Updated Ethereum roadmap, EF stakes ETH and more - The Daily Gwei Refuel #860 - Ethereum Updates

Vitalik Buterin and the EF architecture team, including Justin Drake, Ansgar, Barnaby, and Francesco, published a continuously updated Ethereum protocol roadmap at strawmap.org covering 2026 through 2030 and beyond. The roadmap contains no net-new items, with all entries backed by prior posts or community discussions. It is organized into consensus, execution, and a newly separated data layer covering blobs and call data. Specific items include shorter slot times, fast finality, improved PeerDAS, GigaGAS L1, account abstraction, and binary trees. Upcoming upgrades Amsterdam and Hegelta have reasonably defined contents, while I-star, J-star, and later upgrades are speculative. Items marked with a thinking emoji, including Attester-Proposer Separation, carry explicit uncertainty about whether they will ever ship.

The Ethereum Foundation began staking a portion of its treasury, making an initial deposit of 2016 ETH on February 24th with plans to stake approximately 70,000 ETH total, directing rewards back to the EF treasury. The effort uses open source software Dirk and Vouch by Attestant, which is positive for client diversity. At the time of discussion, total ETH staked on the network was roughly 37,533,000 ETH, making the EF's planned stake about 0.18 percent of the total, equivalent to rank 28 on the staking leaderboard. The EF is staking independently rather than through Lido, Binance, or other major providers. The EF previously avoided staking to preserve credible neutrality and avoid influencing validator decision-making, and the argument made is that the EF's social power over the network already far exceeds its economic power regardless of stake size.

The EF formed a new Platform Team announced February 18th, headed by Josh Rudolph, whose background is in user interface experience and account abstraction. The team's stated goal is delivering the strongest possible Ethereum platform where L1 and L2 best support users, apps, and builders. Account abstraction was identified as the single biggest friction point for average Ethereum users, currently requiring multiple approval steps rather than single-click interactions, with the problem attributed to a combination of technical difficulties and social and coordination challenges. The EF Protocol Support Team also launched its own website detailing its work and how outside contributors can participate, with six core members including Nickso, Josh Davies, Mark Gareya, Mario Havel, Buddha, and Dionysus.

Charles St. Louis, previously of MakerDAO and Delphia, joined the Ethereum Foundation to lead a DeFi initiative, with Yvangby, co-founder of Gearbox Protocol, joining alongside him. The EF's stated DeFi philosophy is called DeFi punk, defined as finance that could not exist without Ethereum rather than finance that is only marginally better than TradFi. Flash loans and AI agent commerce were cited as fitting examples. The argument made is that TradFi adoption of Ethereum will occur regardless of EF involvement because Ethereum offers TradFi a neutral layer of significant value.

Rocket Pool's Saturn One upgrade deployed on mainnet, introducing an RPL fee switch, 4 ETH validators, and Megapools. Rocket Pool was described as the only fully permissionless staking protocol on Ethereum as of 2026, with no reliance on AWS or a select group of trusted professionals, and currently holds almost 500,000 ETH staked at rank ten on the staking dashboard. Entities ranked above it, including Binance, Coinbase, Figment, Kraken, Blockdaemon, Everstake, and Kiln, were characterized as neither decentralized nor permissionless.

The MaxEB change introduced with the Pectra upgrade raised the maximum ETH per validator from 32 ETH to 2048 ETH. Adoption of consolidation has been slow because there is no strong financial incentive and the operational work is significant. Approximately 18 percent of validators currently hold 0x02 withdrawal credentials enabling accumulation beyond 32 ETH, though that credential alone does not confirm consolidation has occurred. Several larger node operators are expected to consolidate toward the end of the year. Reducing the total validator count is necessary to relieve network load, and the urgency increases in a post-quantum context because quantum signatures are larger than BLS signatures, causing slower propagation that would meaningfully degrade network performance at high validator counts.

L2Beat launched a new page called L2Beat Interop tracking 22 bridging protocols, showing metrics including flows, token transfers, transfer sizes, and bridge mechanism types such as burn and mint or lock and mint, with plans to publish complete breakdowns of each protocol similar to its existing L2 coverage.