Topic: New PoS/Proof of Activity Proposal (Read 1332 times)

Yes, there is an incentive to maintain a node with PPcoin occaisionally, but there is no incentive to run the node frequently.
 
1) You gain just as much from running a node at rare intervals as you gain from running a node continuously. The PoS mining reward in PPCoin is 1% of your accumulated coin-age. There are 0 fees paid to you. You could run a node continuously or you could run a node temporarily one year from now. Your reward is the same in any case. The only motivator to run the client is impatience.

2) You are vulnerable to hacking when you run the client. Thus if you want to maximize your risk-adjusted return, the best approach is to keep your node offline until you plan to cash out. This is particularly true if you own a lot of coins. In fact since interest is only 1%/year you may just prefer to keep your wallet encrypted and never mine at all.

3) The security of PoS blocks in PPCoin depends on how many other people are running nodes simultaneously. To do a 51% attack with respect to these blocks, you need to control 51% of the currently mining coin-age. Offline inputs can't be used to mine blocks and thus don't contribute to security.

4) As coin-age accumulates your coins became more and more powerful at mining. This helps some. However, accumulation is capped at 90 days. Assume the attacker waits 90 days to max out his chances. The attacker can then reorganize blocks if he has 51% of the coins that are currently online. The necessary stake for a reorganization is going to be a lot less than 51% of all coins. I'm a little worried by that.

To sum up, the basic issue is that PoS protection effectiveness is proportional to participation. Participation in PPCoin may be adequate to ensure security. I'm not sure. If so, then I should probably significantly weaken my participation incentives. They may be overkill. It is a hard choice to make. Right now I prefer to err on the side of caution.

[I think you could tweak PPCoin just a little bit and mostly fix the potential issue. (e.g. introduce a 90 day cap on how long interest can accumulate before collection, institute some theft protection for PoS miners)]

Thanks cunicula. When you put it that way, it does sound much more reasonable.

Isn't there an incentive to maintain an active mode with PPcoins? It's necessary for generating POS blocks.

Yeah, I think it would be awesome if Coblee IPO'd a litecoin fork. Even more awesome is that there would be no uncertainty about its future. I am very much against a gradual decrease in block subsidy as in bitcoin. That is completely crazy. Who knows when it will break as the subsidy falls? IPO'ing is cool because the currency doesn't need to change over time. If it works after it is IPO'd, it will likely work forever.

Three points:
1) We need active nodes for the system to work. Active nodes are the analogue of PoW in this system. Work is almost irrelevant. Thus the need for some levies to fund incentives for activity.
2) The inactivity levies are not as steep as you think. You have to compare them to something to have a useful measurement. Worst case demurrage taxes compare favorably with best case money supply taxes for national currencies. The worst case scenario levy also compares favorably with bitcoin's fees+subsidy.
3) There are other options for evading the tax besides maintaining the node on your home computer.

 - Active nodes are necessary. Maintenance of nodes needs to be incentivized.

Regardless of how security works, we need active nodes to maintain a P2P system. The strong incentives to maintain nodes are intentional. Based on discussion among bitcoin developers (see irc logs at http://bitcoinstats.com/irc/bitcoin-dev/logs/), the lack of active nodes in bitcoin is a significant problem. If you cannot convince anyone to maintain a cryptocurrency client out of altruism, then the zero trust system breaks down. You will always be trusting someone else to maintain the blockchain for you. That is bad.
The number of active nodes is slowly decreasing over time, see http://bitcoinstatus.rowit.co.uk/. In bitcoin, there is no incentive to maintain a node unless you are a PoW miner. Node maintenance is extremely cheap, but the costs are not zero. To fix this problem, I have created rewards for people who maintain nodes.

 - The levies are not as steep as you think. The PoW miners are already levying unavoidable taxes at a 25% annual rate (soon to be 12.5%). Absolute worst case scenario you pay one-fifth of what you are currently paying to maintain the network.  Best case scenario you earn some money by maintaining an active node.

