Quote from: amincd on July 05, 2011, 06:27:12 AM

Quote

With that premise, that could very well be.

Thanks for conceding that.

You're welcome |-)

Quote from: amincd on July 05, 2011, 06:27:12 AM

Quote

The whole double-spending scenario is proportional to bitcoin value. If value were to go up orders of magnitude, the number of blocks needed to get your investment back goes down the same order of magnitude.

But difficulty, and therefore cost of a double spend attack, is also proportional to bitcoin value, so the rise in the potential reward of a double spend attack, is canceled out by the rise in cost in pulling it off, as bitcoin value increases.

I don't agree, as the cost of pulling it of is proportional to transaction fee cost, which is much lower than transaction value.

Let's look at the current state of bitcoin:

Rewards are 50 (temporally minted) + 0.10 (fees) bitcoins per block.

Resulting hashrate is 11326 (payed by temporally minted) + 23.6 (payed by fees) Ghash / second.

If nothing were to change in the value of bitcoins or transaction fees, I'd eventually have to produce 23.6 Ghash / second for a succesful attack. Which would cost me roughly 15 kW electricity (ca. 3$ per hour) and (very) roughly 1$ depreciation per hour of my hardware. Let's say it costs 5$ per hour all together.

I can then sustain an attack where I forge (double spend) say 100$ (which will not be conspicuous) for 20 hours (which should be more than enough to collect).