Bitcoin discussion and news

[camouflage]

reminded me of the below, not sure if true but what the hell, it features a hi-res picture of a star... i like stars.

Doesn't make your money secure by thermodynamics though, attackers will simply use the human element or a key-logger or something. But in terms of quantum brute-force of bitcoin encryption you're probably ok.

 

[kabbes]

reminded me of the below, not sure if true but what the hell, it features a hi-res picture of a star... i like stars.

...

Doesn't make your money secure by thermodynamics though, attackers will simply use the human element or a key-logger or something. But in terms of quantum brute-force of bitcoin encryption you're probably ok.

It also assumes that nobody develops a short-cut to brute-force it.

 

[Idaho]

It also assumes that nobody develops a short-cut to brute-force it.

Er... Well yes. But right now that would seem like magic to computer science.

 

[kabbes]

What actually is the minimum energy required to switch from 0 to 1, by the way? That bit (oho) isn't explained.

 

[Idaho]

I tried googling "minimum energy to switch 0 to 1" but I didn't understand any of it. It's all maths and physics.

 

[kabbes]

I tried googling "minimum energy to switch 0 to 1" but I didn't understand any of it. It's all maths and physics.

That was a good plan. I did it myself and similarly ran into a bit of a wall.

What is your opinion on the minimum amount of energy to perform a...

This guy, however, seems to think that no energy is needed, theoretically. He says

In general heat is generated only in the case of "irreversible logic operations" (e.g. AND, OR). However, "reversible logic operations" can dissipate arbitrarily little heat (e.g. inverter, Toffoli, Fredkin), hence, theoretically it is not necessary to consume energy for well selected logic operations

Don't ask me whether or not that is true though!

He quotes an equation that I also find here:

https://arxiv.org/pdf/1403.1800.pdf

Finally we derive a relation that connects the minimum
energy and the switch error probability. Provided that a
procedure for zero-power switching 0to1 (1to0)
exists, if this procedure, when applied to the initial state 1 (0) produces no change in the state, i.e. the nal state is still the state 1 (0), then its minimum energy cost
is 2k(B)T ln(2).

He's saying that this equation need not apply if you aren't making irreversible change. But even if you are assuming irreversible chance, the "T" in that equation is temperature. The graphic assumes absolute zero temperature, though, so why doesn't that equation just multiply out to zero?

 

[Idaho]

The basis for all encryption is the fact that it is very easy to multiply two large prime numbers to create a product. That product is the encryption key. But it is surprisingly hard to reverse engineer the process if you only have one of the prime factors. Even with huge amounts of computing capacity, figuring out the encryption key from just one prime (the public key) and the encrypted data takes an "unreasonable amount of time". Currently with free and unlimited use of a super computer, it would take a few decades (maybe centuries) to decrypt a simple message I can encrypt in 5 seconds.

There are so many Hollywood films where you get this kind of dialogue:

"we only have 3 minutes to find the code before the bomb goes off"
"OMG he might have left the code on this USB stick... Oh noes it's encrypted! "
" I think I can break the encryption... "
(cue graphics showing some kind of geometric puzzle being unwound)
" yes! The password is 'kittens'! "

In reality, if the villain had used a freeware pgp encryption tool, it wouldn't matter if our heroes had a deadline of 3 decades... They would be screwed.

 

[kabbes]

The basis for all encryption is the fact that it is very easy to multiply two large prime numbers to create a product. That product is the encryption key. But it is surprisingly hard to reverse engineer the process if you only have one of the prime factors. Even with huge amounts of computing capacity, figuring out the encryption key from just one prime (the public key) and the encrypted data takes an "unreasonable amount of time". Currently with free and unlimited use of a super computer, it would take a few decades (maybe centuries) to decrypt a simple message I can encrypt in 5 seconds.