Money supply growth taxes for holding national currencies are larger than 5%. Money supply growth imposes a tax equivalent to demurrage. i.e. 5% Money supply growth = 5% demurrage in terms of the revenue levied.  See this table for money supply growth rates around the world: http://data.worldbank.org/indicator/FM.LBL.MQMY.ZG
The world average money supply growth is a little over 9%. Even for countries which historically maintain extremely tight monetary policies (such as Switzerland) the average growth rate is still around 6%. The worst case scenario 5% demurrage tax is better than holding your money in Swiss Francs. Swiss Francs tax you at 6% via money supply growth tax. Other developed country currencies are somewhat worse than Swiss Francs. Developing country currencies are much much worse.
Bitcoin is currently printing money at 25%/year, much like a developing country. 5% is five-fold less. Can bitcoin credibly charge less than 5% of the money supply in fees+subsidy? I don't think so. Rampant double-spends will result. Most importantly bitcoin's 5% is unavoidable. The 5% here is avoidable simply by running a client or partially trusting an online wallet service (see below).

 - There are other options for evading the tax besides maintaining an active node on your home computer's IP.

1) You can rent a VPS and maintain your own remote node in the cloud. The remote node only needs to have your stake signing key. If the VPS is compromized, your theft risk is limited to 10% of your balance maximum if you own less than 0.002% of existing coins [that is at least 200 BTC right now]. If you own more, your theft risk is still capped at 20 BTC, provided you notice the theft immediately.  
Currently a VPS costs US$23/year. Recall that doing this can earn you money as well. Future prices depend on blockchain size growth vis a vis growth in computing power. Hopefully computing power grows slightly faster than blockchain size. If not, bitcoin will eventually collapse under its own weight.

2) You can trust an online service provider to maintain a node for you. The online service provider only needs your stake signing key. If the online service provider is dishonest and steals your coins via the stake signing key, theft risk is limited. The limits are the same as those for the compromised VPS. The online service is profitable because active keys earn interest. Presumably the provider takes a cut of these interest profits and passes the rest on to you. The interest rate depends on the number of proportion of dead nodes. Interest rates increase as the proportion of dead nodes increases. This negative feedback ensures a stable level of activity.

3) If you are just worried about exposing your computer's IP address, you can run the node through a VPN.

Finally, if 5% is really still too steep after you consider all this, the rate could be adjusted downwards. If theft risk is too great, this could also be adjusted downwards as well. The theft risk is there to make sure you trust anyone you delegate your signing key to store min(10% of your BTC balance, 20 BTC) for you. Decreasing penalties and lowering theft risk weakens security and increases convenience. Maybe I am paranoid, but I prefer very strong and trustworthy security to inconvenience.

Perhaps if you convince coblee he could introduce something like this in litecoin , or a version of it. After all ppcoin exists for bitcoin so why not a scrypt based pos.

Ive often wondered if you could IPO a coin on a stock market where the "shareholders" receive the demurrage as dividends.

I'm trying to get a better understanding of the demurrage fees. Is it 5% flat per year for any public key address that doesn't maintain an active node? That sounds very steep, and also resource intensive to avoid as every wallet holder would need to keep a computer with an active connection running.

Thanks. I appreciate that.

You seem to be quite literate in technical terms for an economist!

I'm intrigued and I greatly appreciate your creativity and contributions to the field of cryptocurrencies.

I'm a developer by trade but haven't done any work with cryptography or currencies so I doubt I'd be of any use, but I'd love to help if there's some way for me to be useful.

Thanks, please help spread the word. It won't get done until competent open-source developers become interested. Help make that happen!

If you are a developer yourself that will really help. I have trouble communicating effectively with developers because I am an economist.


 

Great work.

I hope and believe that some kind of POS have to be done.

"Recall also, that the necessary proof-of-work investment is much smaller than the proof-of-stake investment."

Does this not depend on whether the coin is Bitcoin or an Altcoin.

"At the same time, the attacker faces costs related to losses on bitcoin-specific investments which are necessary for the attack. It can be assumed that a malicious attack causes the purchasing power of bitcoin to fall to zero. Under such an attack, the proof-of-stake monopolist will lose his entire investment. By contrast, a malicious proof-of-work monopolist will be able to recover much of their hardware investment through resale"

This implies that when it becomes necessary to use ASIC in order to mine bitcoin, those ASIC investment will be a kind of, proof of stake,
if they can´t be resold after a Bitcoin attack unless its just a matter of a simple change of a chip.

Ofcourse real POS rules would also mean an economic loss of the loss of value of coins owned.

In case anyone is interested, I updated the proof of stake wiki with a new blockchain design proposal. I believe that it would be extremely secure. Explain why I am wrong.

https://en.bitcoin.it/wiki/Proof_of_Stake#Cunicula.27s_Note

Even if you don't agree with me at all, perhaps something interesting is in there!