Sharing block templates

The discussion revolves around the computational cost dynamics between sender and receiver in handling transaction identifiers (txids) within a blockchain context. The sender incurs a constant computational cost regardless of the number of txids in a block. Conversely, the receiver might face significantly higher computational burdens, potentially having to process several hundred times more work if unable to effectively cache these txids. However, this setup introduces a vulnerability wherein the receiver's long-lived identifier becomes an easy target for attackers capable of constructing 48-bit colliding transactions. This risk necessitates a continuous defensive posture against such attacks.

Furthermore, the conversation touches upon the protective mechanism inherent in compact blocks, which are safeguarded by the necessity for a matching root hash. This feature ensures that any failure or anomaly, such as a collision, is detectable. A proposed solution to mitigate the impact of collisions involves senders transmitting a small minisketch—specifically, an 8 element minisketch based on the last 64-bits of each txid. This approach would enable receivers to detect failures and recover from a small number of collisions without the need for a minisketch decoder, assuming they can identify the colliding transactions. This method suggests a low-cost strategy for enhancing reliability and integrity in transaction processing.