how does the saying go?
"when the wise man points at the moon, the simple man stares at the finger."
Topic: Wall Observer BTC/USD - Bitcoin price movement tracking & discussion - page 27246. (Read 26691955 times)
June 10, 2014, 10:22:44 PM
how does the saying go?
"when the wise man points at the moon, the simple man stares at the finger."
June 10, 2014, 10:22:33 PM
BTC-E is higher than stamp!
Meanwhile, China doesn't give a damn pushing on $660.
June 10, 2014, 10:21:15 PM
June 10, 2014, 10:21:09 PM
More like panic...
June 10, 2014, 10:18:47 PM
June 10, 2014, 10:18:23 PM
1W MACD bout to cross
lower limit of the channel being challenged!
I have buy orders that should be getting filled as we speak! (*edit - nm, virtex still over 700 and just recently had a spike to 745, damn Canadians just can't keep it in their pants. *)
June 10, 2014, 10:17:52 PM
June 10, 2014, 10:15:05 PM
1W MACD bout to cross
June 10, 2014, 10:14:20 PM
the depth was so low.. think they had to make a nice red candle to scare some weaker hands
June 10, 2014, 10:13:47 PM
Da fuq is up with stamp lately? Its crazy bearish while the other exchanges are fairly neutral.. I cant believe Huobi hasnt tanked yet.
June 10, 2014, 10:11:47 PM
Whoever keeps scooping up the coins from this Beartard on Stamp is going to reap the rewards.
June 10, 2014, 10:05:25 PM
(You are not referring to the P != NP conjecture I suppose? It is just as uncertain, but it has absolutely no relevance to those two assumptions in the bitcoin protocol.)
I am. And it has. At least to 'what the protocol can handle in principle being thrown at it', not 'what the protocol is and does right now'. As long as there are problems that are hard to solve but easy to check, the protocol can be adapted to the threat that a particular problem turned out to be easy to solve after all.First, let me say that I was just reacting to that "pure incorruptible mathematics" bullshit by the reporter. I am not arguing that the lack of mathematical proof iis a flaw of bitcoin. My skepticism does not come from that "flaw" (or from any other strictly technical issue), but from political/social/practical problems, for which my probabilities are obviously different from those of most people in this thread.
As for the P != NP conjecture, I do not need to tell you that it makes sense only if the problem instance size n is allowed to go to infinity, since the difference between polynomial and superpolynomial appears only in the limit, when n is "just about to get there". The two classes are indistinguishable if one assums that n is bounded, say n < 10^500. But the bitcoin mining problem and the signature forging problem have a fixed n, threfore only a finite number of instances; in that case, complexity theory says that both problems are trivial and can be solved in constant time (e.g. by a simple table lookup). Obviously this theoretical conclusion is irrelevant to the practical question.
The sad fact is that theoretical computer science does not have any tools yet that can provide useful lower bounds for problems of fixed size. Even assuming P != NP, there is no way to prove that the current protocol, with specific key and hash sizes, is safe. It may even be the case that forging a bitcoin transaction signature is easier than checking it, or that finding the right nonce for a block is easier than computing SHA256 twice -- independently of whether P = NP or not. We simply do not know.
If someday one were to find fast solutions for those puzzles, then complexity theory indeed says that there must exist safe replacements for them -- assuming P != NP, of course. But the theory cannot identify those replacements, not even tell how many bits one needs to use in keys and hashes to ensure that inverting those functions is hard enough. Conversely, even if P = NP, the current protocol may be safe, or there may exist safe alternatives for those two puzzles. (As you surely know, "polynomial" does not mean "computable in practice", and "superpolynomial" does not mean "impossible to compute in practice".)
The difficulty of those two puzzles, for the currently chosen key and hash sizes, has been verified only experimentally, by tests that have ruled out some weaknesses that people have thought of checking. (For a trivial example, someone surely must have checked that the leading bits of the block hash are not linear functions of the nonce bits, not even approximately.) However, a thorough check would require testing all possible algorithms (or boolean circuits) up to some large size, for all possible inputs -- which is totally outside the realm of practicality.
