Timecoin

[jtimon]

There are trillions of USD in circulation and billions of users this is not an indicator of a good system.

Ok. you're right. People using it doesn't mean it's a good thing.
Also this last sentence was unfortunate:

I guess all this people are just stupid.

I still think it has advantages for the users. The probability of "getting back" a freicoin you spent "soon" (for a good/service you deliver) is greater than for a bitcoin.

[gigabytecoin]

This is a complaint I hear pretty often:

Unfortunately the developer of bitcoin has set a maximum of bitcoins to ever be created, which will make this currency highly deflationary. The more users that will want to use bitcoin, the more they will have to compete for the use of a limited amount of coins. Coins will become worth more and more, meaning less bitcoins will buy more product. Those who will profit from this deflation are going to be the “first come” users, those who created bitcoins when it was still relatively easy.

I believe bitcoin is an interesting proof-of-principle for a user-generated currency, yet it is not a currency that can guarantee price stability. Because of the fixed total amount of coins to be created, the currency cannot be adapted to a growing market.

I think I have a pretty good understanding of why the bitcoin generation scheme is better, but I'm not going to discuss that now, my theories haven't convinced anyone as far as I know.

I propose we try it out. Modify the Bitcoin code to simply continue generating the same number of coins/time forever. You could even call it a time limited currency. At any given time there will be a limit on how many coins exist, maybe this will be god enough to lure some people who are on the limited/unlimited fence. It is still surely better than a centrally issued currency where we can't tell at what rate money will be printed or what the distribution rule will be.

The infrastructure that exists already for Bitcoin can mostly be converted in minutes and this applies to all infrastructure that will be created in the future. A BTC/TC exchange site would be trival to set up and be virtually riskless legally. This would let Timecoin piggyback on all the existing and future BTC to state currency exchange sites. Just go to a BTC/TC exchange then take your BTC to MtGox or wherever. If it catches on there will be direct exchanges I'm sure.

My point is that people don't have to be really really devoted to this idea to make it happen like the developers of bitcoin are. They only have to think that this one difference is even a slight improvement, everything else is already available, practically free. There are even a bunch of miners already set up that could switch on a dime.

If Timecoin allows for a growing market and Bitcoin does not we'll know for sure and don't need to worry about theory.

Empiricism!

Sounds kind of like namecoin, does it not?

[jtimon]

Sounds kind of like namecoin, does it not?

As far as I know, the total supply of namecoins coverge to a maximum like bitcoin's.
The only difference between namecoin and bitcoin is that the namecoin chain allows domain registration/sell transactions and the sells and registrations can be made atomically by paying in namecoins.

Anyone, please, correct me if I'm wrong.

[Transisto]

I did not read this whole thread but here's my thought,

Many people here think from a miner standpoint.
The fact is that generating these BTC is becoming polluting as hell.  (more than 1.5 Gigawatt)

That's one of the reason why a Gold based economy is not a good idea.

When block will become 25btc we'll see great service offers.

[abyssobenthonic]

I'm not fighting against other bitcoin-like currencies. In fact, I'm trying to maintain a list with all proposals, even thought I think is too early to support the competitors when the success of bitcoin is not secure yet.

I agree.  I don't think TimeCoin is in any way ready for prime time until a few reductions in the block reward have come in... until then, it's an interesting thought experiment.

On the other hand, I don't like much Timecoin.
It tries to solve the "deflationary bitcoin" problem, but I don't think the solution is good.
First of all, Bitcoin is only deflationary if there's economic growth.

BitCoin is guaranteed (assuming a non-zero long-term rate of key-loss) to have a maximum usable money supply at some point in time (after which, of course, the money supply will contract asymptotically to zero).  A contracting money supply is generally deflationary.

To avoid timecoin deflation, it issues a constant quantity of timecoins each year (instead of a decreasing quantitiy until convenrgence).
If we assume a constant growth of 1%, Timecoin will become eventually deflationary too, since the constant issued timecoins (50 per block) are a lesser percentage of the total supply each year.

