You could also lose $81,000 in cash by setting it on fire.
Care should be taken with any large currency transaction, digital or otherwise. Though the delicacy of data (from a consumer storage standpoint) is definitely a hit against the concept.
correct me if i'm wrong (i don't know enough about bitcoin), but wouldn't it also be the case that he would've lost his money if he shut down the live CD system before backing up his data?
edit: 1 hr later, upvotes with no correction, so i think "delicacy" does cover it. why would you send $81k to a system that could lose it with a power outage, crash, accidental shutdown, etc?
My point was that 'delicate' isn't nearly strong enough of a word. If you can make diligent backups and still lose everything, the system is beyond 'delicate'.
Another post notes that newer clients have fixed this issue by pre-generating addresses. That's better, but being able to lose all your money because you lose a number just isn't acceptable.
well, i'd expect clients to eventually evolve that allow you to automatically back up your stuff. the currency and concept just isn't that evolved yet.
Serious question: If bitcoins can be destroyed so easily (either by accident or deliberately), how can anyone know the actual number of coins in circulation?
But the market places a higher value on a single btc once they are "out of circulation" because money is scarcer, right? So the only person losing money is you.
Yup. That's a feature of currency systems that are pegged to a commodity. The difference is that instead of coconuts or gold, the scarce commodity is a bunch of random numbers.
The are many problems with such systems, which is why the entire planet moved away from them starting about 100 years ago. Just like the initial users of BitCoin who are sitting on millions of the things are limiting the bitcoin supply in order to slowly sell them for real coin.
Ah, and given there is a known limit on the future amount of Bitcoins in existence, the system may very well end up with no bitcoins at all if all are lost... right?
Highly unlikely. It's more likely that the value of bitcoins goes up slightly with each minute loss of bitcoins, and people start using them in smaller fractions.
The total number of coins held by every BTC address are always known because it's part of the public block chain that every participant in the network holds a copy of. The guy's 8999 BTC are visible in this chain, he just doesn't have the private key to the address, so they are unusable. Counting the total # of BTC that has been awarded is simply a sum across the block chain.
Right, but they've essentially been taken out of circulation. So without the knowledge of how many coins are usable, how can one accurately value the currency?
Maybe this is the wrong mental model, but I picture it like a hard disk slowly filling up with bad sectors. Except in this case, the disk controller can't tell you how much free space the disk has.
It's more like hiding a treasure chest somewhere in the ocean (of cryptographic tokens) and then losing the map (private key) when your ship (filesystem) sinks. (Arrrr!) This is something that happens with physical objects of monetary value as well, including cash itself.
So without the knowledge of how much cash I'm hoarding in my closet, how can you accurately value USD? I don't know how much of the answer is “you can't, exactly, but there's enough other trading that it doesn't matter” (which could apply to Bitcoin as well) versus “the value is fixed by external factors such as the US government”.
I don't understand. What does the "send 1 BTC to yourself" bit of this process do? I'm confused by what exactly is happening that makes the bitcoins disappear. What are you supposed to backup, if not the "wallet"?
To me, it sounds like while its sending the BTC to you and the blockchain is being built/verified, the client somehow puts your wallet.dat in a limbo state where it doesn't have the correct data in it.
(I am not 100% how bitcoin works, but that's how I read it)
When you do a transaction you always spend ALL of the coins in that address, part of the money goes to whoever you're paying, and there rest is returned to you an a new 'change' address. (This helps your anonymity as someone watching cannot tell which amount you spent and which amount you kept.)
Before transaction:
Address1: 9000
Address2: 0
After transaction:
Address1: 0
Address2: 1
Address3: 8999
It sounds like you have wallet.dat that contains the private keys for Address1 and Address2, but not Address3.
The problem seems to be he didn't backup after he initiated the transaction. It's news to me (and apparently a lot of other people) that you can lose bitcoins in this scenario.
One proposed solution is to pre-generate some number of addresses that get stored in your wallet long before you actually use them, so they'll get backed up.
Thanks for the explanation and link. Reading it was like a train wreck...it's awful but you just can't look away.
I don't see how that can be considered in any way a reasonable limitation of the system. I had no idea the "coins" were so fragile.
Every time I think, "Hey, maybe it's time to start accepting bitcoin for our software. It can't hurt, and might even help spur adoption of new digital currencies." Something crazy comes along that makes me think bitcoin isn't so well thought out, after all. I was already a little worried about the "money" being stored on our world-facing system. I'd already assumed I would need to transfer to a private account every day or something to insure that a break-in of our web server (which has never happened in our 6 years of running it, but I know enough about system security to know that a determined attacker could eventually find a way in; which is why we don't store credit card data, software signing keys, or anything else particular sensitive on that server). But, this makes me think bitcoin, as implemented, all but guarantees coin loss, given enough time using the system...possibly very large losses, as in this case.
Seems that a system crash during a transaction would also lead to the same results. And even one bit of data corruption during saving data would do the same.
But, I guess this is the pain one experiences being on the bleeding edge. It wouldn't be called that, if somebody didn't bleed now and then.
It turns out this happened a year ago and the proposed fix I mentioned has already been implemented:
"Since this post a year ago, this feature has been implemented as keypools=100. Personally, I think it is a bad idea and does not address the problem, only pushes it off to the 101st address. At that point, users will come to expect certain backup behavior and then one day (presumably when they have more 'real' rather than 'play' money) it doesn't work as expected.
