wminshew
commented
2 years ago no -- in the current deployment (v1), they must be sorted from smallest (address) to largest. The canonical ordering allows us to prevent duplicate immutable splits, match the hash & cheaply prevent duplicate recipients on-chain. In a future iteration (deployed in next couple of months?) it's likely we'll relax this requirement
In the docs, there is a note implying that the addresses passed needs to be ordered; is this requirement just a need of making the split deterministic?
If my protocol create splits can ensure the ordered addresses are in a order we understand would it work?