Ethereum (ETH) Developer Pedro UID Splits EIP-8130 Into Three Composable EIPs

EIP-8130 Refactored Into Three EIPs

Ethereum's long-running dispute over how to bring native account abstraction into the protocol layer has produced a compromise: the core proposal EIP-8130 is being refactored into three composable EIPs — 8398, 8399 and 8400 — laid out on August 27 by developer Pedro UID in a discussion thread on the Ethereum Magicians forum, with the matching GitHub pull request #12248 opened the same day. Account abstraction is the effort to make smart contract accounts first-class citizens capable of custom verification logic, social recovery, gas sponsorship and transaction batching. Today that functionality runs mostly through ERC-4337 and off-chain bundlers; native AA would move it into the protocol itself. The fight was over how. The keystore camp, embodied by EIP-8130, would manage signature verification through a whitelist of trusted verifiers, keeping verification costs predictable and capped — a design friendly to Layer 2 networks — and ship an account standard with policy controls and session keys. The rival Frame Transactions proposal, EIP-8141, permits unstructured verification with arbitrary EVM code at any stage of a transaction, enabling flows such as an account holding no ETH acquiring funds mid-execution; it maximizes flexibility and allows innovation without a hard fork, at the price of fragmentation risk, with some advanced patterns requiring a private mempool. Ethlabs researcher Derek Chiang's earlier comparison framed the trade-off as simplicity and control versus extreme flexibility on Ethereum's protocol upgrade path.

The new stack layers cleanly: EIP-8398 defines a portable account keystore covering actors, verifiers, account configuration, creation and cross-chain portability; EIP-8399 builds on it with a native AA transaction type, 0x79, adding batching, sponsorship and ordered nonces; EIP-8400 requires both and adds policy, account locking and nonce-less transactions sharing the same envelope. The original EIP-8130 remains unchanged as the predecessor of the three specs. One community reading holds that the Frame camp's flexible approach broadly prevailed while keystore ideas aimed at L2s were folded in — neither side lost outright. Everything remains at the proposal stage, with no slot yet on a formal upgrade schedule.

Glamsterdam Gas Repricing Targets State Growth

The same upgrade horizon carries the first serious gas fee restructuring since Berlin in 2021. Glamsterdam, Ethereum's next major upgrade, includes two proposals that reprice state: EIP-8037 raises the cost of creating new state — accounts, storage slots and contract code — while EIP-8038 raises the cost of accessing existing state, both aiming to align pricing with the resources the network actually carries. The driver is state bloat. The EIP-8037 specification records the state portion of Geth's database at roughly 390GiB as of January 2026; after the block gas limit doubled from 30 million to 60 million, daily new state grew about 3.1x, from roughly 105MiB to 326MiB — an annual pace near 116GiB. Scaled proportionally to a 200 million gas limit, that trajectory reaches about 387GiB per year, breaching the 650GiB performance guideline within a year. The proposals set a common cost of 1,530 gas per byte of new state plus a state-only gas dimension, targeting roughly 120GiB of annual growth under a 150 million gas reference block limit.

The jumps are steep: new account creation rises from 25,000 to 183,600 gas, about 7.3x; a new storage slot from 20,000 to 97,920 gas, roughly 4.9x; and deploying a 24KiB contract alongside an account from 4,947,200 to 37,784,880 gas, about 7.6x. A published impact dashboard re-executed 929,731,274 transactions across 4 million blocks from December 3, 2024 to June 15, 2026 under both rule sets. Under EIP-8037, 174,473,898 transactions that would fail at the original limit succeed once it is raised, while 2,687,652 — about 0.29% — land in a “possibly broken” bucket; EIP-8038 resolves 84,708,228 via a raised limit and flags 3,036,537, or 0.33%. Those counts are transactions, not distinct contracts, since one active application repeating the same flow inflates the tally. A hard floor remains: the fixed 2,300 gas stipend, hardcoded gas forwards, gasleft() branching and presigned transactions with capped limits cannot be rescued by raising limits. An Ethereum Foundation report documents repeated failures on the ERC-4337 EntryPoint and stacks from ZeroDev and Alchemy, plus Across, Socket/Bungee, CoW Protocol and 0x; immutable contracts will need new EntryPoints or factories and a user migration. Both EIPs sit in formal Review with a roadmap target of Q4 2026; the Platåberget validation testnet went live on August 17, while Sepolia, Hoodi and mainnet fork dates remain unpublished. Readers tracking the market in real time can follow live spot and futures prices on Binance.

Q4 2026 Validation Window Opens

Read together, the two threads describe one overhaul: Glamsterdam is set to reshape both what an account is and what persistent state costs on Ethereum. For node operators and wallet teams, nothing is actionable yet — neither the 8398/8399/8400 stack nor the 8037/8038 pair carries an activation block or fork date; both remain proposal-stage text awaiting core developer convergence, much like the recent EIP-8394 proposal to guard $104 billion in staked ETH. The August 17 Platåberget launch, however, opens the validation window now. COINOTAG's read: developers should test against the canonical EIP specifications rather than recaps, because the module boundaries of the three-way AA split — like the new state-gas dimension — will determine what breaks in production; readers new to the upgrade path can start with our Ethereum 2.0 upgrade guide.