If the coin has a goal of 2 minute transaction times, then to me it seems logical that the code should attempt to achieve 2 minute transaction times as regularly as possible.
IMO a 720 block retarget is too long to achieve that goal in an environment where the coin is competing with many others for hash power, and large miners and pools can retarget hashpower at will.
I don't believe it benefits anyone when a coin gets stuck at a high difficulty and it subsequently takes days, weeks, or months to mine through all the remaining blocks. This can and has killed coins, or forced them to hard-fork. It also has the tendency to kill the coins value because it takes a very long time for transactions to get processed and because miners stop seeing any profit from mining it.
Essentially the high difficulty is NOT responding to market forces due to an artificially long retarget time. This is an imbalance in the market, much like a governmental law or regulation that gets in the way of market efficiency. I would think that any ronpaul supporters reading this should grasp this intuitively.
So to my mind, the coin will be best served if the code is modified to use a faster retarget, something like 10 blocks. Or even every block for immediate feedback. Well, I am not an expert.... does anyone know if there is a problem with retargeting immediately?
IMO a 720 block retarget is too long to achieve that goal in an environment where the coin is competing with many others for hash power, and large miners and pools can retarget hashpower at will.
I don't believe it benefits anyone when a coin gets stuck at a high difficulty and it subsequently takes days, weeks, or months to mine through all the remaining blocks. This can and has killed coins, or forced them to hard-fork. It also has the tendency to kill the coins value because it takes a very long time for transactions to get processed and because miners stop seeing any profit from mining it.
Essentially the high difficulty is NOT responding to market forces due to an artificially long retarget time. This is an imbalance in the market, much like a governmental law or regulation that gets in the way of market efficiency. I would think that any ronpaul supporters reading this should grasp this intuitively.
So to my mind, the coin will be best served if the code is modified to use a faster retarget, something like 10 blocks. Or even every block for immediate feedback. Well, I am not an expert.... does anyone know if there is a problem with retargeting immediately?
You can't do that like this, you have to add more blocks to the formula to avoid totally random difficulty changes.
Don't forget, block target is a goal only achieved once you average hundreds of block at a given difficulty/hashrate.
Each block still has a lot of variance in their resolution time when compared one to another : you can wait 10mn for a block, and then just after only wait for 10s, without changing the hashrate.
Fedoracoin has a 10 blocks retarget time, but when you have a stream of 10 lucky block found, difficulty can triple, and you have to wait sometime 30mn before having a more appropriate difficulty, when in an ideal configuration you should only wait 10mn.
720 blocks is too long, right, but you have to take into account at least 100 blocks to have a more stable difficulty. Independently, you can recalculate it more often, even every block. You just have to at least calculate an average over a sufficient amount of previous blocks.
But I'll say it again, it's far too soon to think of a fork, when this coin is only on one single very low volume exchange.
