On the possibility of evil covenants and implications for soft fork proposals

The discussion centers on the comparison between the use of covenants and multisig (multi-signature) arrangements in the context of enhancing security for Bitcoin vaults. It is argued that while covenants offer a direct method to improve security, their benefits could also be achieved, albeit with some compromises, through the use of multisig solutions combined with oracles. This compromise is not seen as equal; for instance, an evil overlord might only face minor inconveniences with multisig, whereas an honest user relying on covenants could be exposed to additional security risks.

The core argument extends to the necessity and demand for trustless vaults within the Bitcoin network. Trustless vaults are portrayed as essential for safeguarding significant assets, potentially worth billions, without relying on an external party's trustworthiness. Currently, services like BitGo, Coinbase, and Casa offer vault solutions using multisig schemes, which indicates a recognized demand and existing implementations of trusted, external vaults. However, these do not fulfill the criteria for being truly trustless since they depend on external validation.

The importance of developing native, trustless vaults that operate without external dependencies is highlighted as a means to potentially enhance security for users. The reliance on an external oracle or system introduces an additional line of trust, diverging from the ideal of a trustless system where validation is entirely decentralized and inherent to the blockchain. The conclusion drawn is that if it is feasible to replace a trusted system with a trustless one without compromising on security or functionality, then such a transition should be pursued. The expectation is that trustless vaults, owing to their superior security guarantees, would likely see higher usage compared to their trusted counterparts.