So, now we need to define these key coin properties:
1) the total supply to be initially issued
2) the block target time
3) The minimum and maximum coin stake age
4) the yearly PoS interest
1)Upto the general consensus, but 1 million should be good to start with?
2) 60 seconds?
3) Maybe weekly and 30 days?
4) 8% or 9%? ( i calculated if we use 9% per annum, total coins in 20 years would be around 20 mil)
1) and 2) seems to me reasonable. Shorter times gives no big advantage.
3) 7d min and 30d max looks good. Shorter periods (say 1-2 days) for minimum age decrease the security of the network significantly.
4) up to community. The note on supply in 20 years: the average coin age with 7d min/30d max when it stakes would be ~17 days. Thus, there are ~21.47 staking periods a year, or ~420 periods in 20 years. The 17 days interest is 0.09*17/365, or 0.004191781 (0.4191781%). Hence, the coin supply in 20 years would be 1mln * 1.004191781 ^ 420, or ~5.79mln. This is close to the theoretical maximum if all coins participate in staking. The actual supply would be smaller in 20 years as any spend transaction resets the coin age and not all coins participate in staking due to lost/not running wallets.
5) I am thinking, if we can limit the max number of coin that an account can keep? This will avoid the over staking of 1 account. Yes, i know what is the next question coming, what is stopping from a person creating multiple account. That would be one of the security measures. Instead of keeping in a single bucket, why dont diversify? My suggestion. I am treating RS as a bank account.
This question is a religious one: some people are happy to keep all their coins in a single address; other people see the address reuse a bad practice. I belong to the latter group. Though, the PoS coin implies the address reuse due to specifics of PoS implementation. Technically we could implement the tx rejection if the destination address total balance would exceed some limit, but then these checks should be implemented in tx & block conformance check functions. This makes a load on nodes quite significant (there is no such thing as a record of a current address balance, the current address balance is a sum of all unspent inputs).
6) Can the wallet come with 2FA? What do think you of this?
It would only make people overestimate the actual level of security: finally, a wallet file is encrypted with some symmetric encryption key, so if a hacker manages to intercept the encryption key then he has a full access to a wallet. The n-factor authentication is relevant there where are no access of a hacker to the system you are authenticating to and you have more than one (virtual) channels of communication with that system, e.g. a voice phone call in either direction, an HW solution to generate a one-time password, a SecurID or alike.
One more question: is the community going to stick with 'Reserve Share Unit' name for a single unit of coin to be created or the community would like to use more common word 'coin' for a unit, so the name of unit became 'Reserve Share Coin'?
I need this to know in order to prepare the wallet's source code.
I was thinking of removing "unit" and just keep RS. I am thinking of this dialogue "RS me x amount of coins", "dude, RS me something..".
Can someone be creative to change the word of RS to something better?
Well, the 'Reserve Share(s)' is also an option, what the community thinks?
outputs welcome!