There are so many Hollywood films where you get this kind of dialogue:

"we only have 3 minutes to find the code before the bomb goes off"
"OMG he might have left the code on this USB stick... Oh noes it's encrypted! "
" I think I can break the encryption... "
(cue graphics showing some kind of geometric puzzle being unwound)
" yes! The password is 'kittens'! "

In reality, if the villain had used a freeware pgp encryption tool, it wouldn't matter if our heroes had a deadline of 3 decades... They would be screwed.

Yes, this is all completely true based on current technology and current knowledge. That bitcoin graphic goes further, however, and postulates arbitrarily powerful computers, saying that it would still be not just "unreasonable" but impossible to crack.

That might be correct, but I'd like to understand the claim a bit more before just accepting it.

It is also far from impossible that mathematics comes up with more short-cuts to prime number factorisation, potentially cutting the speed from the time of the current best theoretical algorithms (a quantum computer could allow log^3 (n) in theory) down to much shorter time periods. And note that current methods do not require you to simply "count" until you get to the right answer.

 

[Idaho]

Yes, we are constrained in our predictions of the possible by our current paradigm. If we could envisage a viable and superior process to the current paradigm, based on the observable limitations and constraints we would not just be predicting the next phase of development, but would in fact be inventing it.

The burden of evidence you demand is too high. To vouch against technology and processes we currently deem to be impossible.

 

[kabbes]

Burden? I just want to know the maths that actually underlies the claim of that infographic.

 

[Idaho]

Oh.. Ok. I don't know anything about that infographic... Sorry.

 

[mod]

Anyone here use blockchain.info

I signed up to them yesterday. Paid them some money to convert into bitcoins, they took money from my account but then rejected the transaction. Should I be concerned? I go the following email from them (via Coinify)...

Transaction Rejected
Unfortunately, we cannot process trade order *******. Due to security regulations the reason cannot be disclosed. Please know that each transaction goes through an automatic risk check performed by risk assessment software which compares and analyses data in order to calculate the risk score for each transaction. We have issued a refund for your order but please be aware that it may take a few more days for the refund to show in your bank account.

We apologize for any inconvenience. If you were attempting to make a trade using credit/debit card, you can try making a trade using bank transfer instead.

 

[kabbes]

Always be concerned. It's the internet.

 

[BigTom]

Burden? I just want to know the maths that actually underlies the claim of that infographic.

The thing that's been confusing me about it is that 2^256 is obviously a very, very, very big number, but put in the context of all the billions of computers (including phones and everything else like that that's really a computer underneath it all) all around the world all doing thousands of calculations each second, I would have thought that we are already having computers do 2^256 bit changing operations collectively and shouldn't we have already run out of energy? Maybe I'm still underestimating just how big a number 2^256 is or overestimating how many computers/calculations there are, or I've simply misunderstood the graphic.

 

[kabbes]

The thing that's been confusing me about it is that 2^256 is obviously a very, very, very big number, but put in the context of all the billions of computers (including phones and everything else like that that's really a computer underneath it all) all around the world all doing thousands of calculations each second, I would have thought that we are already having computers do 2^256 bit changing operations collectively and shouldn't we have already run out of energy? Maybe I'm still underestimating just how big a number 2^256 is or overestimating how many computers/calculations there are, or I've simply misunderstood the graphic.

Let's say in the world there are 1 trillion computers working at an average of 1 GHz (no idea if this is right to several orders of magnitude, but stick with it). That means 10^12 x 10^9 = 10^21 switches from 1 to 0 or back per second.

3x10^6 seconds per year, run it for 30 years (for the sake of argument), and you have 10^8 seconds so far, giving a total of 10^29 switches so far.

10^29 / 2^256 = (2^3.33)^29 / 2^256 = 2^97/2^256 = 2^-141

So on this basis, we would only have used 1/2^141 of all available energy so far, or about 10^-42, i.e. about one million-billion-billion-billion-billionth of the total.

Of your possible reasons for misunderstanding, I think you have just severely underestimated how big a number 2^256 actually is.

