard forks and their EIPs and their process. Um you can see what's uh being developed right now, what is in the next hard fork, and at the same time, you see all the current calls. You There is an overview of uh all the uh so, here are the current hard forks. You see how we progress in in Glömsdal, Dan Hegarty. If you click on it, you see all the all the EIPs that are recently updated, and there are also the current calls and their summaries. So, if you don't want to listen to the whole thing, you can just uh see a transcript. And we have a question, I think. No? Okay, why not? Uh >> You You said something before about Ethereum needs to be verifiable. So, I I do not mention uh if research would be open, etc., which is not true. I was censored by Virgil. I'm not allowed to talk about quantum social and physics and research. That's one point. And the verifiable point, when is the Ethereum Foundation itself on chain verifiable? Like, am I going to see this ever? >> So, so uh again, so first, you were censored on Eth research. That was one question. And the second was about Eth EF being verifiable? >> Yes, Eth EF itself said Ethereum needs to be verifiable. So, I'm asking myself, when is the organization itself on on chain verifiable? >> Okay, so that's a that's a uh that's a different can of worms. Uh so, yeah, let me start with So, yeah, we have different forums. There is the Ethereum research, Ethereum Magicians. So, if you were uh censored in a research, I'm very sorry about that. Uh, there is a lot of spam and a lot of things being censored because uh especially today in the age of AI, it's uh it's uh well, uh I deal with the the people coming in and sometimes I am the one now in the protocol fellowship actually uh have to have to filter through all of this. So, what I I I don't mean to say that you were spam. I'm saying sometimes it's really hard to recognize what is a genuine proposal and that's sort of that's something that I that comes later. >> Virgil, I'm not allowed to talk about >> Virgil. It was a while ago, then. >> Like two weeks ago or something. >> Since he got out of prison. >> Exactly. That's me. THAT'S EXACTLY WHAT I'M TALKING ABOUT. >> OKAY. I I'm surprised that he even has an access to this, but it's Okay, we can chat about it in in a private, I guess. Um, the um uh the E-tree research allows anybody to create posts, which means also AI bots and also many things that again needs to be filtered. Sometimes it's hard to distinguish. Um, and uh but yeah, I don't know. Like I see people right now getting to the core development. We have some people here uh from the fellowship, from like the completely fresh core developers who are creating posts there right now. And uh the I don't think that the main problem is censorship, honestly. I think the main problem there is to just ignore it. That there is so much noise and so many things happening. Again, it's another downside of this open process because there is so many EIPs, so many proposals, so many ideas. Nobody gets time to read through all of them. So, people are kind of scoped in their own domain, in their own ideas and projects, and they uh don't have time to look further than that. So, even if you have come to one of these forums and have a good idea, sadly, it can be often uh uh often just ignored. And you need to do the work of pushing it to coming to the calls, to coming to these events. This is exactly the kind of place to be to be be to uh start ideas and discussions. So, publicly, you know, this is as much public as you can get having uh the forums and uh the repos in public, but otherwise you need to uh you need to be able to push your ideas. Uh the calls, again, you can join all of them. Uh they are being coordinated in the project management repo in the ethereum/pm, where there is bunch of uh uh bunch of uh the calls being announced in the issues. And uh these calls are happening like every week. There is ton of them. Uh so, first of all, the ACD is the main one. All core devs, as the name says, all core devs are coming together to talk about execution about consensus bi-weekly. So, one week is execution, the next week is consensus. Um and uh that's like the main call every Thursday. Uh so, you can just watch a stream on YouTube, or you can actually or you can watch the recording, read the summary. And if you have something to contribute, if you want to share your idea, you can actually uh just uh join the call. Um the other calls are also breakouts on specific topics. So, if we have like some specific EIP or implementation problem, that's something that has dedicated calls throughout the week as well. And um yeah, the uh the all together we have like five plus calls weekly, but mostly like one or two are the most important ones that you want to follow. Uh the uh rough consensus is the way that we actually uh deal with these uh ideas. So, like when we have uh uh some proposal, some research post that that that gets turned into an EIP, and that gets presented on the ACD, and the ACD now deliberates, and they argue about it back and forth, sometimes for short time, sometimes for weeks. Uh but uh it takes so long because sort of everybody needs to be on board, or nobody needs to be against it. We need to have this rough consensus. So, like everybody needs to be in sort of agreement to do this. There is no voting mechanism. There is no like, you know, we count people pro or against, but there needs to be some form of consensus. This is how it looks like. You can join the call and sometimes we even meet in person again like at these events. You can meet the core developers even here and share your ideas. And I believe it has many benefits the way that we do this. We bring people from all over the world. So right now we are starting core seven of protocol full ship which is maybe more structured way how to get people to become core developers. And we have people from everywhere, from Asia, from Africa, from India and couple from Europe and US, but like all over the world basically. So you get the best people or the people who are interested in Ethereum from everywhere, right? And this lets you to um have more diverse background, to be more neutral, but also iterate faster and have like more more knowledge compounded in the ecosystem. People with different expertise being able to wait on like these multi-disciplinary problems. So I think it's it's amazing that we have this diverse and strong core community. There is no other blockchain that would have this. I think the similar projects would be large scale like Linux kernel sort of project, right? Like these sort of uh these sort of large scale free open source project, but it's also a problem. It's also again I will I already mentioned a bunch of stuff here like spam that's coming or bunch of like things that people need to deal with or just miscommunication. And even the client diversity, it's a strength, but it also poses a lot of overhead because testing, let's say we have five or six clients, it's like it should be more actually, but like that's 25, 30, 36 combinations of clients that you need to test. We are talking about number of combinations that your laptop needs to run if you want to test the whole network, right? At least like one client of each. Uh then um uh the uh the rough consensus as nice as idyllic it is, it doesn't always work. As I mentioned, sometimes it takes weeks to deliberate something. Sometimes after months of back and forth and trying to prioritize the features, uh we don't agree on the priorities and even some interesting features get scrapped after some feedback. Uh so like there are many examples of that. And uh uh yeah, all of this uh being done in this like a little bit high-stress environment, like very critical environment where we need the security at scale. We need to be able to test everything very thoroughly. I'm speaking for 20 minutes already and it should be 10 minutes. So I think I'm Yeah, I think I'm all over. So I'm going to I'm going to wrap it up here. But if we if you want to chat, please stop me uh later. I'm here around if you are interested in the core development process. I'm here for you. If you're coming to Prague, we have a hacker space there at Bottle WTF you can visit. And thank you so much. I'm from Vertical Support. Uh sorry about taking so long, guys. Yeah. >> Thank you very