Bitcoin Forum

I think the bitcoin mining power needs to keep growing rapidly for the years to come.

Here is why I think so:
In theory, to successfully "take over" the bitcoin network, one would need to have more processing power than all the other miners combined.
Right now, how much would it cost to achieve this ?
For a total network hashing power of 600Ghash/s, with a GPU able to do about 300Mhash/s (Radeon 6950) costing about $250, it would cost 600,000 / 300 * 250 = $500,000.
This amount is for the GPUs alone. Obviouly one needs to add other components (motherboards, power supplies, ect), then one needs to pay some heavy electricity bills and probably hire a team to build that big GPU cluster.
Ok, let's say, this would double or triple the price.
In the end, we are talking about 1 or 2 millions dollars here.

For an individual, this is rather difficult to achieve.
But, if any governement or big corporation on the planet felt threatened by the bitcoin economy and wanted to shut it down, it could do so relatively easily.
A few million dollars is not much money for them.

Remember that governement have shown willingness to shut down alternative currencies in the past (liberty dollar, eGold, eBullion). BitCoin is different from those currencies, but at some point someone may consider it as a threat.
To make those big players think twice about shelling out money to sink BitCoin, the cost would need to be at least $100 million: a hundred times higher than it is today.
In the long run if BitCoin goes global, the costs would need to be even higher by several orders of magnitude.

So, what I suggest is:
1- keep increasing the mining power fast. Find new way to make people and business participate in mining.
2- be careful not to piss off anyone too much. especially not governements and some 3 letters agencies :)

HTH !
Note: sorry if my english is not so good, I lack pratice :)

Please understand that having more computing power than the entire honest network doesn't allow you to "destroy bitcoin", it only allows rewriting recent (couple weeks, few months at most*) transaction history. Only miners who have their blocks replaced would "lose" money - and consequently everybody that bought their coins - and everybody else would have to wait for their recent transactions to be confirmed again, in case the attacker erases everything.

This could also allow the government in question to double-spend, but if it has no profit motives, it would probably not take the risk of showing itself like that. It's not on their interest, taxes are easier. ;)

Yes, it could cause some awful problems, but in the end, being the target of attacks from a government may have positive effects for bitcoins, meaning, recognition and support from those who don't approve such action. And it could cause bad effects for the images of that government as well.

I don't think it's a good strategy, neither for profit-oriented criminals, nor for political oriented ones.


* I don't know if developers keep moving the "milestone block" at each release as previously planned. Nothing older than this marked block can be replaced. I wonder if there should be a sort of "semi-official" marking of blocks like that, done here in bitcoin.org, in order to avoid long rewrites of the network. This policy could be followed by other clients, eventually.

I agree the number of miners need to grow, but the good news is, with today's rally, it is going to grow fast :)

Also, as caveden said, a takeover wouldn't be that bad, it would probably cause the bitcoin network to halt for a while, while a solution is found, but almost no coins would be lost. We would just need to restart just before the takeover.

It is very possible that I overestimated the problem. :)

However I was talking about some entity willing to shut down BitBoin, not making a profit from it.
In that case, the double spending problem would be used to undermine the trust in the system, not getting rich.
If an attacker is able to create long alternative branches, it would be able to revert many transactions and double spend money. If the attacker is able to do it for an extended period of time (say, a few months) then the trust in the system will be pretty much distroyed forever.

> I agree the number of miners need to grow, but the good news is, with today's rally, it is going to grow fast

It's great the number of miner is growing fast :)
However, I think you need to encourage new miners to mine.
Another thread on this forum is talking about removing the "Generate bitcoin" feature from the base client. Based on the reasonning above, I think the network needs more miners, not less.
Perhaps that instead of mining alone, the base client could join some kind of default/official pool so as to encourage participation and collect a few bitcoin ?

Total hashing power is roughly proportional to the exchange rate.

At present no government or corporation is going to bother attacking Bitcoin because it's far too small and obscure.

When the Bitcoin user base grows, so will its value, and so will the hashing power.

By the time Bitcoin becomes so large that governments start feeling threatened by it, the exchange rate will have climbed to over $100 USD/BTC hopefully.

Feeling threatened is one thing, but until a large bureaucratic institution actively starts fighting Bitcoin, several more years may pass.  

Governments and large corporations have a track record of being slow to react to paradigm-shifting technologies.

Exactly. In addition to that if mining grows faster than exchange rate you're going to need to have people running at a loss.
I don't see that happening. Of course a few people would but then others would drop out bringing it back to equilibrium.
As technology improves and it becomes more popular you're going to have more mining power. Naturally tech grows on the enemy's side too.
Basically what I'm saying is I don't think you're going to artificially increase mining long term. Of course if there was a short term attack I could see people running at a loss for a few days or whatever but not long term.

I'm not convinced the hashing power will grow with the user base by that much.  You'll start seeing more consolidation as the casual miners become no longer profitable, and only those with dedicated rigs in areas with cheap electricity will be able to compete.

Whose image would be more tainted, the victim's or the attacker's? Imagine the headlines: "US government spends millions to hack and steal digital money from US citizens and foreigners".

This would need completely disguised from the public. A government openly doing such attack could even cause some international friction. And I'm not sure it's that easy to hide such an expensive attack from the public, they can't manage to hide even their secret documents!

How about "FBI Cyber Crimes division prevents funneling of money to terrorists, drug dealers and child pornographers through new technology."

That would be a more likely headline and people would say "good".

