So many optimistic comments, but I still don't get what is the consensus algorithm used by the project and how was it picked?