Since MACI v1.1, the protocol featured an implementation of pairwise subsidy, written by @chaosma (original blog post). This feature has not been included in the production version of the circuits (no trusted setup) due to its large constraints.
As it is now removed from the main protocol code, though still available in the v1.2.0 release, this issue aims to track further work on this feature. This work could include attempts to further optimize the circuit, or new implementations.
It should also be noted that there is another similar effort (on stand-by currently) being tracked in #905 - which could be a direct replacement of this feature.
Since MACI v1.1, the protocol featured an implementation of pairwise subsidy, written by @chaosma (original blog post). This feature has not been included in the production version of the circuits (no trusted setup) due to its large constraints.
As it is now removed from the main protocol code, though still available in the v1.2.0 release, this issue aims to track further work on this feature. This work could include attempts to further optimize the circuit, or new implementations.
It should also be noted that there is another similar effort (on stand-by currently) being tracked in #905 - which could be a direct replacement of this feature.