171 sats \ 1 reply \ @0260378aef 15 Dec 2023 \ on: Worse than ordinals - Signing Bare MULTISIG input bitcoin
The problem is not only bare multisig.
If you banned bare multisig, they can do the same with hash values. In both cases (a EC pubkey, a hash function output, there is no structure you can enforce [*], and no way for you to know if the creator has a preimage for what they're publishing. So they can embed data in it.
It's pretty strange how many otherwise very technically knowledgeable people think that we can stop this with some new standardness rule or whatever. We basically can't, unless we do something very drastic to what Bitcoin even is. (Demurrage? )
[*] Apart from the fact that only approx 50% of random 256 bit strings are valid x-coords on the curve, for the former
I guess I should clarify that it's not like there's zero "value" in using bare multisig vs say data embedded in a pubkey-hash standard output here. There's some few bytes of savings by using one output instead of multiple; and from the point of view of coding some jpeg-tracking thing it's always easier to track one output instead of many.
But just talking in general, at the highest level, you can't stop this by disabling one output type
reply