Replies: 2 comments 2 replies
-
I got it wrong. Here, the vrf of each miner is different in each round. the new data look as expect overall |
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
i am not sure, but maybe there is a bug in the block eligibility oracle algorithm. when p is very smaller, probability of each miner is quite small. but the vrfFrac can regard as a randomness. maybe each miner has less chance th get elegibility. os no one have change to generate block or all have chance to generate block.
simulate code :
https://github.com/hunjixin/go-spacemesh/blob/42fcd3ded102d2ca8b06d571bc346cb573e2c3f6/hare/eligibility/oracle_test.go#L58
simulate result:
Beta Was this translation helpful? Give feedback.
All reactions