Unless the pool is recycled (change is returned to a random or cycle of addresses) then this is far more dangerous.
I propose, instead, what is expected. The change should be returned to the same address that the BitCoins were sent from.
I understand this decreases deniability/anonymity. But if someone really is paranoid, they should really be laundering money through multiple addresses in random amounts at random intervals. Sending change to a new address is just an obvious 'paper trail' considering all transactions are public, it doesn't take genius investigator to follow the money."
What's stopping a malicious client from just generating a lot of tiny transactions to fill up the block chain and overwhelm all the other clients that perform the verifications?
I understand that the current database with all the transactions takes up a few hundred MB (and this is why it can take a while to verify a bitcoin transaction).
Couldn't a small number of clients attack the currency by DOSing the network with transactions?
I believe the idea is that a small transaction fee will be expected by the majority of miners, which will make a flood of transactions costly.
Miners are free to include/exclude your transaction in their block regardless of whether a transaction fee was included. If the majority rejects transactions without a fee then it will take a longer time for the transaction to be included in a block.
i don't understand how he lost all his money. aren't all his transactions still stored in the block chain? the network should still agree that he has his money, as he hasn't sent it to anyone (other than 1 BTC)
The transactions are in the block chain, but the key to make new transactions are in the wallet.dat.
That being said, I dont see how any bitcoins were lost given the example in the post. When the wallet from the flash drive was restored, he should have access to the coins again.
Perhaps by spending the coin you put your backup copy of the wallet in a state where it disagrees with "the network". In which case he's making a very important point: backups are not enough, they have to be realtime synchronized backups.
Your key pairs are still valid, though, and they're in the backed up wallet. If those 9k were purchased and a few confirmations were achieved, then everyone with the current block chain will recognize your keys.
That the client is problematic doesn't mean the data/$ is gone.
Except he doesn't have the keys, because the transaction generated a new keypair for the remainder left post-transaction. The new key didn't get backed up because it didn't exist at the time of the last backup.
Absolutely correct. There is no way you can lose your Bitcoins as long as you have a backup of your wallet. The only way to lose Bitcoins is to:
A. Send them to someone else, which will create a new block in the chain that is confirmed by miners.
B. Lose your wallet.dat completely and therefore your private keys forfeiting any way to access BTC that are yours.
A very legitimate way of backup up BTC is to just copy your wallet.dat and blockchain to a thumbdrive and store it in a safe deposit box offsite. If you ever lose your main computer for any reason, restore the backups from the thumb drive, let your blockchain catch up to current, and all your money is still intact.
From the forum post referenced, it appears that this is untrue. If I backup my wallet today to a thumb drive, then spend 1 BTC of my money, then my backup is worthless as it will only have the keys to an empty address. The only way to backup your wallet safely is to store a backup after the most recent transaction.
That is not true. Your wallet.dat doesn't contain your money, the blockchain does. Your wallet.dat just contains the keys to validate that you can spend the money that exists in the blockchain.
If you restore your wallet.dat and let it catch up to the current blockchain, you will still have all your money, unless you really spent it.
edit: Ok, I guess I am mistaken - I see now how this happened with a really old version of the client. Newer versions of the client pre-generate several addresses in your wallet.dat, so even if you only restored an old version of wallet.dat, you would still have access to any money that is in the blockchain that you spent in the recent past. Of course, you should never let your wallet.dat fall too far behind.
If you back up your wallet prior to the transaction of 1BTC, it will not receive the private keys to the change address. Your private keys in your wallet.dat will point to an empty account. That is what this entire post about losing $81,000 is about.
The problem was, that transactions can have "change". So by sending 1 BTC, the client did in fact send 1 BTC to one address and the remaining 8999 BTC to a a newly generated address which is automatically added to the wallet. Since the backup was before this transaction, the backup didn't include the private key to the change address and 8999 BTC were lost.
A really unfortunate incident, as the person simply didn't know this technical detail. It prompted some upgrades to the client, which now generates addresses in advance and stores them in the wallet. When new addresses are needed, the ones from this pool are used first. With a recent client this accident would not have happened. It's still important to know though, that after _sending_ Bitcoins, a wallet might potentially include new private keys that are not in an old backup (if the pool has been completely used up since the last backup).
By the way: All this happened in August 2010, so this wasn't $81,000 at the time.
This is more a rule of thumb for the receiver; if you send a transaction (based on coins which are already well-confirmed), and don't try to send any conflicting transactions, then it shouldn't disappear as long as it gets one confirmation (even if the block chain forks, it will be resubmitted)
So the real post should be: How to lose $81,000 in Bitcoin.
Step 1, go back in time to August 2010, and use the old client
Step 2, get $90 worth of Bitcoin, and don't take proper
precautions with your wallet file.
Step 3, wait for Bitcoins to go up 90,000% in value.
The current client prefetches 100 keys, for your protection, and has many other safeguards in place.
Bottom line - Bitcoin software is becoming more robust as Bitcoins are becoming more valuable.
What happens to the 8999 remaining bitcoins? Does a server somewhere think they are allocated to a person (although it doesn't have any name attached to them), effectively taking them out of the pool forever? Or are the bitcoins forgotten and free to be mined again or allocated to somebody else?
Or is this just a question that's so bad it's not even wrong?
Bitcoin does not really understand people or property. It just understands public and private keys. The 8999 remaining bitcoins are presently signed by a public key that has no corresponding private key -- this means they cannot be used in any transaction. Yes, they are effectively lost forever.
