Author: Ruben Somsen 2021-12-08 22:51:50
Published on: 2021-12-08T22:51:50+00:00
In an email exchange between Ruben and Jeremy, they discuss the possibility of 0 value outputs in Bitcoin transactions. Jeremy argues that as long as a 0 sat output is eventually spent and confirmed, it is not a big problem even if it enters the mempool. He suggests that it would be rational for miners and users to save the transaction spending the 0 sat output to resurrect it if it is not spent again. Additionally, he believes that intentionally trying to keep the 0 value output would be uneconomical compared to creating dust outputs, and therefore, it should not be popular. Ruben agrees with Jeremy's points but sees a possibility where a user non-maliciously behaves in a way that causes the 0 sat output to never be spent. This could happen if another output gets spent at a higher feerate, the tx spending the 0 sat output does not make it into the block due to lower fees, and the user invalidates the tx spending from the 0 sat output. However, Ruben acknowledges that this scenario is somewhat unlikely. The email also discusses a possible consensus change that involves having a UTXO with drop semantics. If a UTXO with a certain flag on it is not spent in the block it is created, it expires and can never be spent. This could be used for 0 value outputs that are present only for attaching fee in the mempool and would be treated like an op_return otherwise. Two cases could be added for this: one bare segwit version (just the number, no data) and one that's equivalent to taproot. This covers OP_TRUE anchors efficiently and ones that require a signature as well. This suggestion is similar to how Transaction Sponsors works, but without full tx graph de-linkage.
Updated on: 2023-06-15T03:23:41.981150+00:00