static const int64 nTargetTimespan = 5 * 60 * 2 * 30;
static const int64 nTargetSpacing = 1 * 40;
5 * 60 * 2 * 30 / 40 = 450
You're correct seems 450 blocks is the actual re-target.
The problem isn't just with the swings though, it needs to re-target faster so the coin cannot be subject to "strip miners". This can only be done by reducing the block re-target or yes putting a fixed percentage limit on diff increase/decrease. Many new coins now re-target every block to help prevent strip mining. By strip mining I mean when difficulty is low large amounts of miners jump on (collective pools etc) and sell the coin straight away, then as soon as diff raises they move on leaving the rest of the network stranded with high difficulty and a long time to go for the retarget (like now ugh). Since these strip miners also sell straight away into buy orders it becomes unprofitable to mine for those left and the coin actually becomes somewhat unusable for a very long time - like right now transferring coins can take multiple hours because of the high diff and low hashrate, blocks are not being found fast enough for transactions to be timely.
Network hashrate: 8,743 Kh/s
Time since last block: 4h 5m 30s //I've seen 10+ hours
Current difficulty: 12.31608598
Est. Next difficulty: 0.08143226 in 236 blocks /
16d 12h 36m at current hashrate.Kudos to the miners on the network at the moment
My recommendation would be to lower the diff re-target to 10 blocks or even every block. We want to support the network having the coin actually usable, for this we need miners finding blocks on time.
p.s CHNCoin was recently revived by implementing diff re-targets every block like that of Peercoin (in November), it's network hashrate was basically 0 and the coins nontransferable, now it's chugging along at 420+ MH/s -
https://bitcointalk.org/index.php?topic=322488.0Thanks to everyone for supporting the coin. This is a great community.
So far as I see it MinerMike is correct in everything he said.
As a matter of preference I prefer min/max limits rather than short re-diff changes. I don't like them too long either but a good balance seems to be the best approach. Either way the end result is almost the same.
As a matter of marketing/popularity/psychology I believe that having a coin occupy the top slot on the profitability charts for a short period of time every now and then is a good thing for the coin.
For the record I am not a strip miner. I was mining Global coin at a time when I was getting all the blocks. In other words I was the only one and the diff was high.
I will support Global Coin no matter what the decision is. I still think it is one of the best investments and undervalued coins available.
Big thanks to everyone mining through this high diff. Looking forward to the update.