What amount of computational power would it take to brute-force recover that private key? For $81,000 it would be worth throwing some considerable hardware resources at it.
I too was considering some Bitcoining but have these nagging concerns:
1. Seems Mt. Gox is the primary exchange of choice of Bitcoins. What mechanism or oversight in Mt Gox prevents Bitcoin exchange rate manipulation by way of wash trading? http://en.wikipedia.org/wiki/Wash_trade
2. The exchange rate data on Mt Gox shows the value of a Bitcoin shot up 10x in about 3 weeks. Hyper deflation to that extreme is not a desirable attribute for any currency.
One economist's 'hoarding' is another's 'capital accumulation'. Note that in order to accrue any form of money, some valuable things must be provided in return (assuming voluntary exchange). So while the hoarder may have pleasure from owning digital or physical bills of exchange, everyone else gets to benefit from real goods and services that have been rendered. Of course, those hoarders that really remove money from circulation are a very small exception.
If there is belief a currency is going to deflate it puts the brakes on spending. Why buy widget x today when you can buy two widget x's next month? Most people will just wait. The hoarding effect is a natural phenomenon, not necessarily a black/white thing.
I understand people use bitcoins for services rendered and a fairly limited number of products - but that is clearly to the benefit of the seller. It would be nice to see some statistics on bitcoin users actually using them as a currency. I'd wager it's less than 10%, esp. now they've become a media darling and a vast number of folks are looking at these like guaranteed lottery tokens.
There are some expenses you can't live without like food, rent, clothes etc. Holding cash in a deflationary (and non fractional-reserve) currency may be the equivalent of a savings account in a bank.
But hey, if people really did decide to lead frugal lives in order to save money, capital would expand faster in areas more essential to human existence. Consider it as a shift toward long-term goals rather than short-term gratification. Of course, the longer the timeframe, the harder it is to predict future demand, so there is a limit to how future-oriented people can get.
Your comment made me realize that computers have long exhibited this property. It was always nerve racking buying a computer because you could get one twice as fast in 18 months so people were hesitant to buy. This always appears to be true of almost all new developing technologies. In fact, all of the fastest growing industries I can think of exhibit this property.
Sure deflation is part of the Bitcoin design but the actual real-world value of a Bitcoin is determined by traders on Mt Gox, and as an exchange Mt Gox doesn't give the impression of being able to prevent common price manipulation schemes.
How long until someone develops a service that will store and back up your bitcoin wallet for you? How long until that service becomes the "central" place for people to manage bitcoins?
Given there is an absolutely finite number of bitcoins doesn't this mean that given enough losses over time we will in fact run out of bitcoins with no mechanism (by design) to replace the supply, adding further deflationary pressure?
Given we're talking about bitcoins this discussion is about as practical as discussing a russian tank invasion of western europe but it's interesting none the less.
Yes it does indeed. However, should you ask such a question on the bitcoin forum it will most likely be brushed aside with a reply of "the net effect is negligable". When in fact, I would argue it is not.
If it is practically infinitely divisible, then doesn't that mean that it is not truly deflationary? For example, if there is only 1 bitcoin left, and it is infinitely divisible, then it is no longer deflationary...?
That wouldn't be inflationary because no more money is being created. No matter how many times you divide that one bitcoin there will always only be one bitcoin when you add it up and thus would still be deflationary because it is still increasing in value.
I suspect the bitcoin supporters actually love this argument, because it's easily rebutted (effectively, "the market will just move the decimal places to account for that") and is not in fact the most likely reason why bitcoins will not work. That sideshow is a lot easier to deal with than the economic problem, which, to steal Tyler Cowen's words, is:
"It is a privately created fiat currency, bundled together with an anonymity scheme for transactions. The anonymity scheme means there is some reason to grant one-time seigniorage to the original currency issuers. Eventually the anonymity scheme, in some form or another, will be available without the fiat currency. At that point, or more likely before then, the fiat currency will fall to near-zero in value."
It is a privately created fiat currency, bundled together with an anonymity scheme for transactions. The anonymity scheme means there is some reason to grant one-time seigniorage to the original currency issuers. Eventually the anonymity scheme, in some form or another, will be available without the fiat currency. At that point, or more likely before then, the fiat currency will fall to near-zero in value
Mr. Cowen is somehow overlooking the relative scarcity factor. Speaking strictly about bitcoins and dollars, we know how much bitcoins will be there in the next 100 years, and we have no idea how many dollars will be the in 2113?
The disconnect here may be that he isn't dignifying the notion that scarcity automatically connotes value, in the sense that for instance my toenails are also more scarce than dollars, and always will be, but are not more valuable than them.
US Dollars are backed by the US government who requires that taxes be paid in them, and pays them out to employees.
That's a lot of trade outside of US which is done in USD and not in local currencies. Are you going to argue that USD is not used as medium of exchange outside of US because you cannot pay taxes with them?
No, that's not his argument at all. You can't pay taxes in Bitcoins in any industrialized economy in the world. Nobody is arguing that the Euro has no value, just that Bitcoin won't.
Bitcoins may have value for the anonymization, the lack of fees, and the ease of micropayments. It's not essential that a fiat currency is backed by a government, it's just the easiest way to create a fiat currency. Bitcoin right now is essentially backed by those merchants who accept it in exchange for goods and conventional currency. Bitcoins have value RIGHT NOW due to this backing, at least partially. Their long-term value is less clear, of course.
