@kyuupichan pointed out that the Median needs to be defined exactly
for the case when timestamps are duplicated. We should consider
updating the protocol to reject such blocks if they are not already
rejected.
Thanks for pointing out this twist in the median computation @kyuupichan, and thanks @schancel for the quick update. The reference works for me, everything looks good - merging.
@kyuupichan pointed out that the Median needs to be defined exactly for the case when timestamps are duplicated. We should consider updating the protocol to reject such blocks if they are not already rejected.