2 ms·

Note that brute-forcing an exponential-time algorithm is not generally possible. If you have an algorithm with a runtime of 10^N, then after N=61 the runtime o

Note that brute-forcing an exponential-time algorithm is not generally possible. If you have an algorithm with a runtime of 10^N, then after N=61 the runtime of the algorithm will be greater than the age of the universe expressed in the Planck Time. For a 2^N algorithm it'd take about N=202. If you're willing to talk in nanoseconds instead of Planck Times (chosen because our current understanding of physics breaks down at that scale), it's more like N=26 for the 10^N case. Practically speaking, though, users of privacy coins are usually looking for plausible deniability rather than total untraceability. The legal standard for criminal cases is "beyond a reasonable doubt"; if there's even a 50% chance that you're just an innocent person caught up by circumstances, you'll have trouble convincing a jury of guilt.

not_a_cop75 7y ago
I like your thinking, but all of this is assuming that runtime for something stays precisely the same no matter what. Most operations are at least reasonably parallelizable, and I would argue that cryptocurrency depends largely parallel operations, which in general should mean that breaking it is also a parallel operation. Plus with advances in quantum computing, regular computing and storage, it's hard to be sure how long any of this information will be stored for or long long someone would crack it for, or wait for it to be crackable. Also, you imply that some amazing computational shortcuts aren't ever going to be arrived at. This historically, is hard to be sure of. Not saying everything will be broken, but more of it is all the time.