Ethereum co-founder Vitalik Buterin has articulated what he perceives as a significant advancement in the transactions of Ethereum. He asserts that recent advancements in account abstraction, state design, and cryptographic proof aggregation are facilitating a more distinct delineation between the functions of transactions and the prerequisites that must be established prior to their execution. Buterin stated that recent investigations into EIP-8141, alternative state models, keyed nonces, and recursive STARK-based mempools have yielded a clearer conceptual framework for understanding transactions. “An action is an effect that a transaction has,” Buterin explained. The distinction may appear predominantly theoretical.
However, Buterin contends that a clear distinction between the two could enable Ethereum developers to enhance them through fundamentally different approaches. Validation and execution have historically been interconnected. Nodes process transactions, verify the necessary information, and ultimately carry them out. Buterin posits that dependencies possess characteristics that render them especially apt for optimisation. Numerous dependencies can be evaluated concurrently.
Certain transactions can be analysed within the mempool prior to their inclusion in a block. Some entities do not necessitate access to the Ethereum state whatsoever. Those purported “pure” dependencies might be verified once by the mempool and subsequently not executed again by each validator. Buterin posits that extensive aggregations of these verifications may ultimately be supplanted by a singular STARK demonstrating that all were executed accurately. “Dependencies can be processed in parallel,” he wrote.
