Bitcoin Forum

I was watching the Bitcoin show (http://onlyonetv.com/2011/09/the-bitcoin-show-episode-045-panel-discussion-thursday-with-yifu-guo-darrell-duane-jared-ostmeyer-and-steven-wagner/) and Yifu Guo mentioned having a 3 year window for bitcoins to be immobile, and then they go back into the mining pool, so that they can be recovered. This seems like a good idea with one caveat. That makes it less desirable to hold onto as long term investments, e.g. bitbills. So I'm thinking about a way around that.

I'd suggest a 10-20 year window, because it's not all that crucial that they come back into circulation, it's only for a time when most bitcoins are completely lost that there are perhaps only 1 million in circulation because 20 million of them have been lost. So I think after that amount of time we should allow those coins to be mined again.

However, I think it's important to allow those who do have a legit claim on old "lost" coins, to be able to access coins if they have the key. So I suggest there be a pool of lost coins that all coins older than the window go into. Claiming coins will be given priority, so if you have a "lost" key, you get the coins from that pool. As other coins are lost that pool will be continually fed, and with a 10 year window it will be increasingly unlikely that that pool will be depleted. The pool will be continuously mined normally at some equilibrium rate with the Loss rate and the Reclaiming rate.

Let me know what you think.

I'm not entirely sure about this as a solution, but I agree that bitcoins will not last without some solution in place.

I hear a lot of talk about how bitcoins will be around for a long time, but with the amount of people using bitcoins and amount of theft that has already taken place, we've already mined and/or lost a considerable amount. Some claim this is not an issue due to the divisibility bitcoins provide, but eventually 0.0001 bitcoin *may* be worth $1. Does this mean that after 1 transaction you lose $50 because of the 0.005btc tx fee? Or are we going to have to continuously update the system as long as bitcoins are around?

In all honesty I think this should be the number one priority if people are actually serious about considering this a long-term investment. Otherwise, it does just really seem to be a great ponzi scheme.

How does one determine if coin in a wallet is 'lost' or simply 'saved'?

As for the transaction fees: They vary.  As more coins are mined, more coins are exchanged, more volume on the BTC network. (If everything goes 'as planned')

If BTC ever gets to .001 = $1 USD, there will be so much volume that the fees will go down.

At least I'm pretty sure that's how it works.

also, how would you prove you have a 'lost' key if you've lost it?

If the coin hasn't been moved in 20 years, it's incredibly likely to have been 'lost' rather than saved. But again, in the very unlikely case that it's been 'saved', that's where the lost and found pool comes in. 'saved' keys can claim from the lost pool the amount of coins their entitled to.

If there aren't enough, then they'd have to wait until more coins enter that 'lost' pool.

The key wasn't lost, you've had it all along. But the system has assumed that you lost it since it hasn't moved in 20 years.

Demurrage would solve this...


Anyway, simply build into the client an automatic ‘transfer’ from one address to another in your wallet every year or so; the user need not even be notified. If an output goes stale for three years or more, that means the either the owner hasn't been seen in two years or the wallet is lost. The output is then up for grabs.

I'm implementing something similar to this in my demurrage-based test block chain. Demurrage fees are assessed as transaction fees when the coins are spent, but each time an output halves in value due to fees (takes about 20 years), those fees, rounded up, become up for grabs. Eventually lost coins are re-entered into circulation.

I see no problem with "lost" coins. They make the original bitcoin scarcity increase. I see no problem with alternate block chains. If their experimental properties prove themselves to be useful, then we will have alternative currencies to chose from.Is there any technical reason that the original bitcoin algorithm could not one day be modified with these properties if they are found useful? If so, then they are relevant discussions.

What problem does this solve? If divisibility is a problem BUT we're willing to modify the protocol, then why not make them more divisible?

it doesn't really solve any problem, actually.

the bitcoin economy will function exactly the same even if the entire world is only circulating 0.00001 btc total.

it would be no problem to create a client that sends/receives nano or pico coins, displays amounts and gets user inputs in nano or pico coins.

It fits the model of gold better. Gold doesn't get lost with no chance of being found again.

Why did Satoshi cap gold at 21 million? Was it arbitrary?

Yes I know the argument, any amount of a money is sufficient.

One benefit is it will keep the bounty on mining high, making it so that people will continue to drive the difficulty level high, making the block chain more secure. It does that without inflating.

my idea of the next gen cryptocurrency is this

1,000,000,000,000 (one trillion/T)coins total, enough for each person in the world to be able to have 100-200 each
2 or 1 min blocks, about 3 or 4-8 years per 1M blocks
100,000 coins per block,(100 thousand/K) no declining reward
instead coins are almost equally divided out of the entire chain once there are no longer any coins to get for free. old small amounts of coins are just removed, like for instance, over time an amount of 10 coins has been abandoned, and is now worth less than 1 coin, this coin will get taken over priority than dividing the entire chain out. so any coins that fit this criteria will always be removed before any division is calculated, so you may not even loose any coins if things work out in a certain way.
this encourages coins to be SPENT AND NOT HOARDED. spend em or loose em, that's how it should be. if you want to save money, buy gold or a house or something, don't ruin a currency.
mandatory fees(.0001+0.01%), they reduce the amount of division needed to do on the entire chain.

these ideas would only work long term if there was very serious mining competition, otherwise miners make way too much money, and the idea is very rough, and has some holes in it, but i think it could work, although the calculations required for just 1 block, even after you hashed it all out would be fairly high. but i also think this helps fix the problem of an ever growing blockchain.

I am convinced that there is a probability that over the long term, in order to keep the block chain from being a mandatory zillion terabyte download, the block chain will be managed by signed checkpoints (call them "superblocks") that simply summarize all of the account balances beforehand so that clients don't have to download and maintain everything from the genesis block to the current just to sell sodapop and bubblegum.  If that ever comes to be, I could see a policy for "superblocks" of dumping old account balances of 0.01 or 0.001 or smaller, perhaps the threshold also takes the age into account, simply so that they don't consume more resources to track than they're worth.

As long as every user must download the whole block chain and dedicate disk space to keeping track of every time joe miner received a payout of 0.00125134 from a mining pool as it is now, it doesn't serve any useful purpose to void any small transactions.

the main thing i am trying to "fix" is the extreme uneven distribution of currency. i think the coins should be moved around from person to person, or you pay for hoarding.

it also works out so that you can eventually get rid of some really old blocks and not even need a genesis block eventually, just some trusted block every few years or so included in your client. and because there are so many coins, you cant really justify keeping .0001 of a coin for more than a month, it simply is a burden on everyone else.

this would only ever be a temporary fix.

(i.e even if you 'fixed' this same problem in USD for example, 80% of the dosh would gravitate back to 20% of the people anyway)

Seriously, if you want to keep even distribution just spread it out evenly and stop having transactions at all.

and redistribute every time someone new pops out or kicks the bucket :D

Go green, recycle btc ;p

Demurrage solves that problem definitively. Do a search for ‘freicoin’ on this forum.

Some people say that it doesn't matter that coins get lost due to divisibility. But assume that an evil hoarder keeps his gold so long that all presume it to be lost. Said person could then, when showing everyone the true amount of availible bitcoins, cause a great the deal of instability.

right now, the possibility of lost coins leads to the risk of an incalculable profit for everybody owning bitcoins. the possibility of recycled coins on the other hand leads to the risk of an incalculable profit for everybody mining bitcoins.
so you basically want to put an awful lot of work into it and make the protocol more complex just for trading one problem that isnt really a problem for another not-really-a-problem.
thats so meaningless you could probably convince the european union to fund it  ;D

This can be fixed by share the demurrage uniformly.

Increasing divisibility means increasing the amount of space taken up by the block chain. It's wasteful. We already have a problem with downloading the block chain as it takes days. We may very well have to increase divisibility of the existing 21 million if they become worth a great deal more. Lost coins would only compound that problem.

There is no accounting for entropy in this system. That's a problem. The reason this solution is better than increasing divisibility when needed is because this only needs to be done once.

And I'd rather make it more profitable for people doing work for the system than for people holding coins and doing nothing for it.

Think of it like a storage fee.

First of all, the idea in the bitcoin show is not new. You may be interested in this thread:
https://bitcointalk.org/index.php?topic=9295.0

As some people pointed out, demurrage would solve this problem.
The only problem with the solution proposed here is with very durable stores such as bitbills. Apart from that, no one is punished, hoarders can move their coins from one address to another from time to time.

