Improved reward functions
This MR updates the reward functions as follows:
block_reward (0, e) = (e / 32) * r * 80 block_reward (p > 0, e) = (e / 32) * a endorsing_reward (0, e) = (e / 32) * (1-r) * 80 endorsing_reward (p > 0, e) = endorsing_reward (0, e) / b
where the constants are
r = 1/2,
a = 6,
b = 1.5.
The new reward functions are meant to better protect against various attacker strategies, and in particular when taking inflation into account.
An analysis of these functions is currently underway.