Efficient transactions schemes are not necessarily good stores of value, as Cowen points out; more importantly, if that's primarily where Bitcoin's value is derived from, Bitcoin will be worth zero when a better transaction scheme arrives.
Efficient transactions schemes are not necessarily good stores of value
Yes, but universally accepted scarce tokens are good stores of values. And currently dollars and euro, and almost every other currency are just that. Universally accepted scarce tokens.
And bitcoins are both an efficient transaction scheme and a scarce tokens. And, by design, a lot more scarce. Of course, it has not yet bootstrapped itself, but I see that as a possibility. And damn convenient possibility, too.
Dollars are backed by the full faith and credit of the United States Government, which is why they are universally accepted, and which is why US Treasuries establish the risk-free rate.
Bitcoins are accepted by online gamblers in much the same fashion as casino chips, by ideologues offering marginal services as political statements, and (allegedly) by criminal enterprises.
If bitcoins are mostly interesting to you as a way of reasoning through how economies would work in a world without established governments and economies, that's fine, but could you just say that? Because it's hard to pick out which of your arguments are pie-eyed what-if's.
You can't pay taxes in Bitcoins in any industrialized economy in the world.
Taxes is not what a money makes. Colonials, Greebacks, French Assignats. They all were used to pay taxes. On the other side the Somali shilling is (still) used as money. The russian (tsarist) ruble was used as money during 1918-1920 Civil war. Hell, shells were used as used as money.
As for taxes in bitcoins, it can be just of matter of time. You know the tune: Cheaper to tax than to fight
Your argument then is that despite the fact that the participants of the largest economy on Earth are effectively forced at gunpoint to use US dollars, dollars aren't necessarily any more valuable than cryptographically random numbers, because other currencies have in the past stopped circulating.
Yes, and no. Yes, because, the taxes are not the magic wand that your paint them to be. No, because, I will not recommend anyone to put all their savings in bitcoins.
Also, define valuable, and we'll talk. I can easily foresee situations (even now) when bitcoins are valuable to me than dollars.
Mr. Cowen is somehow overlooking the relative scarcity factor.
But you can create an arbitrarily high number of bitcoin-like electronic currency units, simply by taking the exact bitcoin implementation and changing the genesis block. Bam, Patiocoins, with guaranteed scarcity, anonymity, and a new goldrush phase -- so if you got dumped rather than pumped, you have a new opportunity to start at the top of the pyramid scheme this time.
The equilibrium is that sooner or later expected returns on starting a new distributed pump-and-dump dwarf expected returns of getting in late on bitcoins. And then poof.
Couldn't I make the same argument for choosing another precious metal other than gold and convincing the world treat that as money. It seems to me that the only argument against it is that historically it hasn't happened. (except for cigarettes, Yap Rai, silver, platinum, and others I'm probably forgetting.)
It's somewhat arbitrary which precious metal we use, but it's not entirely arbitrary.
There's the usual reasons we use e.g. gold for money. It's fungible, ductile, rare and essentially impossible to forge. That rules out cigarettes and Yap rai as stable currencies.
As far as gold vs. silver vs. platinum, the inflation rate (i.e. the rate at which new metal comes out of the ground) is also important. Gold inflates slower than silver, simply because there's more silver in the ground.
As far as gold vs. platinum, it's important to look at what happens if the world decided to move to gold or platinum as a currency. Gold is already used as a store of value, while platinum isn't. If everyone moved to gold as currency, it would become far more valued. If the same happened to platinum, its value would go stratospheric. There are essentially no platinum reserves (compared to gold), so its price would be even more volatile, and its price would spike far more than gold.
Platinum is also much more concentrated, geographically. There are like 5 working platinum mines in the world, and like 4 of them are in South Africa. South Africa has more of the world's platinum than the middle east has of the world's oil supply.
Those objections are not insurmountable, but human civilizations have been using gold & silver as currency for thousands of years, but the same is not true of platinum. Don't underestimate inertia.
Currency has, in fact, never been a wholly stable affair. Before paper money, allowing non-gold monies was a way of inducing inflation. But silver or platinum or cigarettes are clearly exact matches for gold while Patiocoins could automatically be that (and Patio is "famous" - that should be enough for a few people).
So this should only reinforce the point that money and currency can't just exist as a "natural" abstraction generated by an algorithm but rather must depend on social conditions (any society's currency system is transparently dependent on the overall configuration of that society's member's beliefs about the world).
But you can create an arbitrarily high number of bitcoin-like electronic currency units, simply by taking the exact bitcoin implementation and changing the genesis block
Yes, you can. If your X-coins will be in some respect better than the original ones they may even supplant bitcoins. But you'll have to put of effect in order to convince people to use them. With uncertain results. Currently bitcoin has network effects working for bitcoin and against other crypthographic/p2p currencies. To change that will require a major effort.
The history of bimetallism (gold & silver) is relevant in this respect. The final result is that your are better using the medium of exchange the majority using, so usually there only one money left.
Is anyone else concerned that the number of Bitcoins in circulation could be an issue?
I'm fully aware of the explanation that it doesn't matter when Bitcoins are lost because they are divisible to 0.00000001, but I'm not convinced that it makes sense to be doing business on that scale.
I see this as a psychological problem. It's hard for most people to really conceptualize value of very small numbers. Yes, people will just be converting to "real" currency to get an idea of what the value is, but shouldn't a successful currency stand on its own? Shouldn't we be able to understand the numbers we're dealing with?
> It's hard for most people to really conceptualize value of very small numbers.