The proposed solution is also nice for storage (demurrage also solves the problem here).

Another good side effect is that lost coins would reward miners, something that some people say is critical:
https://bitcointalk.org/index.php?topic=6284.0

Freicoin also has perpetual reward for miners. Freicoin is different from maaku's proposal because it distributes evenly the demurrage charged between miners (after the maximum monetary supply is reached, the reward for miners is constant. Well the fees are still variable), instead of some of them being luckier than others. I still don't know why he wants his system to be like that.

Because my proposal trades that inconvenience to the miners with a reward for the user: those demurrage fees are treated as transaction fees, so their transaction is processed quickly and at less cost. These may be heretical words on this forum, but I'd rather reward the consumers in the economy than the miners.


But we should move this debate elsewhere if you want to continue it.

Then perhaps we should have demurrage for coins which haven't been moved in 10 years, rather than uniform throughout the bitcoin economy.

If all you are trying to do is recover lost coins, that's fine. But that's really more of an added benefit to demurrage, whose primary purpose is much more expansive.

for anyone who has read my post, can they say anything that is wrong with it, because i think it could really work.

A better Lost Coins recycling program:

Instead of losing all 50,000 coins at once if someone forgot about it for 3 years...

Make it so you lose only 10% of coins that's been idleing for 1 year

so an address with 50,000 coins would have 45,000 after 1 year, 40,500 after 2nd year... and so on

if this is indeed lost BTC, eventually it will gradually be redistributed into the econ again...

if this is someone's idle address, he would still be able to recover more than half of it after 5 years.

I'm not sure I understand your proposal. Is it demurrage only for very old or small accounts?

the basic idea is every bitcoin address will give up a % of coins to the person who makes the next block once all coins have been made. there is never a declining reward for blocks, coins will always eventually get recycled, and it discourages hoarding the currency. the amount you loose per block would be very small, but if you were to just keep your coins over the course of about 6-10 years, you would have nothing left if i did my math right.

additionally it resolves the issue of the chain getting too large, because over time, all the coins in old addresses will be gone anyway, so you only need a trusted person to include a valid block in a client that gets you started.

What you describe is demurrage (http://en.wikipedia.org/wiki/Demurrage_(currency)). We have a forum for a fork of bitcoin with this feature.
We've not implemented it yet. In fact there's many open discussions.
For example, you say "once the maximum base is issued", I prefer from the beginning. Most people want a 4% or 5% rate, you're proposing 10%. The curve of generation, the maximum base (some people think that it shouldn't exist and the supply of money should be elastic), charge the demurrage automatically or at the moment of the transaction (so that the demurrage acts as a tx fee for the miner instead of being part of a constant reward), etc.
If you're interested, please visit: http://www.freicoin.org/

the concept is very similar, although my idea has some major differences.

there is always 1T coins once they have been made.
there is no fixed %, you only get reduced by whats required
once you get below a x amount for y blocks, it just gets taken. this is to get rid of stray coins quickly instead of just deducting small useless amounts forever.

so the basic way reductions are calculated is by

1. take all coins that meet the criteria
2. 500,000-from #1
3. divide out what you need to make your 500,000 coins from each account.

this process does not start until the 1,000,000,000,000 coins have been made

i think it would also be beneficial to have every block consolidate every previous block. so instead of having a huge number of small blocks, just have about 10,000 big blocks(more if faster block times)

In your idea, you have 10,000 copies of a database, each with several millions or billions of rows.  And you need two do three passes through it every minute or two, and the third pass will require a hash of each of those millions or billions of rows.  And when a miner finds a new block to send out, everyone in the system needs to do several millions or billions of hashes just to verify it.

By the time that is feasible on regular computers, we'll have wristwatches running full bitcoin nodes and storing the entire bitcoin blockchain internally.

Not to mention that your system produces the starting conditions of every hyperinflation ever, namely the thought that "I must get rid of this fucking currency as quickly as I possibly can because it will be worth less in a couple of minutes".

