mirror of
https://github.com/status-im/eth2.0-specs.git
synced 2025-01-10 10:45:46 +00:00
33a05109ea
I think we want `first_committee[epoch % len(first_committee)]` as opposed to `first_committee[slot % len(first_committee)]`. The reason is that if the shuffling happens infrequently and `len(first_committee)` is a multiple of `SLOTS_PER_EPOCH` then the proposers will not be sampled fairly.
Taking this logic further, we may want to avoiding always picking the proposer from `first_committee`, e.g.:
```
validators_at_slot = []
for crosslink_committee, _ in get_crosslink_committees_at_slot(state, slot, registry_change):
validators_at_slot.append(crosslink_committee)
return validators_at_slot[epoch % len(validators_at_slot)]
```