Open dholth opened 7 years ago
It looks like in this implementation i_relative is always i_min * (1 << (log2(i_relative)). That should make it possible to save a little space by storing only the number of doublings as a uint8_t.
i_relative
i_min * (1 << (log2(i_relative))
It looks like in this implementation
i_relative
is alwaysi_min * (1 << (log2(i_relative))
. That should make it possible to save a little space by storing only the number of doublings as a uint8_t.