Key Insights:
- In the latest Ethereum news, Vitalik Buterin now supports leanPCS on ETH as a route to scalable proof verification.
- EIP-8288 would aggregate proofs while blocks retain claims and input-output hashes.
- Developers continue weighing post-quantum schemes, leanVM design, and the 2027 upgrade pace.
Vitalik Buterin says his view on Ethereum news of leanPCS has shifted from doubts to strong support, placing the proposal at the center of a discussion about scalable proofs on Ethereum.
His explanation of EIP-8288 describes a system for handling computations whose results matter on-chain, while the original data need not remain visible. The proposal also intersects with post-quantum security plans, where developers are weighing technical choices against a crowded upgrade schedule.
Ethereum News: How 8288 Could Reduce Onchain Proof Data
Based on the Ethereum news, Buterin said EIP-8288 targets work that needs to be performed on data once, with the chain retaining the answer.
In an X post, Vitalik Buterin listed signatures, STARKs, Merkle branches, and fully homomorphic encryption computations as possible uses. He identified signatures and STARKs as the most important examples.

The size of cryptographic data could become a concern as Ethereum considers quantum-safe systems. Buterin compared 65-byte ECDSA signatures with WOTS or SPHINCS signatures measuring one to six kilobytes, depending on parameters.
He also compared 200- to 1,000-byte Groth16 or PLONK proofs with STARK proofs around 128 kilobytes.
Under his description, a user would publish a STARK, SPHINCS signature, or polynomial commitment opening. Each node would collect new objects and produce a STARK containing claims about them.
Those claims on Ethereum news could state that a public key signed a message or that a program produced a particular output.
The nodes’ proofs would then be aggregated again, eventually producing a master proof for a block.
The network would rebroadcast input and output hashes, estimated at about 64-96 bytes per claim, along with proofs. However, the original signatures and proofs would not need to be placed on-chain under this design.
The block would still carry the statements that need verification, such as message and key details or program inputs and outputs. Buterin said those statements would use a specialized frame type. The Ethereum proposal’s planned sub-EIPs could separate this process into distinct components.
LeanPCS Offers Flexibility While LeanSTARK Awaits an ISA
The three approaches differ in the object users create. LeanSPHINCS uses a hash-based signature scheme, whereas leanSTARK allows users to prove a computation. LeanPCS focuses on polynomial commitment schemes, which prove that a polynomial evaluates to a specified value.
Buterin said leanSPHINCS could support quantum-safe accounts and multisignatures, with 4,096 gas for a 64-byte signature. Its scope would remain tied to one signature scheme. LeanSTARK, by contrast, could support different signature schemes, privacy protocols, and multisignature logic.
That flexibility would require Ethereum to specify an instruction set architecture, or ISA, for user programs. Buterin said agreement could take time because the ISA would be difficult to upgrade. He also said developers could later consider exposing it inside regular contracts.
LeanPCS would require fewer choices, according to Vitalik Buterin, including a field specification.
He cited a 128-bit binary field as a natural option and mentioned a recursive tower field extending to 256 bits. He said leanPCS could be less efficient on-chain than leanSTARK, but more flexible off-chain.
Buterin described a possible sequence on the Ethereum news: implement leanPCS first, then add leanSTARK after developers settle the ISA.
He said a PCS oracle could enable STARK verification for about 20,000 to 40,000 gas. The ETH crypto discussion therefore includes both proof efficiency and decisions about which capabilities arrive first.
Ethereum News: Post-Quantum Roadmap Faces Timing Questions
The technical debate on the Ethereum news follows disagreement over a proposed 2027 schedule involving Glamsterdam, Hegotá’s execution and consensus layers, and I.
Soispoke.eth argued that four forks in 2027 would not be realistic. Barnabé Monnot called I a stretch goal and said work through H remained ambitious, though some core developers considered it feasible.
Soispoke raised questions about signature schemes, leanVM, hash selection, cryptographic agility, and post-quantum consensus designs.
Following that, Monnot agreed these issues required discussion and said leanVM could be a major consensus-layer blocker. He also said developers were considering when to freeze signature choices.









