I've actually started implementing a new algorithm, from scratch. It performs faster re-adjustment, limited to 1.2 or 0.8 difficulty change between individual blocks. But it also limits the difficulty change to 3.0 or 0.33 compared to the last 120 blocks difficulty average. This means that the diff can rise ~3.0 times in 6 blocks, and it can fall to ~0.33 times in 5 blocks. In the linked formula's the impact of the new blocks themselves are not calculated in the 120 blocks average.
The idea behind this is that it will be able to handle large joins and leaves, but won't be tricked into settling on a high difficulty too fast.
Thoughts?
DGW3 was developed at a time and for a network that had far smaller spikes than we see today.
I still think what we need is a faster reaction to the hash rate spikes and drops we see.
The more time it takes to settle in to the desired diff for the actual hash rate the more time we give pools to take advantage of it and normal miners will suffer after a drop and the time it takes to bring the diff back down.
The idea I wrote about a few pages back and some others picked up too is to give the newer blocks in the interval a higher weight than the older ones.
Lets say we split the 24 blocks into 4 parts when calculating nActualTimespan.
older ---> newer
part1, part2, part3, part4 (each part 6 blocks)
a more conservative approach:
part1: block times * 1 (counted like 06)
part2: block times * 2 (counted like 12)
part3: block times * 3 (counted like 18)
part4: block times * 4 (counted like 24)
sum / 60 = weighted average
weighted average * 24 = nActualTimespanWeighted
a more agressive approach:
part1: block times * 2^0 (=block times * 1) (counted like 06)
part2: block times * 2^1 (=block times * 2) (counted like 12)
part3: block times * 2^2 (=block times * 4) (counted like 24)
part4: block times * 2^3 (=block times *
(counted like 48)sum / 90 = weighted average
weighted average * 24 = nActualTimespanWeighted
Splitting it into more parts would make it react even faster, both on the way up and on the way down.
Giving the newest i.e 1 or 2 blocks additional weight can even speed it up more; giving the very latest blocks too much weight could lead to an exploit for an attack.
It should be an easy mod to our DGW3 diff algo with lower risk than a completely new algo.