4 ms·
And yet the market dynamics of that wouldn't be all that bad. They'd never get 100% of the blocks, so all they'd be doing is slowing down the flow of transacti
by EdHominem 10y ago
And yet the market dynamics of that wouldn't be all that bad.
They'd never get 100% of the blocks, so all they'd be doing is slowing down the flow of transactions and increasing the fees (because people would be paying for fewer spots in the non-DoS blocks, which would incentivize other miners.
Bitcoin scales fees, instead of scaling transactions/second, so for a while it wouldn't be economical to use it for small purchases where the fees would be almost as much as the price, but for larger purposes, and as a store of value, it'd work just as well.
- runeks 10y agoI agree that it wouldn't be all that bad, but I disagree that they can't cause a complete standstill if they want to. If they have more than 50% of mining power, they simply refuse to build on top of the honest blocks (from the rest of the network). The clients will follow the longest chain. However, it's a costly attack to maintain in electricity alone -- with a dwindling income, since your income is denominated in the currency you're attacking -- and it wouldn't take that long for bitcoiners to get sufficiently pissed and change the hashing algorithm, thereby disarming this attacker completely.
- bduerst 10y agoIf the goal was to cause a complete standstill, then they wouldn't need to take over the mining, all they would need to do is flood the network with transactions with minimal fees. Not hard to do for someone who manipulates currencies and has large cash reserves. The miners would keep mining but Bitcoin users wouldn't be able to get a transaction through unless they paid a large transaction fee.
- jcoffland 10y agoToday's mining software already ignores transactions like this. Miners select the highest paying transactions first. When the block is full they ignore the rest until the next round. This makes flooding the market expensive.
- bduerst 10y ago>Not hard to do for someone who manipulates currencies and has large cash reserves.
- jcoffland 10y agoSure you could push the transaction cost up for a little while if you're into throwing money away.
- bduerst 10y agoIt's not throwing away if you were going to flood that currency anyways, and just used BTC instead of gold as the intermediary. I did the math above, and if you wanted to save money on attacking the network with your own mining, you could spend up to $71 per transaction on every transaction.
- proto-n 10y agoNot necessarily "large", only larger than the "minimal fees" you mentioned earlier.
- bduerst 10y agoUsing just over 4 million antminer S9's at 3.2 cents/kWh (bulk cost of coal power) to generate the 1,350,000 TH/s required to get 50% control of the network would cost you $3.8 billion over a 3 month time period. At 7 transactions a second, it would take 54 million transactions to flood the network over a 3 month time period, meaning you could spend $71 on each transaction fee. If you wanted to stop the bitcoin network and wanted to save money on the attack of mining, you could force everyone to make a "minimal" fee of up to $71 to make one Bitcoin transaction. As the costs of mining increase, this is only going to increase as well.
- EdHominem 10y agoBitcoin is also traded off-network so the market would remain liquid and you'd be paying miners an extra $300,000 per block so they could easily afford the transaction fees to settle large amounts. You'd just make Bitcoin a reserve currency.
- bduerst 10y agoSidechains still require the bitcoin network to validate eventually, otherwise they are just another payment system. They also don't scale with users. Ethereum & Monero have become popular enough that most people would just switch to that rather than go to lightning. Then the whack-a-mole begins.
- EdHominem 10y agoYeah, and the cost to do that in this scenario is $72. Comparatively expensive, but you're flooding the network with money... Anchoring a sidechain just increases the resources needed to tamper with it, as if it itself had a higher proof of work. It doesn't need to happen every ten blocks or any set number. As often as possible is good enough. Sidechains can even have sidechains. If the $72 was too expensive that chain can support hundreds of other chains anchoring into it, spreading the costs. But most people would still just use exchanges and as long as the exchanges settle often enough that their government-mandated floats cover the difference, what do they care if the underlying network itself is slow?