That's why we do all the tricks to workaround the problem of long numbers. We're measuring the length in ~6.68458134 × 10-12 AU, time in 9,192,631,770 periods of radiation, etc. etc. all the time. We just call them 1 meter and 1 second instead. Once a "reasonable amount" of BTC reaches either 1k BTC or 1m BTC, someone will come up with a new, cute and short name for those - I don't think we'll have a problem with small / big numbers here.
It's difficult to determine the real volume of BTC transactions because of the structure of the Bitcoin Algorithm.
Bitcoin transactions have two outputs. In order to send a fragment of money, you need to create one output for the fragment, and the remaining output to put the "change" into your own wallet.
For example, if you have a wallet with 10,000 BTC, and you send 1 BTC to someone else, there will be two "outputs" - consisting of 1BTC and another with 9,999 BTC. An external observer won't (easily) be able to determine if this was a transaction of 1BTC, or 9,999 BTC.
Well and the MtGox volume last Friday was roughly half a million. USD.
On the other hand, given how much effort is going into promoting bitcoins I wouldn't be surprised if some folks are just trading them with themselves in order to pump up the numbers.
Anyway, if people have gone from "ehh, I'll stick twenty bucks in just to see what happens" to "I'm gonna buy eighty-one thousand dollars worth of these things!" then it confirms my suspicion that the world is about to run out of bigger fools to sell 'em to.
100,000 might as well be 20 bucks to a fund manager who has some high-risk fun-money set aside, or to some dot-com billionaire who thinks the crypto aspects are neat.
I really hope this is the case with the insanely large buys, and it's not someone mortgaging their home.
At the time this happened bitcoins were trading for less than 10 cents. But I have seen six-figure deals go by on the ticker at #bitcoin-market.
Thirty day volume over at mtgox is 6 million dollars. Granted, much of this might really same money going back and forth for some day trade action, but it's still way more than people who don't follow bitcoin would ever think.
While it seems to be same as throwing money into fire, it is more of a lack of atomic transaction problem. The backed up wallet has the state snapshotting the account balance. If the new transaction is invalid, it can always be rolled back to the old snapshot. That's how transactional database works. It looks like the new transaction is not atomic. Part of it has been committed (published) and part of it not saved. The system crashes and the account is in the inconsistent state, thus losing everything.
That's nonresponsive to Cowen's point, which is that as soon as Bitcoin stops being an efficient way to conduct anonymous transactions (and "probably before then"), Bitcoin's scarce fiat currency will be worth near zero.
You say, "yeah, sure, but scarcity matters". Well, actually, Cowen says it doesn't in this case.
Well, actually, Cowen says it doesn't in this case.
Can I (bit)coin the Tyler Cowen fallacy?
So Tyler Cowen says it doesn't matter? Frankly, I have a hard time trying to remember when it was last time that I heeded to TC financial (or other) advice. Maybe, when that moment will come people will just not bother asking TC what they should do.
You are refuting a straw man appeal to authority instead of the actual argument, which is so short it fits in just a couple of sentences and is included in its entirety upthread.
Your/TC argument above mistakes a handy feature of bitcoin protocols for the whole thing, and then argues than as soon as that feature will be replicated everything will be lost.
It's fun to look at the reply history and notice that someone new discovers these posts every month like clockwork and reminds this person how much he paid for pizza.
As others have mentioned, exchange rates have exploded since then and the titles are hyperbole. In May 2010, the exchange rate was about $0.005 (when 10k BTC was worth about $40). I don't have exact numbers, but I'm guessing around August 2010 the exchange rate was closer to $0.05 (or 8999 BTC worth ~$500). Given that gridecon says he bought the coins over time, I doubt he lost more than a couple hundred bucks of "investment".
Update: Based on forum posts in August 2010, seems like the rate was $0.06. I was close!
Sometimes I wonder how I missed such opportunities, or how I actually didn't figure them out. In the dawn of bitcoin, it was easy to mine 10K coins with my computers. That would make 10 years of an engineer salary in my country!
I wonder the same thing. I stumbled on Bitcoin from at tweet about a week before it hit parity with USD. I managed to earn 100+ BTC that week, and now it's worth almost $1000. If I came two weeks later, I wouldn't have been as lucky. If I came two months earlier, I would have been ten times as lucky.
Bitcoin is getting lots of traction these days. I have checked the forum and some people are looking to buy coins (small amounts $100-$150).
I wonder if this buzz is not made by someone holding a couple thousands, if not dozens of thousands of coins and looking to hyper-inflate the market and profit at selling them with these high prices.
Before I buy any bitcoin, I need some questions answered:
1. Currency Exchange: This needs to be legal and available. I don't want to look for a trader in some forum to do the exchange.
2. Stability: I probably want to have my money doubled by tomorrow, but I prefer not to lose them.
3. Popularity: If no buyer accepts bitcoins, so why are they worth it?
Unless these issues get solved, bitcoin is now just a bubble. And like always, the idiots are going to pay the consequences...
117 comments
[ 6.0 ms ] story [ 402 ms ] threadCare should be taken with any large currency transaction, digital or otherwise. Though the delicacy of data (from a consumer storage standpoint) is definitely a hit against the concept.
edit: 1 hr later, upvotes with no correction, so i think "delicacy" does cover it. why would you send $81k to a system that could lose it with a power outage, crash, accidental shutdown, etc?
My point was that 'delicate' isn't nearly strong enough of a word. If you can make diligent backups and still lose everything, the system is beyond 'delicate'.
Another post notes that newer clients have fixed this issue by pre-generating addresses. That's better, but being able to lose all your money because you lose a number just isn't acceptable.
The are many problems with such systems, which is why the entire planet moved away from them starting about 100 years ago. Just like the initial users of BitCoin who are sitting on millions of the things are limiting the bitcoin supply in order to slowly sell them for real coin.
Maybe this is the wrong mental model, but I picture it like a hard disk slowly filling up with bad sectors. Except in this case, the disk controller can't tell you how much free space the disk has.
So without the knowledge of how much cash I'm hoarding in my closet, how can you accurately value USD? I don't know how much of the answer is “you can't, exactly, but there's enough other trading that it doesn't matter” (which could apply to Bitcoin as well) versus “the value is fixed by external factors such as the US government”.
(I am not 100% how bitcoin works, but that's how I read it)
When you do a transaction you always spend ALL of the coins in that address, part of the money goes to whoever you're paying, and there rest is returned to you an a new 'change' address. (This helps your anonymity as someone watching cannot tell which amount you spent and which amount you kept.)
Before transaction:
Address1: 9000
Address2: 0
After transaction:
Address1: 0
Address2: 1
Address3: 8999
It sounds like you have wallet.dat that contains the private keys for Address1 and Address2, but not Address3.
The problem seems to be he didn't backup after he initiated the transaction. It's news to me (and apparently a lot of other people) that you can lose bitcoins in this scenario.
One proposed solution is to pre-generate some number of addresses that get stored in your wallet long before you actually use them, so they'll get backed up.
I don't see how that can be considered in any way a reasonable limitation of the system. I had no idea the "coins" were so fragile.
Every time I think, "Hey, maybe it's time to start accepting bitcoin for our software. It can't hurt, and might even help spur adoption of new digital currencies." Something crazy comes along that makes me think bitcoin isn't so well thought out, after all. I was already a little worried about the "money" being stored on our world-facing system. I'd already assumed I would need to transfer to a private account every day or something to insure that a break-in of our web server (which has never happened in our 6 years of running it, but I know enough about system security to know that a determined attacker could eventually find a way in; which is why we don't store credit card data, software signing keys, or anything else particular sensitive on that server). But, this makes me think bitcoin, as implemented, all but guarantees coin loss, given enough time using the system...possibly very large losses, as in this case.
Seems that a system crash during a transaction would also lead to the same results. And even one bit of data corruption during saving data would do the same.
But, I guess this is the pain one experiences being on the bleeding edge. It wouldn't be called that, if somebody didn't bleed now and then.
"Since this post a year ago, this feature has been implemented as keypools=100. Personally, I think it is a bad idea and does not address the problem, only pushes it off to the 101st address. At that point, users will come to expect certain backup behavior and then one day (presumably when they have more 'real' rather than 'play' money) it doesn't work as expected.
Unless the pool is recycled (change is returned to a random or cycle of addresses) then this is far more dangerous.
I propose, instead, what is expected. The change should be returned to the same address that the BitCoins were sent from.
I understand this decreases deniability/anonymity. But if someone really is paranoid, they should really be laundering money through multiple addresses in random amounts at random intervals. Sending change to a new address is just an obvious 'paper trail' considering all transactions are public, it doesn't take genius investigator to follow the money."
I understand that the current database with all the transactions takes up a few hundred MB (and this is why it can take a while to verify a bitcoin transaction).
Couldn't a small number of clients attack the currency by DOSing the network with transactions?
Edit: found this, I guess they know it "might be" a problem https://en.bitcoin.it/wiki/Weaknesses#Denial_of_Service_%28D...
Miners are free to include/exclude your transaction in their block regardless of whether a transaction fee was included. If the majority rejects transactions without a fee then it will take a longer time for the transaction to be included in a block.
That being said, I dont see how any bitcoins were lost given the example in the post. When the wallet from the flash drive was restored, he should have access to the coins again.
That the client is problematic doesn't mean the data/$ is gone.
A. Send them to someone else, which will create a new block in the chain that is confirmed by miners. B. Lose your wallet.dat completely and therefore your private keys forfeiting any way to access BTC that are yours.
A very legitimate way of backup up BTC is to just copy your wallet.dat and blockchain to a thumbdrive and store it in a safe deposit box offsite. If you ever lose your main computer for any reason, restore the backups from the thumb drive, let your blockchain catch up to current, and all your money is still intact.
If you restore your wallet.dat and let it catch up to the current blockchain, you will still have all your money, unless you really spent it.
edit: Ok, I guess I am mistaken - I see now how this happened with a really old version of the client. Newer versions of the client pre-generate several addresses in your wallet.dat, so even if you only restored an old version of wallet.dat, you would still have access to any money that is in the blockchain that you spent in the recent past. Of course, you should never let your wallet.dat fall too far behind.
If you back up your wallet prior to the transaction of 1BTC, it will not receive the private keys to the change address. Your private keys in your wallet.dat will point to an empty account. That is what this entire post about losing $81,000 is about.
The problem was, that transactions can have "change". So by sending 1 BTC, the client did in fact send 1 BTC to one address and the remaining 8999 BTC to a a newly generated address which is automatically added to the wallet. Since the backup was before this transaction, the backup didn't include the private key to the change address and 8999 BTC were lost.
