Summary of Proposal
Ethereum is fragmented, technically and socially. 70+ live rollups behave like separate countries, and nobody can say which teams are alive, where to deploy, or what is stranded inside the projects that quietly stopped. The Fragmentation Hub maps that reality in public and makes the case for coordination without uniformity.
Motivation and Rationale
The problem we want to put in one room
Every serious conversation in Ethereum right now runs into the same wall. More than 70 live L2s secure tens of billions of dollars while behaving like separate countries, each with its own liquidity pool, its own bridge and its own deployment of the same contract. On 29 March at EthCC in Cannes, Gnosis, Zisk and the Ethereum Foundation launched the Ethereum Economic Zone on the argument that Ethereum does not have a scaling problem, it has a fragmentation problem. The question of whether Ethereum converges or keeps splitting is the defining structural argument of this moment.
We want to widen it beyond the technical layer, because the fragmentation that hurts builders most is social. A developer arriving in Ethereum today cannot answer basic questions: which teams are actually alive, who is working on this problem already, where should this be deployed, who maintains the library this depends on. Projects launch, raise, ship and then quietly stop, and the users, the liquidity and the code do not stop with them. There is a growing graveyard behind Ethereum that nobody has catalogued, and its existence changes how responsibly people should be building.
Why a hub and not a talk
Fragmentation cannot be settled from a stage, because it is a live disagreement rather than a body of knowledge. The competing answers, from synchronous composability to intents, shared sequencing, based and native rollups and chain abstraction, are all in flight, and their proponents rarely sit in the same room for longer than a panel. A four-day space lets those people argue in front of an audience that can interrupt, and lets a builder come back on day three with a real design question rather than raising a hand at a panel session.
It also lets us build something that takes four days to build. Two of our formats are physical artifacts that only exist because attendees keep adding to them across the whole conference.
How this supports Ethereum’s core properties
Fragmentation is the thing standing between CROPS and compounding. Privacy that works on one rollup is not privacy for Ethereum. Censorship resistance routed through a single bridge or a single sequencer is a chokepoint wearing the language of decentralisation. Security spending split across 70 islands buys less safety than the same spending on one. Open source maintenance duplicated twenty times is wasted maintenance. Getting the rails right is what allows a censorship-resistant, private, secure, free and open source tool to be built once and to work everywhere. That is why we think a hub on fragmentation belongs in a year devoted to going back to the roots.
India is one of the largest remittance economies in the world. Money moving between family and friends across borders is, for millions of people, the most consequential financial transaction they make, and the existing rails are slow, expensive and opaque. Ethereum was built to make exactly this kind of value transfer accessible. But a remittance corridor that only works if both sides happen to be on the same rollup is not a solution, it is a demo. Fragmentation is what stands between Ethereum and the version of this that feels no different from the tools people already use, and better. We want that conversation held in Mumbai with people who actually send and receive this money, not described about them from a stage.
Implementation
The standard Community Hub package covers almost everything we need. We would use it as follows.
- Two cork boards. One becomes the Fragmentation Map, the other the Graveyard. Both are described under Programming and both run for all four days.
- Two whiteboards. Architecture sketching during the design clinics, and the running question wall.
- LED screen. Live ecosystem dashboards during open hours, presentations during sessions, and short recorded interviews with India-based builders on what fragmentation costs them in practice.
- Worktable and seating. Coworking and office hours. Builders should be able to sit down and work in the hub between sessions, which is how most of the useful conversations actually start.
Two asks beyond the standard package
Both subject to what production can accommodate.
- Two wireless microphones. Our debate format does not work if the room cannot hear the two people arguing, and hub spaces are noisy.
- A low riser or soapbox, roughly a single step. Not a stage. Enough that whoever has the floor is visibly holding it. If this is not possible we will run the same format at floor level.
What we source and bring ourselves
- Printed board materials, including the starting set of map and graveyard cards, so that the installations are populated from the first morning rather than starting empty.
- “Fragments”, a glass mosaic workshop. A running hands-on table where attendees set individual pieces into a single shared panel over four days. It is the hub’s argument in physical form: the pieces stay distinct, the whole still holds. It also gives people something to do with their hands while they talk, which is when they say the interesting things.
Programming
Two things that run for four days
The Fragmentation Map. A physical map of who is actually building what, and where. Attendees add their team, their chain, their standard, their unanswered dependency. It starts partly seeded and is filled by the room. At the end of Devcon we photograph it, digitise it and publish it openly, so the hub leaves behind an artifact rather than a memory.
The Graveyard. The other board catalogues what Ethereum has already abandoned: dead chains, sunset bridges, unmaintained repositories, and what is still stuck inside them. It is uncomfortable and it is honest, and we expect it to start more arguments than any session we could schedule.
Daily themes
Day 1, Map It. How fragmented is Ethereum actually, in data rather than in vibes. Liquidity, deployments, standards, teams. The map board opens and the room fills it.
Day 2, The Competing Answers. Synchronous composability and the Ethereum Economic Zone, intents and ERC-7683, shared sequencing, based and native rollups, Superchain and Agglayer interop, chain abstraction. One slot each. A proponent presents, the room interrogates, and no winner is declared by us.
Day 3, The Graveyard. Stranded liquidity, unmaintained code, sunset bridges, exit rights, and what responsible sunsetting of a chain or protocol would actually look like. This is the day we think nobody else at Devcon 8 is programming.
Day 4, Build for Greater Coordination. Design clinics for builders on how not to become another island, followed by the closing debate.
Daily rhythm, 9:00 to 18:00
| Time | Block |
|---|---|
| 09:00 - 10:00 | Open hours, coffee, map and graveyard check-in |
| 10:00 - 11:30 | Anchor session for the day’s theme |
| 11:30 - 13:00 | Roundtable, multiple communities on one question |
| 13:00 - 14:00 | Open coworking and office hours |
| 14:00 - 15:30 | Second anchor session or design clinic |
| 15:30 - 16:30 | The Debate Fight |
| 16:30 - 18:00 | Open hours, board contributions, informal discussion |
The mosaic table and henna station run through all open hours, staffed alongside the boards.
The Debate Fight
Our signature format, run once a day and proposed as our Devcon stage session. A stated motion, two speakers per side, fifteen minutes, and a room vote taken before and after. The side that moves the most people wins, which rewards persuading opponents rather than performing for allies. It is deliberately playful, and it is the format we have found that gets a room of builders to say what they actually think about a contested question.
Motions we would open with:
- Ethereum should stop shipping new layer twos until the existing ones can talk to each other.
- Fragmentation is a feature. Convergence is central planning.
- A rollup you cannot exit is not an Ethereum rollup.
- Sunsetting a chain responsibly is more valuable work than launching a new one.
Possible speakers
We list them as people whose work defines this question and, in most cases, whom one of our two communities has already hosted. We have deliberately weighted the list towards protocol researchers over ecosystem representatives, because a hub full of chain advocates would defeat its own purpose.
Ethereum Foundation research and coordination. This is where we would concentrate the programme, because the Foundation is currently the only body treating fragmentation as a coordination problem rather than a product opportunity, and its own work is spread across teams that rarely present together.
- Yoav Weiss, Account Abstraction team, on the Ethereum Interop Layer and the wallet-centric approach to making the L2 set feel like one chain. The single most on-topic invitation available to this hub.
- Julian Ma, Josh Rudolf and Josh Stark, authors of the March 2026 post on how L1 and L2s can build the strongest possible Ethereum, which is the Foundation’s clearest public position on fragmentation and the first revision of that model in roughly five years. We would want at least one of them to present it and take questions on it.
- Justin Drake
- Ansgar Dietrichs, on based rollups, preconfirmations and where sequencing centralises.
- Barnabé Monnot, on the protocol economics of a fragmented execution landscape.
- Toni Wahrstätter, on L2 data, fees and what the numbers actually say about where activity sits.
- Francesco D’Amato and James Smith, both of whom have spoken at our conference, on consensus and settlement.
- Mario Havel, an ETHPrague regular, on client diversity and FOSS maintenance as its own species of fragmentation.
- Gopi Mehta, TUM Blockchain Club, on the developer cost of fragmentation: building the same application across multiple Ethereum L2s and measuring where compatibility ends and chain-specific work begins.
Independent measurement, which is what keeps the hub honest. The Foundation’s own position commits it to working closely with the people who monitor and validate L2 security properties, and we want that evidence in the room rather than each team’s self-description. Sergey Shemyakov and Sebastian Kugler of L2BEAT, both of whom have spoken at our conference, alongside Bartek Kiepuszewski; and Lorenz Lehmann of growthepie on cross-rollup activity data. Their work would seed the Fragmentation Map on day one.
Proponents of specific approaches, one slot each on Day 2. Friederike Ernst and Jordi Baylina, both of whom spoke at our 2025 conference and who initiated the Ethereum Economic Zone; Benedikt BĂĽnz on shared sequencing; and Marek Olszewski, who spoke at ETHPrague and whose team took an independent L1 through a full migration into an Ethereum L2, which makes him one of very few people who can speak about convergence from experience rather than from a roadmap.
Deliberate opposition, because a debate needs one. Péter Szilágyi, who spoke at ETHPrague 2026 and has been among the most consistent critics of the rollup-centric roadmap, and Jeff Burdges, whose speaker biography at our own conference read “Trouncing rollups”. We would rather host the strongest available argument against our own framing than run four days without one.
We are keeping programming slots open for India-based contributors and for speakers surfaced through this thread during the review period. If you work on this and want a slot, reply here.
Cross-Hub Collaboration
This year’s RFP asks for stronger bridges between hubs. We are unusually well placed to build them, because fragmentation is not a vertical slice of Ethereum. It is the layer underneath everyone else’s topic, and it degrades all of them.
We would rather give away hours of our programme than run four days of our own voice. Looking at what has already been proposed in this category, the concrete joint sessions we would want:
- Standards fragmentation and competing ERCs, with the EIP Hub. Where does duplication across ERCs come from, and what does the process do about it.
- The networking layer across rollups, with the P2P Networking Hub. Fragmentation below the execution layer, measured rather than asserted.
- Cross-domain extraction, with the MEV & Fair Ordering Hub. What shared sequencing fixes and what it centralises.
- Why single-rollup privacy is not privacy for Ethereum, with the Privacy Hub.
- Security budget split across twenty islands, with the Security Hub, and exit rights and sovereignty with the Sovereign Infrastructure Hub.
- Duplicated maintenance as a species of fragmentation, with the Open Source Hub.
We will approach selected hubs during onboarding and shape our schedule around what they can commit to. If any of the above are selected and want a slot in our space, take one. The Debate Fight in particular is built to host other people’s disagreements, not only ours.
Audience
- Builders already shipping on Ethereum who want to understand where the ecosystem is converging and adapt their product to it rather than be stranded by it.
- New builders who are about to make deployment and architecture decisions, and who should make them with fragmentation in mind rather than discovering the cost eighteen months later.
- Attendees outside protocol circles, including application developers, students, researchers and non-technical contributors, who want a legible picture of where Ethereum is heading and currently have nowhere to get one.
We keep both depths served by design: the map and the graveyard are readable by anyone who walks in, the anchor sessions and clinics go as deep as the room can take, and the debates work for both because a good argument does not require prerequisites.
On India. Our team is based in Munich and Prague, and we want to be straightforward that we are not a local community. Our former president has lived in Mumbai for many years and will lead our outreach to local communities in the run-up and on the ground. Beyond that, we would rather commit to open slots and genuine collaboration than overstate connections we do not yet have. We will use this thread during the review period, and the onboarding calls if selected, to find India-based contributors and co-hosts, and we will hand over programming time to them rather than merely inviting them to attend.
Team
The communities
TUM Blockchain Club e.V. is the student organisation responsible for blockchain and Web3 education in Munich and the organiser of the TUM Blockchain Conference, the world’s largest student-organised blockchain conference. The 2025 edition brought together more than 1,100 attendees and over 125 speakers across 57 talks, 12 panels, 4 workshops, an ideathon and a researchathon co-hosted with the Ethereum Academic Secretariat. The Ethereum Foundation has sponsored the conference twice. The club has been chain-agnostic and even-handed between companies and ecosystems since its founding, which is a policy rather than a posture, and it is the reason we think we can host this particular argument without tilting it.
ETHPrague has been bringing the Ethereum community together in Prague since 2021. Over the years it has become a platform connecting Ethereum builders, researchers, developers and the wider technology community, with a strong focus on bringing global voices into dialogue with the local ecosystem. ETHPrague has hosted leading figures from the Ethereum ecosystem including Vitalik Buterin, Péter Szilágyi, Dankrad Feist and Mario Havel. In 2026, Marian Goodell, CEO of Burning Man, joined the programme, while in 2025 Sir Tim Berners-Lee took part in a panel discussion with Vitalik Buterin. ETHPrague does not claim that its organisers are core developers. Its role is to create a space where leading developers, researchers, technologists and thinkers can meet, exchange ideas and connect with the local community. Production of ETHPrague and its year-round activities is handled by Duct Tape.
The two organisations already work together: TUM Blockchain Club is a listed partner of ETHPrague 2026. This coalition was not assembled for this application.
The people responsible
All seven of us are attending Devcon 8 India regardless of the outcome of this proposal, so staffing the space from 9:00 to 18:00 across all four days, and being present for setup on 2 November, carries no conditionality.
Kerem Eskici (Vice President and Head of External Relations, TUM Blockchain Club) and Alexandra KrejÄŤĂková (core organiser, ETHPrague) manage the hub and are the points of contact for Devcon, for speakers and for the other hubs.
The wider team is Felix Rihacek, President of TUM Blockchain Club; Gopi Mehta, former President, who has lived in Mumbai for many years and leads our outreach to local communities; Eva Freiberger, former board member; Annemarie NieĂźner; and Srini Jagannath. All six TUM Blockchain Club organisers have run the TUM Blockchain Conference at the scale described above.
Neutrality
The hub is about the problem, not about any one answer to it. The Ethereum Economic Zone appears in our programming as one of several competing approaches and as evidence of where the conversation has moved, not as a position we are advocating. Day 2 is built so that each approach gets equal footing and is questioned by the room.
- No token promotion, no product pitches, no branding, no single-vendor sessions.
- Speakers state their affiliation at the start of every session, and we will say plainly when a speaker is presenting their own team’s approach.
- No approach gets more than one anchor slot on Day 2, and we do not close a debate by declaring which side was right.
- TUM Blockchain Club takes no ecosystem or protocol sponsorship for this hub, consistent with the chain-agnostic policy it has held since founding. Organisers with any commercial affiliation to a proposed approach will not curate or moderate that session.
- House rules will be posted visibly in the space, and an organiser is present throughout opening hours.
Languages
The programme runs in English. Our organisers can additionally support conversations in German, Czech, Hindi, Marathi and Turkish, and we will seek volunteers for further Indian languages through this thread and the onboarding process.