Care to describe further this attack vector? A 50% + 1 attack is about in the $10+ million range, but still doesn't "override the code".

You don't need majority. And the attack isn't about stealing money, its about taking down the system and making people lose faith in it. Remember: to make money all you need is some unlikely put options and for knowledge of market instability.

It sounds like you're talking about market manipulation.

So, please enlighten us: how is Bitcoin "mathematically flawed"?

So you are sitting on two closehold, major 0days in bitcoin? I do not mean any disrespect but why would we believe you?

Unrelated Question: Why do you have a pgp cast5 encrypted message in your profile?

You are off by an order of magnitude. It would take $9M of FPGAs to have a clear probability of performing a majority attack https://bitcointalk.org/index.php?topic=67969.msg821634#msg8... This is without even counting hosting costs: power, cooling, space.

It isn't a majority attack. Infact, today, it would probably be cheaper since bitcoins are a third of what they were when I found it. All you need to do is crash the market and buy put options against it, you don't need to create nonvalid transactions.

Put options? On Bitcoins?? Who is going to be the counter party to those options?

The reason why options work is because there are clearinghouses that will guarantee the transaction. Without a regulated clearinghouse, you're subjecting yourself to counterparty risk, which as we discovered in 2008, is a significant risk especially during market crashes.

Why would you call this a "mathematical" flaw? Your original post implied there was some algorithmic issue, when you're really just proposing dumping the exchange market -- which is a completely independent mechanism of trade.

> Also the only way to get bitcoins securely is to buy and sell them for cash. And even then there are ways of figuring out who is who.

This is ridiculous. The psuedoanonymity of bitcoin allows for laundering services which enhance your plausible deniability. You can't practically figure out "who is who" if I send a delayed transaction through a popular laundry service.

I buy and sell bitcoins for cash from people on craigslist. No identifying information is given. Guess its not the "only way."

Why does this make Bitcoin "mathematically flawed" and different from other markets? If you find a way to crash any market, be it a penny stock or a type of commodity, for example by spreading terrible rumors, you can always profit with put options.

Is your point that the Bitcoin market is still very small and can be manipulated?

Who exactly will sell me put options on BTC?

You could probably find a willing "I'm putting all of my life savings in Bitcoin because it is rapidly appreciating" type to do the deal, it's free money as far as they see it.

I've looked and I could not find anyone like that.

You could try following up with this guy: http://falkvinge.net/2011/05/29/why-im-putting-all-my-saving...

(for reference, HN discussion was here: http://news.ycombinator.com/item?id=2596475)

That guy is just buying bitcoins, he's not interested in writing put options.

Right. I'm taking the piss out of bitcoin as a whole here, I don't actually think the bitcoin market is mature enough to make shorting on a large scale feasible nor any individual put transaction advisable given the lack of clearinghouses and counterparty risk mentioned by steve8918.

But... the guy has made it clear he believed bitcoin's value would rise, so selling a put option at - say - current value would be advantageous to him. Offer a low enough strike price, high enough premium, and word your offer in layman's terms ("you get money now and might get money later if bitcoin rises or might have to pay money if it falls") and he could be tempted to enter in an appropriate contract.

Producers and merchants trying to hedge their cash flow.

It would actually be very useful to have a liquid options market to help producers and merchants overcome the current volatility.

Could you name one? I've never heard of anyone selling a put option on BTC.

The cost of such attacks will go down as bitcoin ages.

As I understand it the bitcoin rewards of mining go down over time. If the rewards go down over time, eventually it won't be profitable to mine bitcoin. So the speed of transactions goes down and the network becomes weaker because theres less cooperating computer time on it.

Maybe I just don't get bitcoin. But if true that seems like a fundamental problem.

Block rewards are not the only way to earn from mining bitcoins. You also earn transaction fees of processed transactions. As block rewards dissipate, transaction fees will naturally increase to compensate.

This is the answer I was looking for. Thank you.

Fees won't naturally increase. If I can mine profitably at the minimum amount of transactions enforced by the protocol I'll raise the difficulty as I bring on more miners and squeeze out people relying on transactions to profit. Eventually someone will come along who does mining as a loss leader (a bank?) and will offer transaction fee-less transactions to customers which will squeeze out the last of the profitable miners. Competition will only push the mining profit lower and lower.

Alternatively, a really busy Bitcoin network will be limited by the maximum block size and people will pay for priority inclusion in a block. However, this limitation is totally artificial and doesn't scale with the increasing amounts of bandwidth and disk space available. Miners have an incentive to "suck up" to Bitcoin users - without them their mining has no point. If the Bitcoin users demand bigger block sizes so they don't have to pay transaction fees I think miners will have to take the partial hit from bigger blocks rather than the total hit of lack of blockchain use.

"If the rewards go down over time, eventually it won't be profitable to mine bitcoin."

Yes.

The idea is the following: If the amount that can be mined per unit of computation is decreased, then fewer will be created. And if there are less of them being created and more of them demanded, then their value increases. So this balances out. Also, computational power gets cheaper with the passage of time.

The largest caveat I see there is the part about more of them being demanded. (or equivalently, the market equilibrium of demands for bitcoin holdings shifting towards greater demand)

I detect confusion here, and I'd recommend you take some time on the bitcoin wiki to clarify:

> "If the rewards go down over time, eventually it won't be profitable to mine bitcoin." > Yes.

No. Because the other part of mining is that you get transaction fees for it. Mining isn't a luxury, it's necessary for the network to function (otherwise double-spending isn't prevented). So as time goes on and less coins can be made by mining, transaction fees will increase to compensate and keep mining profitable.

> Also, computational power gets cheaper with the passage of time.

But, mining does not get cheaper because the network (by design) adjusts the mining difficulty to keep the incoming block rate constant. So in relative terms, mining will never become significantly more or less expensive with time.

> The largest caveat I see there is the part about more of them being demanded.

Not a problem. This is solved because bitcoins are infinitely divisible (up to 8 decimal places currently and this can be extended if needed). So if there's more demand, the value of coins increases but you can still trade them in whatever meaningful quantities you like. This isn't something you can do with cents, but you can trade one millionth of a BTC just fine.

My points were purely economic. The OP was just asking a simple economic question, and I gave a simple answer. It doesn't need to be explained in the mechanisms employed by the design of bitcoin.

"Not a problem. This is solved because bitcoins are infinitely divisible"

I detect confusion here.

I'm not talking about lack of divisibility, but a lack of demand. Increased divisibility doesn't facilitate demand. It's just a nice feature of something that could be money or commodity.

Water is infinitely divisible but that doesn't get it demand. If people don't exercise demand for bitcoin, (as in, they don't want it) divisibility doesn't solve the problem.

I haven't seen any evidence that transaction fees will increase. It looks more likely that both block rewards and transaction fees will approach zero, mining will become unprofitable for everyone, and then difficulty will drop.