BIP370: introduce PSBT_IN_PREVIOUS_OUTPOINT #1297

pull dr-orlovsky wants to merge 1 commits into bitcoin:master from dr-orlovsky:patch-13 changing 1 files +6 −16

dr-orlovsky commented at 1:50 PM on April 5, 2022: contributor

There is no single case I am aware of when PSBT_IN_PREVIOUS_TXID and PSBT_IN_OUTPUT_INDEX are used separately. Thus, I propose to keep all data about the spent output in the input map as a single data structure/key. This helps implementation and simplifies error handling: the situation where just txid or vout with the second part of the data missed for the spent outpoint are present now will be impossible.

CC @achow101

BIP370: introduce PSBT_IN_PREVIOUS_OUTPOINT

There is no single use I am aware of for `PSBT_IN_PREVIOUS_TXID` and `PSBT_IN_OUTPUT_INDEX` to be used separately. Thus, I propose to keep all data about the spent output in the input map as a single data structure/key. This helps implementation and simplifies error handling: the situation where just `txid` or `vout` with the second part of the data missed for the spent outpoint are present now will be impossible.

099b64304f

apoelstra commented at 3:49 PM on April 7, 2022: contributor

Concept NACK. This seems like it would cause unnecessary churn, strictly increase complexity (we cannot remove the old fields), and also there is no canonical encoding for the conjunction of these fields.
achow101 commented at 4:21 PM on April 11, 2022: member

NACK for the same reasons mentioned by @apoelstra
dr-orlovsky commented at 10:59 AM on April 18, 2022: contributor
1. I was assuming that the BIP370 is a draft, i.e. the old fields can be removed before it has become final standard
2. The canonical way of encoding them together is the way they are encoded in any transaction input (OutPoint structure)
luke-jr closed this on May 5, 2022

Contributors

dr-orlovsky

apoelstra

achow101