Author: Peter Todd 2017-02-24 02:58:11
Published on: 2017-02-24T02:58:11+00:00
In an email exchange between Bram Cohen and Peter Todd, the two discussed the use of merkle trees as a method for compactly proving the current state of an output. Todd referenced his own article which proposed a tree that would require random access updates, but position entries in the order they were added instead of sorting by their hash. Cohen argued that updating to indicate spent status would lead to issues of TXO size and cache coherence on updates, and that it was better to use a basic fundamental data structure. Todd clarified that he was referring to the ability to extract a compact proof from an MMR that a given output is unspent, and asked if Cohen agreed or disagreed with this statement before discussing performance characteristics.
Updated on: 2023-06-11T21:39:19.505715+00:00