 

[Idaho]

Let's say in the world there are 1 trillion computers working at an average of 1 GHz (no idea if this is right to several orders of magnitude, but stick with it). That means 10^12 x 10^9 = 10^21 switches from 1 to 0 or back per second.

3x10^6 seconds per year, run it for 30 years (for the sake of argument), and you have 10^8 seconds so far, giving a total of 10^29 switches so far.

10^29 / 2^256 = (2^3.33)^29 / 2^256 = 2^97/2^256 = 2^-141

So on this basis, we would only have used 1/2^141 of all available energy so far, or about 10^-42, i.e. about one million-billion-billion-billion-billionth of the total.

Of your possible reasons for misunderstanding, I think you have just severely underestimated how big a number 2^256 actually is.

I reckon the CIA can crack it by getting a whizz kid who bucks the rules to tap furiously at a keyboard for a few minutes.

 

[camouflage]

I reckon the CIA can crack it by getting a whizz kid who bucks the rules to tap furiously at a keyboard for a few minutes.

Only if there's a timer with a red LED display counting down somewhere.

 

[joustmaster]

 

[Idaho]

Just awful on a surprisingly varied number of levels.

 

[Backatcha Bandit]

Much quicker to use Richard Pryor, now he hacks like a boss...

 

[Idaho]

Much quicker to use Richard Pryor, now he hacks like a boss...

To be fair, in those days cyber security was so new/lax that it probably was that easy.

 

[Wolveryeti]

Quantum computing will screw *all* computer security - banks, government, military, personal. It's just that bitcoin users are the only ones taking about it.

It's the developer turf wars & mistrust (see this article), and ironically, its distributed nature that make BC particularly vulnerable though. If you cant get consensus over something as simple as increasing the transaction limit, what hope is there for implementing a complete change in the encryption standards?

Post quantum cryptography exists: Post-quantum cryptography - Wikipedia

When quantum computing bursts onto the scene, if the organisations running centralised sysadmin have any sense, they will be already be running these new encryption standards, or can at least force conformity onto their network's users through system-wide updates. I dont see that happening for Bitcoin on a reasonable timescale due to the above reasons - by which time it will be too late. Loads of burned folks will never trust it again.

 

[mwgdrwg]

Bitcoin v Quantum Computing:

 

[Ranbay]

pfft everyone has hacked a Gibson.

 

[mauvais]

Never mind what quantum computing (not a given in any timeframe) will do to encryption, consider what it'll do to mining. It's basically a licence to print money and devalue everyone else's.

 

[Wolveryeti]

Never mind what quantum computing (not a given in any timeframe) will do to encryption, consider what it'll do to mining. It's basically a licence to print money and devalue everyone else's.

I still think the encryption thing is worse. There is a system limit of 21m bitcoins, of which maybe 16m have been mined already. A dilution of the money supply of that order folks could probably handle. Money which can randomly and unpreventably be hacked and spent... not so much.

 

[Idaho]

There are a few different reasons why there isn't yet a consensus on transaction limits their isn't yet the urgency and there are issues with the suggestions so far. I think a major and immediate threat to the entire network would be a different thing.

This is all speculation though as q-bit computing is still a way away.

 

[kabbes]

This is all speculation though as q-bit computing is still a way away.

This is true. It's also arguable that solving the problems with quantum computing will necessitate solving so many other nano-level problems that the systematic impact on our entire society might be so great as to render a great many social structures simply obsolete. Maybe we'll be in a Star Trek-style post-money world. Bravely seeking out new civilizations to rapaciously exploit.

 

[DotCommunist]

This is true. It's also arguable that solving the problems with quantum computing will necessitate solving so many other nano-level problems that the systematic impact on our entire society might be so great as to render a great many social structures simply obsolete. Maybe we'll be in a Star Trek-style post-money world. Bravely seeking out new civilizations to rapaciously exploit.

 

[kabbes]

"Fuck you, Bones. I think it looks cool."