And yes, I'm sure you can find ways to tweak the system to mitigate the technical problems, but I doubt that you'll ever be able to do anything about people's preference for money that gains value over time, instead of losing it.

my idea hardly resembles hyperinflation where you take wheel barrels down to the store to buy a loaf of bread. once all 1T coins have been made there is no more inflation. although it may look like it from your end.

i think that the average person would find that it would still be cheaper to use this currency than USD and visa, inflation and the fees would be far lower and thus cost less money. after all, most people spend most of their money the same month they get it. so they would not be paying much.

a new system could be added on top of the one i already suggested. you could put coins in a address that makes them not able to be removed. i guess there could be 2 types of addresses or something. you send coins to the savings type and they cant be removed for 1 year or so. also, coins can not be added to an address once it has coins. coins not taken out will suffer from the same demurrage as normal coins. coins can be put back into savings immediately. the reason you can not be able to set the amount, is to prevent abuse by setting either really short times or really long times with not intent on ever getting them back, effectively destroying them. this system would allow you to save coins without having to loose all your money in 6-10 years or whatever amount of time.

@ ctoon6
Apart from the additional work required from the miners that kjj points out, your system has another flaw.
Your system will not have a perpetual reward as you claim. People will just move their coins in the quantities and times you incentive. That's good for storage and lost coins, but doesn't solve the security potential problem.

@maaku
Your system doesn't warranty the reward to solve the potential security problem and neither solves the lost coins problem. Until the account get completely destroyed by demurrage, the system can't take the lost coins, because it's waiting for a transaction from that account to charge the demurrage.

If we have, for example, a lost account with 1000 coins, and your proposed 4% demurrage...

In 10 years
0.96^10 = 0.664832636, the account will still have 664.832636 coins
In 20 years
0.96^20 = 0.442002434, the account will still have 442.002434 coins
In 50 years
0.96^50 = 0.129885794, the account will still have 129.885794 coins
In 100 years
0.96^100 = 0.0168703194, the account will still have 16.8703194 coins
In 200 years
0.96^200 = 0.000284607675, the account will still have 0.284607675 coins
etc

On the other hand, my system will be reintroducing the lost coins through demurrage because it is charged automatically on every account no matter if there's transactions from it or not.

Yeah, but my proposal also includes a new transaction type where fees can be claimed as a bounty each time an output halves in value, solving that problem.

Are you aware that at 4% the bounty takes almost 17 years to be claimed?

0.96^16.97 = 0.500199006
0.96^16.98 = 0.499994857

Yes, I'm aware of that and consider it a feature. I don't think that lost coins is an issue, except in the very long-term.

I can understand that you're not much concerned with lost coins.
But do you really consider how your system (compared to mine) treats the lost coins a feature?

I do, because for those ~17 years the owner of those coins would have received a benefit from the levied fee had he chosen to spend them.

I'm sorry I don't understand. There's no owner to receive any benefit. The coins are lost. You have to be more clear.
Do you mean when the owner doesn't lose them? I understand that part. I mean in the specific context of lost wallets.

How do you differentiate coins in lost wallets from coins that just haven't been spent yet by their anonymous owners?

Answer: you can't.

I know. That's why (for this lost wallet case) charging the demurrage when a transaction happens or after 17 years instead of charging it each block cannot be an advantage.

I think we just want them to prove they have the coins. In order to avoid the demurrage charge, they must show control of the coins by moving them. I don't think that's unreasonable to require every 20 years to avoid that charge.

What maaku and I are proposing is charging the demurrage in every case (being the wallet lost or not). The difference is that I wwant to charge it automatically each block (the implementation is easy) instead of every 17 years or when the owner moves the coins.

I'm starting to like the idea of demurrage for bitcoin. Demurrage could help bitcoin circulate with its use-it-or-lose-it design and address the extra deflation caused by accidental bitcoin burning. The bitcoin network is very expensive to run and should not be used for long term savings, it should be used to drive economies. There are other investments that can be used for long term savings. On another note: physical bills can be created that are counterfeit-proof (if you trust the creator) and can have an expiration date-stamp applied.

Ahahah, this is so false! What about gold in ship that sank? What about the gold used in technological things? Good luck recover gold in billions of hardware around the world, and so on...

isn't bitcoin about freedom?
who are you to tell me, what i have to/can/cannot do with my money? why do i have to prove anything to you?

this is ridiculous.

if you want demurrage, create a new fork, so i can choose not to use it.

