Quote from: LeoC on January 08, 2014, 11:41:48 PM

If the stars align by some miracle and this coin gets that type attention, then we will we see what happens.
Until then, you are theorycrafting, something you have done over and over and over without really bringing much to the "reality" table.
You have consistently shown that you have no idea how these altcoin systems work and constantly bring up far fetched coding examples into the equation such as the loyalty rewards.
I can say without a fathom of a doubt that your systems would fail immediately.

To fix the issue, you need to understand the problem. The problem isn't coin jumpers themselves. The problem is how the current coding reacts to these coin jumpers. In this case, the reaction would be a wild oscillation in difficulty. My proposal was simply to limit the range of this oscillation. This gives regular folk a chance to catch the train and hopefully stay on it. Anything else you would like to add?

No, the problem is neither the coin jumpers themselves, nor how the coding reacts to the coin jumpers. The coin jumpers are not a given. The problem is that the code specifically rewards coin jumper behavior, which creates the coin jumpers who are actually just people responding to incentive to try to maximize profits - and seen this way - who can blame them. If coin jumping is specifically rendered unprofitable, by implementing strategic coding changes, then by definition, profit-seeking behavior would no longer involve coin hopping into and out of Catcoins (or else they would be jumping in and out of mining for uneconomic reasons, and there is no evidence people who do this are a significant factor in what we are suffering). I submit that this is exactly an accurate model of how incentives cause coin jumping, which in turn cause wild difficulty fluctuations. Without fixing incentives, coin jumpers will continue to seek "maximum profits" by means of jumping in and out of Catcoin mining, which is a different modality from choosing the coin and mining it over an extended duration. This in turn must inherently destabilize the difficulty levels of the coin. It is a very simple cause-and-effect chain.

Quote from: LeoC on January 09, 2014, 12:02:01 AM

There is no way to combat coin jumping unless EVERY single altcoin got together and created some impossibly difficult shared code to prevent people from doing so.

This is untrue. If any one altcoin (or even any significant number of pools supporting that one altcoin) stopped paying for coin jumping behavior, the coin jumpers would respond to the new economic incentives and leave the coin out of their coin hopping activities. Just ask any coin hopper if they would move their gear to mine Catcoins, if they knew that during easy difficulty times, they would receive only a fraction of the coins, while those who had been mining would receive the full amount of coins. I can pretty much guarantee they will say no way and that they would stay away from the coin. No coordinating with other altcoins is required.

Quote from: LeoC on January 09, 2014, 12:02:01 AM

Those mean the same thing, so is the coding a problem or not?

There are subtle differences in meaning depending on how you use language. Your formulation implies that coin jumpers are a given - like they are changes in the weather - that they will inevitably come regardless of what changes you make to the code - like it will rain once in a while no matter what you do to your home - the best you can do is prepare for them. The only thing you can do is respond to the inevitable comings and goings of the coin jumpers then is by means of clever methods to adjusting difficulty in a way that outsmarts them, and frustrates their plans to maximize profits. My formulation by contrast specifically says that coin jumping is a volitional act on the part of those seeking to increase profits - that the positive incentive to coin jump is embedded in the code - and that this can be repaired by application of engineering principles - and coin jumping would stop once there is no economic incentive to engage in that conduct. Moreover, I would say when profits are involved, any attempt to cleverly outwit profit-seekers by any static formula would eventually by defeated by profit-seekers who will always apply creativity to find a way to beat the averages by understanding the rules and incentives set by the code. We would end up seeing the same problems over and over again in different forms ad infinitum, until the source of the problem is finally repaired (or the coin dies or lingers in obscurity for lack of such a repair).

Quote from: hozer on January 08, 2014, 11:38:52 PM

What do you predict if we retarget every block based on the last 36 block average?

With rational-only economic calculations, I think coinhoppers will still have an advantage, but since it's a gradual change, rather than a step change, I think the psychology and market dynamics will be dramatically different.

In a sense, it's hard to apply the model directly to this proposed difficulty adjustment formula because the model I am using presupposes some kind of predictable difficulty over time - but apart from that, and in a casual application of the model and observing the coin markets - I would predict the coin would have a very randomly looking profitability graph, and it would land at random times, for a couple of minutes at a time, at the top of profitability charts - and get random spikes of major hashing power - and long-term miners would see on-going profitability of less than LTC, while sophisticated coin hoppers find ways to monitor the blockchain and know exactly when to jump in and snipe the few coins that can be had before the averaging formula move the difficulty out of rage of profitability again. There are coins like Galaxycoin exhibiting this kind of wild fluctuation in difficulty/profitability, and the reputation of that coin was not helped by being known for this type of fluctuation. I do not see this as an immediately catastrophic change like, like the suggestion to cap difficulty level changes to something narrow (which has a good chance of killing the coin in the aftermath of the first significant spike in the value of the coin in the markets) - but I do see it can drain 1) the reputation of the coin from the fluctuation and 2) enthusiasm for the coin among miners who see consistent profitability below LTC. These may then prevent the coin from gaining momentum to see any significant value spikes to begin with - and would tend to linger long-term as a low value coin without much hope of increasing in value. But like I said, this one is hard to predict because it does not fit the model I am using, and this is partly conjecture on my part. I apologize if I misrepresented your formula or intent in any way.

Etblvu1