With no known shortcuts, exhaustive enumeration is pretty much the best way we now have to solve those puzzles. That is all one can say about the issue.
Not to forget, there's some speculation in the more tinfoil-hatty segments that the NSA may have supplied algorithms that are "pre-weakened". Seems unlikely but with everything that has been coming out lately, perhaps not to be dismissed out-of-hand so quickly.
June 10, 2014, 10:01:03 PM
June 10, 2014, 09:50:07 PM
(You are not referring to the P != NP conjecture I suppose? It is just as uncertain, but it has absolutely no relevance to those two assumptions in the bitcoin protocol.)
I am. And it has. At least to 'what the protocol can handle in principle being thrown at it', not 'what the protocol is and does right now'. As long as there are problems that are hard to solve but easy to check, the protocol can be adapted to the threat that a particular problem turned out to be easy to solve after all.First, let me say that I was just reacting to that "pure incorruptible mathematics" bullshit by the reporter. I am not arguing that the lack of mathematical proof iis a flaw of bitcoin. My skepticism does not come from that "flaw" (or from any other strictly technical issue), but from political/social/practical problems, for which my probabilities are obviously different from those of most people in this thread.
As for the P != NP conjecture, I do not need to tell you that it makes sense only if the problem instance size n is allowed to go to infinity, since the difference between polynomial and superpolynomial appears only in the limit, when n is "just about to get there". The two classes are indistinguishable if one assums that n is bounded, say n < 10^500. But the bitcoin mining problem and the signature forging problem have a fixed n, threfore only a finite number of instances; in that case, complexity theory says that both problems are trivial and can be solved in constant time (e.g. by a simple table lookup). Obviously this theoretical conclusion is irrelevant to the practical question.
The sad fact is that theoretical computer science does not have any tools yet that can provide useful lower bounds for problems of fixed size. Even assuming P != NP, there is no way to prove that the current protocol, with specific key and hash sizes, is safe. It may even be the case that forging a bitcoin transaction signature is easier than checking it, or that finding the right nonce for a block is easier than computing SHA256 twice -- independently of whether P = NP or not. We simply do not know.
If someday one were to find fast solutions for those puzzles, then complexity theory indeed says that there must exist safe replacements for them -- assuming P != NP, of course. But the theory cannot identify those replacements, not even tell how many bits one needs to use in keys and hashes to ensure that inverting those functions is hard enough. Conversely, even if P = NP, the current protocol may be safe, or there may exist safe alternatives for those two puzzles. (As you surely know, "polynomial" does not mean "computable in practice", and "superpolynomial" does not mean "impossible to compute in practice".)
The difficulty of those two puzzles, for the currently chosen key and hash sizes, has been verified only experimentally, by tests that have ruled out some weaknesses that people have thought of checking. (For a trivial example, someone surely must have checked that the leading bits of the block hash are not linear functions of the nonce bits, not even approximately.) However, a thorough check would require testing all possible algorithms (or boolean circuits) up to some large size, for all possible inputs -- which is totally outside the realm of practicality.
With no known shortcuts, exhaustive enumeration is pretty much the best way we now have to solve those puzzles. That is all one can say about the issue.
much better
some guy once said "if you wanna be a good writer, write what you know."
*I'm paraphrasing of course*
June 10, 2014, 09:43:05 PM
(You are not referring to the P != NP conjecture I suppose? It is just as uncertain, but it has absolutely no relevance to those two assumptions in the bitcoin protocol.)
I am. And it has. At least to 'what the protocol can handle in principle being thrown at it', not 'what the protocol is and does right now'. As long as there are problems that are hard to solve but easy to check, the protocol can be adapted to the threat that a particular problem turned out to be easy to solve after all.First, let me say that I was just reacting to that "pure incorruptible mathematics" bullshit by the reporter. I am not arguing that the lack of mathematical proof iis a flaw of bitcoin. My skepticism does not come from that "flaw" (or from any other strictly technical issue), but from political/social/practical problems, for which my probabilities are obviously different from those of most people in this thread.