Because you are using other people's resources for your benefit, Bitcoin can tell you what to do in order to keep your bitcoins. I'm not a fan of demurrage for its own sake. Only for the long run purpose of recovering lost coins to keep the supply stable and memory costs lower. I want them to be able to avoid the charge. It's only moving the coins from one address to another. And it's only once every 20 years. The but hell, why not once every 100 years. It won't affect you in your lifetime and it will keep bitcoins stable in the long term.

I would rather bitcoin be changed, and that's why we discuss this stuff. We may lose the argument, but if we win, you'll still have the option of creating your own fork without it.

As to gold being lost in a sunken ship
http://www.msnbc.msn.com/id/18736741/ns/technology_and_science-science/t/deep-sea-booty-million-coins-found/

Hardware..
http://en.wikipedia.org/wiki/Computer_recycling

no, you are the ones who want to change things.
if bitcoin is changed the way you want it to, i have no choice, you force me to accept your rules, even if i don't want to.
that's the whole point.

in bitcoin there is no way to tell, if coins are lost or not <period>
if you want a way to tell if/how many/which coins are lost, you have to create a fork.

if i buy a BitBill (or some similar "offline-bitcoins") today, you want them to become worthless in 20/100/1000 years and i can't think of any reason to do that.
you want me to send my coins from address A to address B every 20something years and again i dont get it, what difference does it make? none

it's my money, i can do whatever i want with it.
if i wanna keep it offline for the next 50years and give it to my grandchildren, that's up to me,
if i wanna destroy it, i can and there's nothing you could do about it.

everythings fine as it is, if you want to change it, create a fork, don't force others to play your rules.

No. My proposal of demurrage "for its own sake" is going to be in another fork because most bitcoin holders and miners won't accept it.
But if people want to change things, the rules are hard to change, not written on stone. The change that he proposes would only force you to move your coins from time to time. I don't see how that can "hurt you" much.

Changes in bitcoin are possible and happen. For example, if we want to have more divisibility, we have to change the protocol.
To allow more complex transactions (https://bitcointalk.org/index.php?topic=38928.0), the protocol is going to be changed soon. You can try to convince us not to change it, but I don't think that you're going to be the majority.

If you don't like demurrage to restore lost coins and save storage space (and reward miners for longer and therefore have better security) you can try to change our minds. With arguments, not just saying "you can't do it because it goes against my freedom".
You're free to sell your bitcoins if you don't like the change. A change in the protocol doesn't happen overnight. I'm still waiting for namecoin to have merged mining, for example.

those kind of changes don't do any harm to my money,
yours would.

and i gave you a few examples how it can hurt me.
if i'm a non-techie person and want to "hoard" my coins offline, using BitBills for example, you want me to destroy them after X years (and import the BTCs to an online wallet), just to not lose my coins.
i bought those BitBills for a reason, i paid for those coins, it's up to me what todo with them.
no protocol-changes ever should change that, or at least should be on their own fork (maybe i would use that fork too, by choice, not forced to).

My original proposal was to allow recovery of coins from a lost coin pool if you did indeed retain the key, which would protect things like BitBills.

The only problem this might run into is if the lost pool doesn't contain enough coins to cover your recovery claim, in which case you'd have to wait for more coins to enter the lost pool. But they would eventually be made whole.

This is a misunderstanding of a social experiment. If the greater good deems it necessary to tear down your house to build a Walmart, then that is the price of living in a social experiment of laws. There are many investment vehicles for long term storage of value that you may use that will not cost me electricity, such as stock, gold and real estate, but then you will bear the cost of maintaining them. Bear in mind, that any changes to bitcoin will require very good reasons, and that is one reason why we have a forum.

I've got it.

Instead of having to move them, you could just send the address some coins. All you'd have to do is send the minimum transaction to the address to keep it active, which would push back the demurrage date another 20 years. I think that would be good enough to demonstrate control of the wallet. and you wouldn't destroy the bitbills.

That would work too, but without saving space, people have still to maintain the chain from that ancient block where the address first appeared in an output.

I suppose you're right, but one step at a time. this would be the most amenable option for now.