A really unfortunate incident, as the person simply didn't know this technical detail. It prompted some upgrades to the client, which now generates addresses in advance and stores them in the wallet. When new addresses are needed, the ones from this pool are used first. With a recent client this accident would not have happened. It's still important to know though, that after _sending_ Bitcoins, a wallet might potentially include new private keys that are not in an old backup (if the pool has been completely used up since the last backup).
By the way: All this happened in August 2010, so this wasn't $81,000 at the time.
General rule of thumb is it takes about 60-90 minutes to be guaranteed that your bitcoins have transferred.
Bottom line - Bitcoin software is becoming more robust as Bitcoins are becoming more valuable.
Or is this just a question that's so bad it's not even wrong?
1. Seems Mt. Gox is the primary exchange of choice of Bitcoins. What mechanism or oversight in Mt Gox prevents Bitcoin exchange rate manipulation by way of wash trading? http://en.wikipedia.org/wiki/Wash_trade
2. The exchange rate data on Mt Gox shows the value of a Bitcoin shot up 10x in about 3 weeks. Hyper deflation to that extreme is not a desirable attribute for any currency.
I understand people use bitcoins for services rendered and a fairly limited number of products - but that is clearly to the benefit of the seller. It would be nice to see some statistics on bitcoin users actually using them as a currency. I'd wager it's less than 10%, esp. now they've become a media darling and a vast number of folks are looking at these like guaranteed lottery tokens.
But hey, if people really did decide to lead frugal lives in order to save money, capital would expand faster in areas more essential to human existence. Consider it as a shift toward long-term goals rather than short-term gratification. Of course, the longer the timeframe, the harder it is to predict future demand, so there is a limit to how future-oriented people can get.
http://www.appapillai.com/blog/2010/06/13/the-pound-sterling...
http://en.wikipedia.org/wiki/Troubled_Asset_Relief_Program
I don't know any bank that understands or will deal with 'bitcoins'. LOL!
Given we're talking about bitcoins this discussion is about as practical as discussing a russian tank invasion of western europe but it's interesting none the less.
And it supports microtransactions. The way I've heard it told, you could run an entire world-size economy off of one bitcoin.
So unless you have the situation in which everyone loses their wallets at once, it's just business as usual. Nothing to be concerned about.
"It is a privately created fiat currency, bundled together with an anonymity scheme for transactions. The anonymity scheme means there is some reason to grant one-time seigniorage to the original currency issuers. Eventually the anonymity scheme, in some form or another, will be available without the fiat currency. At that point, or more likely before then, the fiat currency will fall to near-zero in value."
Mr. Cowen is somehow overlooking the relative scarcity factor. Speaking strictly about bitcoins and dollars, we know how much bitcoins will be there in the next 100 years, and we have no idea how many dollars will be the in 2113?
This is what makes dollars valuable. Their scarcity is far less relevant.
Bitcoin needs to be backed by some kind of utility to have value. Scarcity alone is not enough.
That's a lot of trade outside of US which is done in USD and not in local currencies. Are you going to argue that USD is not used as medium of exchange outside of US because you cannot pay taxes with them?
Yes, but universally accepted scarce tokens are good stores of values. And currently dollars and euro, and almost every other currency are just that. Universally accepted scarce tokens.
And bitcoins are both an efficient transaction scheme and a scarce tokens. And, by design, a lot more scarce. Of course, it has not yet bootstrapped itself, but I see that as a possibility. And damn convenient possibility, too.
Bitcoins are accepted by online gamblers in much the same fashion as casino chips, by ideologues offering marginal services as political statements, and (allegedly) by criminal enterprises.
If bitcoins are mostly interesting to you as a way of reasoning through how economies would work in a world without established governments and economies, that's fine, but could you just say that? Because it's hard to pick out which of your arguments are pie-eyed what-if's.
Taxes is not what a money makes. Colonials, Greebacks, French Assignats. They all were used to pay taxes. On the other side the Somali shilling is (still) used as money. The russian (tsarist) ruble was used as money during 1918-1920 Civil war. Hell, shells were used as used as money.
As for taxes in bitcoins, it can be just of matter of time. You know the tune: Cheaper to tax than to fight
Also, define valuable, and we'll talk. I can easily foresee situations (even now) when bitcoins are valuable to me than dollars.
But you can create an arbitrarily high number of bitcoin-like electronic currency units, simply by taking the exact bitcoin implementation and changing the genesis block. Bam, Patiocoins, with guaranteed scarcity, anonymity, and a new goldrush phase -- so if you got dumped rather than pumped, you have a new opportunity to start at the top of the pyramid scheme this time.
The equilibrium is that sooner or later expected returns on starting a new distributed pump-and-dump dwarf expected returns of getting in late on bitcoins. And then poof.
There's the usual reasons we use e.g. gold for money. It's fungible, ductile, rare and essentially impossible to forge. That rules out cigarettes and Yap rai as stable currencies.
As far as gold vs. silver vs. platinum, the inflation rate (i.e. the rate at which new metal comes out of the ground) is also important. Gold inflates slower than silver, simply because there's more silver in the ground.
As far as gold vs. platinum, it's important to look at what happens if the world decided to move to gold or platinum as a currency. Gold is already used as a store of value, while platinum isn't. If everyone moved to gold as currency, it would become far more valued. If the same happened to platinum, its value would go stratospheric. There are essentially no platinum reserves (compared to gold), so its price would be even more volatile, and its price would spike far more than gold.
Platinum is also much more concentrated, geographically. There are like 5 working platinum mines in the world, and like 4 of them are in South Africa. South Africa has more of the world's platinum than the middle east has of the world's oil supply.
