19:17:54 <sarang> Worked on some multisig stuff today, and I have a working proof-of-concept for MPC inversion, albeit under very weak security guarantees
19:19:28 <sarang> Downside is the Paillier encryption stuff adds complexity and a separate static key per player (can be reused between signatures)
19:21:16 <sarang> Right now I assume honest-but-curious players, who follow the MPC correctly but may be interested to learn other players' secrets
22:20:22 <sarang> Some very initial draft code on the MPC inversion, of poor quality (should not be used anywhere): https://github.com/SarangNoether/skunkworks/blob/inverse-mpc/inverse.py
22:22:31 <sarang> For a group of players holding additive shares of a secret scalar field element `x`, it collaboratively computes `1/x*U` for a fixed group element `U`
22:25:41 <sarang> (this is the key image format for Triptych, RCT3, and one version of Omniring)