Ethereum’s development roadmap is taking a clearer shape after core teams wrapped up a key interoperability meeting in Svalbard, Norway. The Ethereum Foundation confirmed that the Glamsterdam multi‑client development network is now operational, with enshrined proposer‑builder separation (ePBS) and gas repricing via EIP‑8037 progressing steadily. Meanwhile, features originally slated for Glamsterdam – including FOCIL, Verkle Trees, and account abstraction – have been shifted to Hegotá, setting that fork up as a late‑2026 “cleanup and hardening” release.
On the execution layer, ePBS has completed end‑to‑end testing and is running stably across almost all client implementations. This architecture separates block building from consensus, marking a formal step toward protocol‑level MEV management. EIP‑8037, now in final draft, introduces a cost_per_state_byte model that targets approximately 60 GiB of annual state growth at a 300 million gas block limit. Under this framework, contract deployment costs rise roughly 10x and new account creation becomes 8.5x more expensive, while separate metering for code deposit preserves deployability for large contracts such as Uniswap pools.
Scalability and censorship‑resistance improvements that were previously part of the Glamsterdam scope now belong to Hegotá. The FOCIL (Fork‑choice Inclusion Lists) prototype has a runnable implementation, and the account abstraction requirements have been defined, with the next phase entering a multi‑client devnet validation stage. Hegotá will also introduce Verkle Trees, which could cut node storage requirements by up to 90% and pave the way for stateless clients. The Glamsterdam mainnet activation is still targeted for the first half of 2026 – though Q3 appears more realistic after the Soldøgn interop devnet concluded in early May – while Hegotá is positioned as a late‑2026 “cleanup and optimization” fork addressing technical debt in Ethereum’s data structures.
Beyond the code, a leadership reshuffle in Ethereum’s Protocol Cluster was announced. Will Corcoran will coordinate zkVM proofs and post‑quantum consensus research, Kev Wedderburn will lead zkEVM development, and Fredrik will oversee protocol security and the “Trillion Dollar Security” initiative. Original leads Barnabé Monnot and Tim Beiko will gradually step back from management, and researcher Alex Stokes is taking a leave. The Foundation noted that under the outgoing structure, the group completed the Fusaka upgrade in December 2025, which introduced PeerDAS for data availability and boosted mainnet gas capacity.
As always, timelines remain subject to testnet outcomes, and users are reminded that upgrade‑related volatility carries risk despite an orderly‑looking roadmap.
