4

I understand the output of the SHA-256 hashing function is a random number between 0 and 2^256. Does the randomness of this output imply that there cannot be a, so to say, honey hole in the hashing.

For instance this reddit post, http://www.reddit.com/r/BitcoinMining/comments/1rysqh/is_there_something_strange_about_block_272754_and/

They point out that someone relayed two blocks very close together, approximately 10s apart. Could they have found an occurrence where two block solving hashes sit close to each other?

I guess this kind of doesn't make sense though. Once a block is found, or a transaction is accepted, and even for each 3 seconds(timestamp) the block header is changed. So you would really could never find a block with the same hash just one number off (same block header, with two blocks one with nonce and the other nonce+1), because the first block will not be included in the second one, the second will become a dead fork in the blockchain?

7

No, the SHA2 hashing function is not vulnerable to any known attack that significantly reduces the difficulty of finding a block, but it is a highly random process.

Randomness is funny sometimes. Statistically it it works out to be even over a longer time period or larger population, but if you focus in on a small set of samples sometimes you see a cluster or a gap. This is why the estimated hash rate of the network on blockchain.info is woefully inaccurate on anything less than a multi-day average, and bitcoin is based on a 2 week correction period instead of a smaller number.

(which is why the unlucky miner can't expect to find 2 blocks quickly "because" he had a drought, and you should not expect a gap to "make up for" earlier luck, it's the population that matters.)

  • I think this answer is correct, but I'd like to add a comment regarding potential foul play with quick consecutive blocks by the same miner/pool. The vulnerability, already widely discussed, is where a block found by the miner is kept private and not broadcast immediately, giving the miner/pool a slight head start on the next block, at the potential risk of having the original block go orphaned. – ktorn Dec 3 '13 at 6:01
  • I agree, that the luck will even out in the end, but that doesn't show that you couldn't find a block every hash right? You have the same chance each block, isn't there then a chance that you could find n blocks in a row? Though the chance is probably amazingly incomprehensibly low. – KDecker Dec 3 '13 at 6:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.