The gross creation rate of TimeCoin never goes below zero... if no TMC are ever lost, 50/100 trillion is still greater than zero (the gross creation rate will asymptotically approach zero).  If there are periods where the rate of coins lost to key-loss is greater than the creation rate, the money supply will have a short bout of contraction before equilibrium is restored (barring a case where the loss rate monotonically increases... the only situation I see that occurring in is one where users simply abandon TimeCoin, in which case there's nothing that can be done at that point).

Contraction of the money supply tends to lead to deflation.
Increase of the money supply tends to lead to inflation.
A constant money supply (zero net long-term money creation) is the option least likely, IMO, to lead to deflation or inflation.

My proposal is Freicoin, but nobody seem to like demurrage nor Gesell here.

The issue with Freicoin is that it exacerbates the contracting supply feature of BitCoin (if it preserves the continuously decreasing block reward), as demurrage implemented by destroying coins is equivalent to losing coins through key loss.  Implementing demurrage through transaction fees is also likely IMO to have the effect of reducing transaction volume ceteris paribus as people hope that clients that don't enforce demurrage become predominant before they need to spend.

[jtimon]

The gross creation rate of TimeCoin never goes below zero... if no TMC are ever lost, 50/100 trillion is still greater than zero (the gross creation rate will asymptotically approach zero).  If there are periods where the rate of coins lost to key-loss is greater than the creation rate, the money supply will have a short bout of contraction before equilibrium is restored (barring a case where the loss rate monotonically increases... the only situation I see that occurring in is one where users simply abandon TimeCoin, in which case there's nothing that can be done at that point).

To be honest, I'm not much worried about wallet losses.

Contraction of the money supply tends to lead to deflation.
Increase of the money supply tends to lead to inflation.
A constant money supply (zero net long-term money creation) is the option least likely, IMO, to lead to deflation or inflation.

The key thing here, I think, is that artificial monetary inflation/deflation leads to unwanted results. I agree, a constant supply would be better that anything else but a magical formula that tells the system how big the reward have to be to maintain "stable prices". Too complex (if possible).

The issue with Freicoin is that it exacerbates the contracting supply feature of BitCoin (if it preserves the continuously decreasing block reward), as demurrage implemented by destroying coins is equivalent to losing coins through key loss.  Implementing demurrage through transaction fees is also likely IMO to have the effect of reducing transaction volume ceteris paribus as people hope that clients that don't enforce demurrage become predominant before they need to spend.

No. The nominal reward is constant in freicoin. The total supply will converge when demurrage fees equal the reward.
Implementing destruction of money supply (to be able to keep rewarding miners) by "destructive fees" would only solve the problem of wallet losses and, as you say, would discourage trading.
Although demurrage diminish some effects of price deflation, the main point of demurrage is to reduce interest rates, not purchasing power of money.

[abyssobenthonic]

The key thing here, I think, is that artificial monetary inflation/deflation leads to unwanted results. I agree, a constant supply would be better that anything else but a magical formula that tells the system how big the reward have to be to maintain "stable prices". Too complex (if possible).

It's impossible in practice to determine whether prices are stable or a reflection of differences in preferences.  Thus it's absolutely reasonable to adjust the components of a price index; it's impossible to know how to properly adjust said components, especially in an economy offering a plethora of choices.  Such adjustment is also an invitation to rent-seeking.

No. The nominal reward is constant in freicoin. The total supply will converge when demurrage fees equal the reward.

That's what will happen with TimeCoin as well, as long as key loss occurs at a non-zero rate.  Demurrage effectively imposes an extra level of key loss (so it may converge faster, but will probably also gyrate more wildly).

Although demurrage diminish some effects of price deflation, the main point of demurrage is to reduce interest rates, not purchasing power of money.

Is reducing interest rates necessarily a good thing?

[jtimon]

No. The nominal reward is constant in freicoin. The total supply will converge when demurrage fees equal the reward.

That's what will happen with TimeCoin as well, as long as key loss occurs at a non-zero rate.  Demurrage effectively imposes an extra level of key loss (so it may converge faster, but will probably also gyrate more wildly).

If the reward is greater than the wallet losses, the supply won't converge.

Although demurrage diminish some effects of price deflation, the main point of demurrage is to reduce interest rates, not purchasing power of money.

Is reducing interest rates necessarily a good thing?

I think so.
http://www.community-exchange.org/docs/Gesell/en/neo/part5/2.htm
This and the following two chapters give an explanation why "interests are bad".

[abyssobenthonic]

No. The nominal reward is constant in freicoin. The total supply will converge when demurrage fees equal the reward.

That's what will happen with TimeCoin as well, as long as key loss occurs at a non-zero rate.  Demurrage effectively imposes an extra level of key loss (so it may converge faster, but will probably also gyrate more wildly).

If the reward is greater than the wallet losses, the supply won't converge.

That's true presuming that long-term change in number of unspendable coins is not positively correlated with the number of spendable coins over the period.  I'd say that the case for that presumption is weaker than the case for a positive correlation.

[jtimon]

No. The nominal reward is constant in freicoin. The total supply will converge when demurrage fees equal the reward.

That's what will happen with TimeCoin as well, as long as key loss occurs at a non-zero rate.  Demurrage effectively imposes an extra level of key loss (so it may converge faster, but will probably also gyrate more wildly).

If the reward is greater than the wallet losses, the supply won't converge.

That's true presuming that long-term change in number of unspendable coins is not positively correlated with the number of spendable coins over the period.  I'd say that the case for that presumption is weaker than the case for a positive correlation.

I'm not sure I did understand you, but are you saying that 50 timecoins (or wathever the reward is) will be lost every 10 minutes on average?

[abyssobenthonic]

That's true presuming that long-term change in number of unspendable coins is not positively correlated with the number of spendable coins over the period.  I'd say that the case for that presumption is weaker than the case for a positive correlation.

I'm not sure I did understand you, but are you saying that 50 timecoins (or wathever the reward is) will be lost every 10 minutes on average?

If the number of coins lost in the average ten-minute period is generally increasing as the number of spendable coins increases (equivalent to a positive correlation between losses over any time period and the supply), then there is a number of spendable coins where the expectation of coins lost in the next ten minutes equals any positive constant you care to name.

If the rate is 0.00009% of spendable per block (which annualizes to about 0.5% per year, which I suspect is the right order-of-magnitude), then a reward of 50 TMC implies about 526 million TMC as the equilibrium supply; 2 TMC implies a 21 million supply at equilibrium.  TimeCoin implemented as BitCoin with a 2 TMC minimum block reward would not fork until the BTC reward drops to 1.5625.

FreiCoin isn't BitCoin plus demurrage... it's TimeCoin plus demurrage (or BitCoin plus constant block reward plus demurrage)

[jtimon]

That's true presuming that long-term change in number of unspendable coins is not positively correlated with the number of spendable coins over the period.  I'd say that the case for that presumption is weaker than the case for a positive correlation.

I'm not sure I did understand you, but are you saying that 50 timecoins (or wathever the reward is) will be lost every 10 minutes on average?

If the number of coins lost in the average ten-minute period is generally increasing as the number of spendable coins increases (equivalent to a positive correlation between losses over any time period and the supply), then there is a number of spendable coins where the expectation of coins lost in the next ten minutes equals any positive constant you care to name.

If the rate is 0.00009% of spendable per block (which annualizes to about 0.5% per year, which I suspect is the right order-of-magnitude), then a reward of 50 TMC implies about 526 million TMC as the equilibrium supply; 2 TMC implies a 21 million supply at equilibrium.  TimeCoin implemented as BitCoin with a 2 TMC minimum block reward would not fork until the BTC reward drops to 1.5625.

I think now I understand. The bigger the total supply, the more units will be lost (with each generated block). That's why the supply of timecoin would converge too: when the losses equal the rewards.

FreiCoin isn't BitCoin plus demurrage... it's TimeCoin plus demurrage (or BitCoin plus constant block reward plus demurrage)

Well, if timecoin and freicoin will have a constant supply and bitcoin a decreasing supply, maybe it's more appropriate to define freicoin as timecoin with demurrage.
I was considering that bitcoin and freicoin will have constant supply and timecoin won't.
The constant block reward of freicoin could just be deduced from constant supply (in the end) and demurrage.
I was ignoring the effects of wallet losses in supply.