Ethereum's 66-proposal Hegotá list is a budget, and only one line item is funded

Ethereum Foundation researcher Toni Wahrstätter says developers have 66 Ethereum Improvement Proposals under consideration for Hegotá, the upgrade scheduled for 2027, and that the list will be cut down over the next few core developer calls to proposals with a realistic prospect of implementation, devnet, testnet and mainnet readiness inside that year ^[1][2][28]. Anything that misses moves to a later fork ^[3], and on the current cadence, with Glamsterdam targeted for the second half of 2026 and Hegotá for 2027, that slip is measured in about a year at best ^[26].

The arithmetic is the story. Exactly one proposal is locked in: FOCIL, or Fork-choice Enforced Inclusion Lists, EIP-7805 ^[4]. That leaves 65 items competing for whatever capacity is left ^[5]. FOCIL would let a committee of validators assemble lists of pending transactions that block builders are obliged to include, removing unilateral discretion from any single builder ^[6].

The privacy cluster is the visible contest. Wahrstätter says the Protocol Architecture team wants Frame Transactions (EIP-8141) prioritised alongside FOCIL ^[14]. Frames would let accounts define their own rules for how transactions are authorised, executed and paid for, which in practice means one party covering gas for another, bundling operations, or using alternative signature schemes without abandoning existing addresses ^[7][8]. Two supporting proposals fix the plumbing: Keyed Nonces (EIP-8250) gives an account independent counters per transaction stream, so one stuck transaction no longer blocks everything behind it ^[9], and EIP-8272, Recent Roots, lets a transaction prove validity against a recent cryptographic record rather than state that may shift while it sits in the mempool ^[10]. Transaction Assertions (EIP-7906) is also in the package, allowing wallets to set conditions on what a transaction can do after submission ^[13]. "Together with Frames, these enable privacy pools where the pool itself can pay fees, removing the need for intermediaries," Wahrstätter wrote. "Add FOCIL support and privacy transactions also gain protocol-level inclusion guarantees" ^[12]. The claimed payoff is less external infrastructure: fewer specialised relays and intermediary services, and fewer places for metadata to leak ^[11]. It does not make Ethereum private by default; ordinary transfers between standard addresses stay fully visible ^[15].

Now the part that decides the allocation. The same list contains repricing work for data and state growth (EIP-8131, EIP-8279, EIP-8368) to support gas limits reaching toward 600 million, shorter slot times (EIP-8198), changes to issuance and validator rewards (EIP-8363), anti-correlation penalties (EIP-7716), EVM simplifications, and the first formal discussions of zkEVM and post-quantum cryptography ^[16][18]. Developers are also weighing separating block-access-list data from the execution payload and testing optional zkEVM proofs on mainnet ^[27]. Wahrstätter's own framing is that the pricing and scaling work, less glamorous than privacy, is among the most important short-term progress available ^[19]. That is the competing bid, from the same person making the privacy case.

The constraint is time, not enthusiasm. "A fork can't be a wishlist by the community or core devs jamming on what Ethereum should eventually become," Wahrstätter wrote. "Instead, we have to decide what Ethereum should become next, and what has to wait" ^[20]. He also noted developers were 256 days into Glamsterdam and still aiming to ship it by the end of 2026 ^[21], and cautioned that they "can't do everything at once and still expect to ship on time" ^[22].

Watch the next core developer call, scheduled for Monday at 14:00 UTC ^[24], for whether the Frames trio moves from advocacy to assigned implementations and devnets. Watch, too, whether the repricing EIPs quietly absorb the capacity: Vitalik Buterin has been pushing privacy and quantum resistance as priorities ^[25], but scoping decisions are made by whoever can demonstrate a testnet, not by roadmap sentiment.