Marchhill, an Ethereum contributor, has put forward a draft initiative designed to facilitate generic proof verification alongside sponsored batch processing. This approach permits multiple transaction claims to utilize a shared proof while maintaining distinct payments and signatures for each. Within this structure, sponsors cover the cost of batch proofs, whereas individual member transactions pay for their specific claims. Additionally, valid subsets of these transactions may be incorporated into a block without necessitating that members re-sign their transactions.
Current fee structures and resource constraints are tentative, hinging on future performance metrics. The offering exists as a draft pull request lacking an assigned EIP number or a set date for mainnet deployment. Submitted to Ethereum's EIP repository on Nov. 13 by Marchhill, this proposal intends to activate the LEANSTARK_SCHEME reserved in EIP-8288 and establish regulations for sponsored batches. The initiative is still in development and has not yet been accepted as an Ethereum network upgrade.
Draft Proposal Details and Status
The draft proposal consolidates generic verification, identity proof, resource limits, dependency summaries, and batch payments into a single submission. This document replaces an earlier pull request, identified as #12654, though an official EIP number and a dedicated discussion thread have yet to be assigned. The core of the draft involves a payment structure where a sponsor pays for every retained batch leaf, defined as an individual proof element that includes multiple claims. Payments from member transactions will be based on the claims they submit themselves.
A builder could include any valid subset without requiring members to sign again. Consequently, an individual whose transaction was prepared for inclusion would not need a new signature merely because other participants from the original group were excluded, as per the proposed mechanism.
The draft proposal maintains that individual transaction nonces and receipts will remain intact, where a nonce defines a transaction's position within an account's sequence and a receipt logs its execution outcome. Under this approach, sharing proof work would keep those records attached to their respective separate transactions.
Payment matching is scheduled to occur during finalization, with recursive proofs verifying the sources utilized in the batch. The submission separates the effort required to validate claims from the effort needed to confirm that necessary payments accompany them. Discounted claims and sponsor payments would be directed straight to block builders, while ordinary proofs covering a single statement would stay eligible for the public mempool where transactions await inclusion in blocks.
Vitalik Buterin and Thomas Coratger have introduced EIP-8288, a proposal to move the aggregation of cryptographic signatures and proofs away from Ethereum's primary execution layer. Within this framework, transactions would pinpoint specific statements requiring verification, while a recursive STARK mechanism would confirm their validity at the block level. These STARKs function as cryptographic evidence showing a computation occurred correctly without forcing every verifier to re-run it, and recursion enables one proof to validate others by merging their outcomes.
While EIP-8288 presently keeps the generic leanSTARK scheme reserved against use in dependency frames, Marchhill's submission aims to unlock that option. This approach also swaps the base draft's raw data-hash statement for a stable digest that represents a collection of claims. The architecture builds on prior conversations regarding cryptographic verification, notably Buterin's backing of complex cryptography managed via aggregated proofs; his remarks focused on the foundational structure, whereas Marchhill's document details a concrete setup for verification and payments.
According to the submitted revision, native verification incurs a provisional cost of 15,400 gas while ordinary generic verification requires 572,000 gas. The intrinsic proof share is allocated 5,100 gas, whereas a leaf payment draws 675,000 gas and a batch claim receives 4,200 gas.
The draft notes that binding payments introduce fixed work distinct from standard verification, necessitating separate charges for leaf payments. Neither claims nor payments will attract extra intrinsic proof share costs. Because these figures depend on joint calibration, they represent neither a finalized fee structure nor a guaranteed reduction in user expenses.
The draft proposal establishes resource controls featuring a provisional block limit of 16,384 distinct dependencies, a metric applied independently of gas consumption alongside an inclusion-list fit rule. Within this framework, a single leaf node is permitted to contain up to 256 claims, whereas a proof step may accommodate 16 inputs; neither of these specific figures functions as a block-wide limit.
Mechanics of Payment and Verification
Prior to activation, the proposal mandates measurements covering the complete processing schedule, late sponsor repair mechanisms, and per-proof charges that must fit within transaction gas limits including required overhead. The benchmark material distinguishes between measured individual-node and single-machine block results against modeled 16-worker figures, requiring that any evidence and remaining performance gaps be recorded in the documentation.
The draft proposal introduces a shared payment model for batch proofs while maintaining concrete commitment construction and encodings within the pinned proof-system specification without adding new networking messages. It notes that an invalidated sponsorship could result in wasted builder work which inclusion gas would not reimburse, and it does not introduce new networking messages.