Seriously they could spin the hell out of it, they already made the LD look like they're evil and they didn't give the gov. nearly the material to work with that we are.

> Whose image would be more tainted, the victim's or the attacker's? Imagine the headlines: "US government spends millions to hack and steal digital money from US citizens and foreigners".

All it takes is a few key people in the governement starting to think BitCoin is somehow "bad for the US dollar" and then they will find some official or non-official agency to take care of it.
After all, governement secret agencies of all the countries around the world are known to do much worse than hacking a network...
And there will be no bad press. When caught, they'll just say: "we terminated a network of internet hackers designed to undermine the US currency", or something along that line.

To avoid this scenario, I can think of 2 things to do:
1- Continue to make clear that BitCoin is good for everyone (including governements), not a way to avoid taxation or screw governements.
2- More miners, so that BitCoin is more resistant to attacks.

Now that's a catchy headline ! :)

I honestly don't think you're going to convince governments that btc is good for them. I'm not sure it is. Please tell me how if I'm missing something.
I see the benefits (if any besides off the books dirtyness) being very few and the cons being huge for the gov.

Attacking bitcoin only makes it stronger.

BitCoin motto:

"Building Character through Adversity"

Depends upon which side of government that you are talking about, the public side that is honestly serving the public, or the shadow side that desires to continue to control the monetary system.

What about those governments that choose to help Bitcoin because of the potential threat that it poses to the US $ hegemony?  China doesn't want to see their $3 Trillion in reserves suddenly drop in value, but if they did they could destroy the US$ value in days by simply dumping all of their treasuries upon the market.

Iceland, however, would be vastly favored by directly supporting Bitcoin, openly or otherwise.  Bitcoin would be far better for their own small economy than the Euro.

The national debt grew by 3 trillion in 2 years.  Then dumping 25% of the bonds might cause rates to go up.  The fed will just buy some of them.

Hashing power of the network will always be related to bitcoins exchange rate. As the rate grows so will the hashing power.

As a direct consequence of this, the cost to aquire roughly 1/2 of the network that is, to have 1/2 of the networks hasing power, and therefor be able to do damage to the network will be at a fairly fixed ratio of the overall bitcoin economy.

So right now the bitcoin economy is worth about 6Million USD, it would cost about 2Million USD (you say) to have 1/2 the networks hashing power.

If my theory(it's just an idea, but it makes sense to me) is correct, then when the bitcoin economy is worth 100Million USD it will cost about $30Million USD to do this attack.

Economy | Attack price
1Billion   |300Million
6 Billion  |1Billion
10 Billion |3 Billion

Essentially the more the bitcoin economy grows, then the more expensive it will become to attack the network, because the networks power will have grown along with the economy. It will get to the point where there would be much cheaper methods of attacking the network than brute forcing it.

Also some variables you are missing in your cost calculations, rent (machines have to be housed somewhere), cooling, operators (government workers are not free). And that's assuming the government has somewhere they can just stick these.

If a government or other large organization wanted to shut down Bitcoin, there'd be much more effective ways than trying to take over the hashing power of the network. Shutting down the exchange sites and getting large ISPs to block Bitcoin traffic would do a lot of damage for a lot less money.

There are ways - granted cumbersome - around such ISP blocks (think torrents). I think the worst they can do is threaten with jail time anyone who decides to operate an exchange site. If they block the gateways between the "real" economy and the bitcoin economy it would make things a bit more difficult for us.

Wouldn't it be possible to create some kind of decentralized exchange?

Yep theoretically it can be done. I would imagine it would be in the form of an application or "addon' to the official bitcoin software. It would essentially work like the existing OTC bitcoin marketplaces already in place BUT it would not vulnerable to shutdown because it would not be hosted on web servers.

I've got some ideas as to how this should work. One would be that the addon would operate a parallel P2P network which stores and broadcasts all available bid/ask order (like a stock market exchange but decentralized). After you decide how much you wish to buy your oder would be matched with the best available price for the number of "lots" (how many bitcoins you wish to buy) you want. Then if your order was met by a single person (the purchase CAN be spread across multiple sellers even at different prices if your total order cannot be met by a single seller at your asking price) he/she would then receive in encrypted format payment details and turn you'd receive his/her's payment details. This software could run it's own peer to peer escrow service whereby whatever bitcoins you are due to receive would be spread out across the network and in a sense every participant in the marketplace would own a tiny tiny fraction of your transaction UNTIL the settlement terms have been met - ie you paid the seller. Now where it gets tricky is how do you ensure the safety of the sellers and still allow buyers to use a wide variety of payment methods. Also how do you keep the seller's privacy?

I can think of some ways but they'd be a tad bit cumbersome and open to other risks. There are also tiny technicalities that need to be addressed, such as what happens of the seller decides not to release the bitcoins from escrow. I think that depending on the method of payment used by the buyer the network should automatically release the funds, but since this would require that a certain database of various transaction clearance time frames be stored by the network I think a better idea would be to have it simply expire at 30 days regardless of the method of payment.

Also we would need a distributed user rating/feedback system. This would also have to be stored by the network, but since this would be simple text which can be compressed easily it should not pose too much of a bandwidth burden.

This addon could also use the bitcoin network to bootstrap itself. When it first starts up it would send a request to the various bitcoin nodes it normally connects to anyways asking whether that particular node is running this software and if so whether it is a willing participant in the network - ie whether it accepts incoming connections.

Just throwing out some ideas.