Author: Yuval Kogman 2019-07-19 17:45:17
Published on: 2019-07-19T17:45:17+00:00
In a discussion on the bitcoin-dev mailing list, Mike Brooks suggested giving scripts the ability to refer to data on the blockchain in order to reduce transaction sizes. However, Yuval, another participant in the discussion, expressed skepticism that this feature would actually be utilized by users, given that address reuse is discouraged and mining fees are usually pushed onto customers. Furthermore, Yuval argued that unless extensively utilized, this change would impose an externality on validating nodes by making pruning impossible and allowing random reads on all historical data, not just the UTXO set. Yuval suggested that this redundancy might be better addressed as optional functionality in the p2p layer, which could potentially help with IBD (initial block download), though peer latency is currently the limiting factor. Additionally, he mentioned that encoding transaction IDs as a (block height, offset) could save several bytes per input, but the complexity of this approach could introduce new DoS vectors.
Updated on: 2023-06-13T20:10:32.607845+00:00