remove output blackballing

[tobtoht]

#8756

[UkoeHB]

I did not check if everything related to blackballing was removed, if anything was missed it can be cleaned up in a future PR.

[jeffro256]

I'm not approving the changes in principle yet, but this PR does successfully remove only blackball features and should not affect the behavior of anything else.

[tobtoht]

Rebased.

[jeffro256]

I believe we should keep the output blackballing feature as long as there still exists the possibility of poisoned outputs (e.g. Mordinals). If/when full-chain membership proofs get merged into mainnet, then it would be safe to remove blackballing entirely.

[tobtoht]

I disagree. The current blackballing implementation is not helpful at mitigating effective ringsize reduction from large amounts of non-uniform outputs (such as Mordinals - which was mitigated), as it would require manually generating and importing up-to-date blacklists before sending a transaction. (About half of selected decoys are < 1 day old, so at least daily). Transactions that do not include these outputs in their rings are finger-printable on chain, reducing transaction uniformity and increasing traceability for the small subset of users that would go through the effort.

A better solution for output blackballing would be to implement it on the daemon side and have it mark undesirable outputs in the get_outs call. This would not require storing blackballed outputs in RingDB. In this way, all users would benefit and transaction uniformity is not harmed.