Bitcoin Forum

So, I was thinking and came up with an idea.. I wonder if it is even remotely possible for a current or new alt-coin to implement.


FRC and TRC have both now been plagued by coin hoppers.. people shooting the difficulty to a very unprofitable level and then dumping for the next best thing.. leaving difficulty way high and the legitimate users of the coin in a rough spot.

Is it possible to have some kind of 'difficulty decay' system outside of the typical readjusts?

For example: TRC is now 55% as profitable as BTC to mine.. if there was a way for the network to see "oh gee, for the past 5 hours we have been at a 5x block time cycle, lets go ahead and start dumping down the difficulty to alleviate this issue"

Or maybe if other coins just adopted the PPC/NVC method of constant gradual adjustment., you wouldnt get this absurd jumps that screw everything up.

What are your thoughts?

FRC developers are indeed looking into improving the responsiveness of difficulty adjustments.

If difficulty goes back down quickly, you'd have other coin hoppers with ASICs jumping in. Then difficulty would rise again.

Is this potentially an issue with Bitcoin as well? E.g., there's an earthquake and 50% of the ASIC machines around the world are damaged beyond repair. How f'd is bitcoin?

/me runs off with some gassoil in search of asicminer's farm  ;D

The bitcoin testnet (https://en.bitcoin.it/wiki/Testnet) has a difficulty adjustment rule related to what you are describing.


Reseting back to the minimum would be too extreme, but a moderate adjustment down after a period of slow blocks could work.

I think bitcoin will be fine, but the half of the world destroyed by your "earthquake" would be f'd.

Maybe a difficulty could be based on the last N blocks found in T time?  Just a thought.

I proposed a more reasonable situation: Avalon Asics have a manufacturing flaw that they don't know about, but breaks the system after x number of hours. That's far more reasonable.

True, but they still wouldn't go down all at once, because their age-since-manufacture is staggered. For it to be really catastrophic, the bug would have to be based on the date or something like that. Hard for me to imagine such a bug.

True, but their age-since-manufacture isn't staggered very much. And when you average the some were built a couple days before others and some were received in the mail a few days before others, it's not such a large difference. We also know that the usage on the machines will pretty much be identical: IE: people who buy avalons will assuredly all be running them 24/7. So although failure would be staggered, it could potentially happen all within a small time frame.

Indeed it could be someone else's ASIC hash farm (Avalon did say their goal was to have a ASIC farm, not sell machines to people) that gets flooded or raided, or whatnot.

I still think an adjustment every block based on a rolling average block discovery time of the last X blocks where X is say some number of hours or maybe a days worth of blocks.

I think what you would see is that when hoppers jumped on difficulty would rise pretty aggressively and when they top out difficulty and blocks slowed, difficulty would drop quickly as well for network to resume normal operations.  I think this would likely discourage hopping because the ramp up of difficulty would be more rapid according to real time than current difficulty adjustment schemes and network will resume normal behavior quicker, so the hash rate pump-n-dumps will be less rewarding and less impacting overall.... well that is the theory at least....