Those objections are not insurmountable, but human civilizations have been using gold & silver as currency for thousands of years, but the same is not true of platinum. Don't underestimate inertia.
Currency has, in fact, never been a wholly stable affair. Before paper money, allowing non-gold monies was a way of inducing inflation. But silver or platinum or cigarettes are clearly exact matches for gold while Patiocoins could automatically be that (and Patio is "famous" - that should be enough for a few people).
So this should only reinforce the point that money and currency can't just exist as a "natural" abstraction generated by an algorithm but rather must depend on social conditions (any society's currency system is transparently dependent on the overall configuration of that society's member's beliefs about the world).
Yes, you can. If your X-coins will be in some respect better than the original ones they may even supplant bitcoins. But you'll have to put of effect in order to convince people to use them. With uncertain results. Currently bitcoin has network effects working for bitcoin and against other crypthographic/p2p currencies. To change that will require a major effort.
The history of bimetallism (gold & silver) is relevant in this respect. The final result is that your are better using the medium of exchange the majority using, so usually there only one money left.
It's the same as:
How to Lose $81,000 in Bitcoins
1. Buy 9,000 BTC on one of the exchanges over time. 2. Transfer them all to someone else 3. Don't get thank you
I'm fully aware of the explanation that it doesn't matter when Bitcoins are lost because they are divisible to 0.00000001, but I'm not convinced that it makes sense to be doing business on that scale.
Official explanation here: https://en.bitcoin.it/wiki/Myths#Lost_coins_can_t_be_replace...
I see this as a psychological problem. It's hard for most people to really conceptualize value of very small numbers. Yes, people will just be converting to "real" currency to get an idea of what the value is, but shouldn't a successful currency stand on its own? Shouldn't we be able to understand the numbers we're dealing with?
That's why we do all the tricks to workaround the problem of long numbers. We're measuring the length in ~6.68458134 × 10-12 AU, time in 9,192,631,770 periods of radiation, etc. etc. all the time. We just call them 1 meter and 1 second instead. Once a "reasonable amount" of BTC reaches either 1k BTC or 1m BTC, someone will come up with a new, cute and short name for those - I don't think we'll have a problem with small / big numbers here.
According to Wikipedia, "A decibel is one tenth of a bel, a seldom-used unit."
In other words, people will adjust. Someday, we'll be dealing in millicoins or whatever. And not think twice about it.
1. Buy $81,000 in bitcoins
2. Wait.
Seriously, did this actually happen to someone? Are people actually buying tens of thousands of dollars worth of bitcoins?
Bitcoin transactions have two outputs. In order to send a fragment of money, you need to create one output for the fragment, and the remaining output to put the "change" into your own wallet.
For example, if you have a wallet with 10,000 BTC, and you send 1 BTC to someone else, there will be two "outputs" - consisting of 1BTC and another with 9,999 BTC. An external observer won't (easily) be able to determine if this was a transaction of 1BTC, or 9,999 BTC.
http://www.bitcoinmonitor.com/ -- red dots.
Well and the MtGox volume last Friday was roughly half a million. USD.
On the other hand, given how much effort is going into promoting bitcoins I wouldn't be surprised if some folks are just trading them with themselves in order to pump up the numbers.
Anyway, if people have gone from "ehh, I'll stick twenty bucks in just to see what happens" to "I'm gonna buy eighty-one thousand dollars worth of these things!" then it confirms my suspicion that the world is about to run out of bigger fools to sell 'em to.
http://forum.bitcoin.org/index.php?topic=782.0
I really hope this is the case with the insanely large buys, and it's not someone mortgaging their home.
Thirty day volume over at mtgox is 6 million dollars. Granted, much of this might really same money going back and forth for some day trade action, but it's still way more than people who don't follow bitcoin would ever think.
You say, "yeah, sure, but scarcity matters". Well, actually, Cowen says it doesn't in this case.
Can I (bit)coin the Tyler Cowen fallacy?
So Tyler Cowen says it doesn't matter? Frankly, I have a hard time trying to remember when it was last time that I heeded to TC financial (or other) advice. Maybe, when that moment will come people will just not bother asking TC what they should do.
You sound very insecure about your point of view.
Well, good luck with that.
It's fun to look at the reply history and notice that someone new discovers these posts every month like clockwork and reminds this person how much he paid for pizza.
As others have mentioned, exchange rates have exploded since then and the titles are hyperbole. In May 2010, the exchange rate was about $0.005 (when 10k BTC was worth about $40). I don't have exact numbers, but I'm guessing around August 2010 the exchange rate was closer to $0.05 (or 8999 BTC worth ~$500). Given that gridecon says he bought the coins over time, I doubt he lost more than a couple hundred bucks of "investment".
Update: Based on forum posts in August 2010, seems like the rate was $0.06. I was close!
I wonder if this buzz is not made by someone holding a couple thousands, if not dozens of thousands of coins and looking to hyper-inflate the market and profit at selling them with these high prices.
Before I buy any bitcoin, I need some questions answered:
1. Currency Exchange: This needs to be legal and available. I don't want to look for a trader in some forum to do the exchange.
2. Stability: I probably want to have my money doubled by tomorrow, but I prefer not to lose them.
3. Popularity: If no buyer accepts bitcoins, so why are they worth it?
Unless these issues get solved, bitcoin is now just a bubble. And like always, the idiots are going to pay the consequences...