As for the P != NP conjecture, I do not need to tell you that it makes sense only if the problem instance size n is allowed to go to infinity, since the difference between polynomial and superpolynomial appears only in the limit, when n is "just about to get there". The two classes are indistinguishable if one assums that n is bounded, say n < 10^500. But the bitcoin mining problem and the signature forging problem have a fixed n, threfore only a finite number of instances; in that case, complexity theory says that both problems are trivial and can be solved in constant time (e.g. by a simple table lookup). Obviously this theoretical conclusion is irrelevant to the practical question.
The sad fact is that theoretical computer science does not have any tools yet that can provide useful lower bounds for problems of fixed size. Even assuming P != NP, there is no way to prove that the current protocol, with specific key and hash sizes, is safe. It may even be the case that forging a bitcoin transaction signature is easier than checking it, or that finding the right nonce for a block is easier than computing SHA256 twice -- independently of whether P = NP or not. We simply do not know.
If someday one were to find fast solutions for those puzzles, then complexity theory indeed says that there must exist safe replacements for them -- assuming P != NP, of course. But the theory cannot identify those replacements, not even tell how many bits one needs to use in keys and hashes to ensure that inverting those functions is hard enough. Conversely, even if P = NP, the current protocol may be safe, or there may exist safe alternatives for those two puzzles. (As you surely know, "polynomial" does not mean "computable in practice", and "superpolynomial" does not mean "impossible to compute in practice".)
The difficulty of those two puzzles, for the currently chosen key and hash sizes, has been verified only experimentally, by tests that have ruled out some weaknesses that people have thought of checking. (For a trivial example, someone surely must have checked that the leading bits of the block hash are not linear functions of the nonce bits, not even approximately.) However, a thorough check would require testing all possible algorithms (or boolean circuits) up to some large size, for all possible inputs -- which is totally outside the realm of practicality.
With no known shortcuts, exhaustive enumeration is pretty much the best way we now have to solve those puzzles. That is all one can say about the issue.
June 10, 2014, 09:38:15 PM
did you try and it failed?
no Fundz - 0.00150771 already gone
is'nt that like 15 thousand bits on facebook?
June 10, 2014, 09:36:00 PM
did you try and it failed?
no Fundz - 0.00150771 already gone
June 10, 2014, 09:35:12 PM
My most conservative projection:
We're due for a margin cleaner IMO
We're due for a margin cleaner IMO
June 10, 2014, 09:23:39 PM
sweet! it worked!! thanks!!!
(not really, I didn't claim it
)but it just got me thinking that anyone who says "thanks" would be linking their
I also advocate that random people just say thanks to obscure the bit-trail.
Umm I got it... Thanks Adam!
June 10, 2014, 09:22:37 PM
wow, adam's offering free beers and you could drop a pin in here
(guessing everyone is now trying to figure out how to import private keys )
so I'll take this time to post a chart
those are *almost* completely arbitrary lines I drew a week ago, I think we could go sideways longer than most,
I'm also not certain that this will break to the upside, but that's probably just me hoping for another chance at sub-500 coins...
*(edit cuz I missed rudy's post)*
did you try and it failed?
(guessing everyone is now trying to figure out how to import private keys
)so I'll take this time to post a chart
those are *almost* completely arbitrary lines I drew a week ago, I think we could go sideways longer than most,
I'm also not certain that this will break to the upside, but that's probably just me hoping for another chance at sub-500 coins...
*(edit cuz I missed rudy's post)*
damn, 1 second late
pre-paid drinks are super cool idea now powered by bitcoin
pre-paid drinks are super cool idea now powered by bitcoin
did you try and it failed?
Jump to: