Researchers in distributed systems have been focusing on consistency and liveness in Byzantine consensus and state machine replication. However, bad actors can manipulate transaction order in public blockchains, leading to MEV exploitation. Fair-ordering protocols aim to address this security gap by introducing transaction order-fairness as a third consensus property, enhancing transparency and resistance to adversarial reordering.

Receive-Order-Fairness (ROF) is a strong notion of fairness but is impossible to achieve universally due to network constraints. The Condorcet paradox illustrates the challenges in achieving perfect fairness, leading to the proposal of weaker fairness definitions like batch order fairness. This highlights the complexity of maintaining fairness in asynchronous and synchronous networks.

Hedera Hashgraph’s approach to fairness through median timestamping is susceptible to manipulation by a single adversarial node, undermining the fairness of transaction order. Aequitas and Themis protocols introduce Block-Order-Fairness (BOF) to relax fairness definitions, ensuring fair transaction order while addressing communication complexity and liveness issues in distributed systems.

Aequitas faced limitations in achieving BOF, leading to the introduction of the Themis protocol with improved communication complexity and scalability. Themis employs Batch Unspooling, Deferred Ordering, and Stronger Intra-Batch Guarantees to enhance fairness in transaction ordering, allowing for efficient consensus in large networks. This evolution in consensus protocols emphasizes the importance of provable fairness over perceived fairness in decentralized systems.

Read more at Coin Telegraph: The impossibility of perfect fairness in transaction ordering