[RFC] Canonical input and output ordering in transactions



Summary:

In this context, Wladimir J. van der Laan suggests dropping the discussion of IsStandard or consensus rules. Mark Friedenbach raises concerns about the proposal's compatibility with SIGHASH_SINGLE, which plays a role in assurance contracts among other things, and mentions that sometimes the ordering is set by the signing logic itself. However, randomization cannot be used if the ordering is unconstrained. Therefore, the proposal still makes sense as an alternative to randomization. Wladimir believes that moving towards verifiable, deterministic methods where possible is a good approach.


Updated on: 2023-06-09T22:17:24.589286+00:00