Author: Gregory Maxwell 2012-06-19 17:59:04
Published on: 2012-06-19T17:59:04+00:00
On June 19, 2012, Alan Reiner expressed his concern about an app developer who updated their RB tree code which caused a significant portion of the network to disagree on the correct root. This could potentially be disruptive and difficult to track down. Reiner emphasized the importance of good comprehensive tests and a well-specified algorithm in order to prevent similar issues from occurring. He likened worrying about implementers getting it wrong to worrying about them getting SHA256 wrong.Reiner suggested that a PATRICIA tree/trie would have been ideal for this situation due to its completely deterministic structure. He also noted that it is an order-of-magnitude more provable than libJudy trees.
Updated on: 2023-05-19T03:51:37.810343+00:00