sowbug
commented
8 months ago - Each performer hears local rendition of the composition
- Any performer can contribute
- [magic] each local rendition predicts the current contribution(s) based on information provided by CRDT
- On replay, CRDT is complete and no prediction is needed