Last week, we interpreted the upcoming L2Beat @l2beat reclassification and its impact on Ethereum Rollups. Now, let's delve into the solution proposed by Cartesi for this higher standard: the PRT (Permissionless Referee Tournament) anti-fraud mechanism, which places the Cartesi Rollups SDK in its second phase.

PRT is Cartesi's first solution to this challenge (followed by the Dave proposal): a permissionless decentralized method to validate state transitions, thus resisting Sybil attacks without requiring extensive resources or trust assumptions. It aims to provide a secure, decentralized, and ultimately fast way to resolve disputes. Unlike older models that are susceptible to Sybil attacks or rely on permissioned validators, PRT aims for permissionless scalability while minimizing the burden on honest validators.

How does it work? PRT uses computational hashes. Participants no longer merely commit to the final state, but rather to the entire computational path. Thus, PRT not only exposes liars but also reveals lies.

Disputes are resolved using a bracket-style tournament. Validators can work as a team, and due to halving the claims each round, honest validators only need to complete logarithmic work in logarithmic time. Even ordinary hardware is sufficient to handle large-scale Sybil attacks.

PRT has launched an upgraded honeypot, but without an economic model. Delay attacks still exist, so it is currently not recommended for high TVL applications. However, it propels Cartesi toward its second phase Rollup deployment and aligns us with the new standards of L2BEAT.

Complete thread: