Random Number Generators

[Smilin]

Originally posted by: Peter
The silliest cryptography attempt was an early revision of my local bank's internet software ... the Java applet would insist you wave the mouse around in its window for a few seconds before continuing

It may sound silly but depending on how they implemented it that might not be an entirely stupid idea.

 

[Mark R]

Originally posted by: Shalmanese
A related question: How much would it cost to build a real random number generator into a CPU or chipset and do any CPU's have one (ie. Sparcs, Alphas etc). Are there any plans to put one into a consumer class CPU?

The circuit itself is small, the problem is ensuring that it works reliably in the harsh environment of a PC system.

Intel have integrated a real random number generator into their chipsets starting with the i810.

 

[heliomphalodon]

Originally posted by: dejitaru
Particles directly affected by those which came before them, hence predictable.
Even the best RNG's are notably flawed.

Entropy is not mathematically or physically possible.

What on Earth are you talking about? Fundamental particles like electrons have no memory, they have quantum states.
These states evolve over time, but their random transitions are not predictable.
Exactly how are "even the best RNG's" "notably flawed"?
When did entropy become mathematically and physically impossible?

 

[Smilin]

What on Earth are you talking about?

What he said.

 

[dejitaru]

Fundamental particles like electrons have no memory, they have quantum states.

I'm aware.

These states evolve over time, but their random transitions are not predictable.

Prediction requires a bit of research on the particle, is not practical with current technology. The motion of particles are still the effects of some cause.

Exactly how are "even the best RNG's" "notably flawed"?

See above.

When did entropy become mathematically and physically impossible?

Since its inception.

 

[heliomphalodon]

Your most recent comments illuminate nothing. If you contend that entropy is not possible, explain why you believe this to be the case. If you contend that it is possible to predict a random quantum transition, explain how this is done. If you contend that all random-number generators are flawed, describe these alleged flaws.

I challenge you to display a physical theory that is coextensive with QED but which allows the prediction of, say, the exact moment when an isolated atom of Carbon-14 will decay.

 

[Fallen Kell]

The other thing you need to worry about is if the random number generator you use actually will generated randome numbers over the range that you are looking for. Many methods out there will not do this. Especially functions like rand(low, high) in C++ (from MS at least). You will need to really generatate a random floating point number between 0 to 1 and then scale that up to the range you actually want to use.

Remember doing a project for college in which we needed to "roll dice" and many people could never "roll" a "6" because of the algorithmn used by MS was incorrect and not complete across the range.

 

[drag]

I beleive the term used for discusing the random movement of particles is "brownian motion". I learned thsi from playing the text-based adventure game based on the book "hitchhiker's guide to the galaxy",

HINT: Use hot tea to get good brownian motion...

Hey were did that whale come from?

Hey, Ford. Your turning into a peguin. Stop that, please.

Oh look at that. What is this around me? What ever it is there is a awfull lot of it. I'll call it "wind". I am sure I will think of a better name for it, later. Why is it so loud?
Oh look at that. It's big and brown and is comming at me very very fast. I think I should name it something.... mmmn it's getting very big.
MMM round.. big and round... ig and round... Ground! that's what I'll name it! ground!
"Hi, Ground!" I wonder if it is friendly..


And I am not sure exatcly what it uses, but one of the random number generators that linux uses is the general state of hardware for the random number seed, Like time, harddrive speed, mouse movement history, video buffer, keyboard input and timing, just what ever is easily accessable I guess. Then It pushes it thru a algorithm.

 

[dejitaru]

Your most recent comments illuminate nothing. If you contend that entropy is not possible, explain why you believe this to be the case. If you contend that it is possible to predict a random quantum transition, explain how this is done. If you contend that all random-number generators are flawed, describe these alleged flaws.
I challenge you to display a physical theory that is coextensive with QED but which allows the prediction of, say, the exact moment when an isolated atom of Carbon-14 will decay.

I've explained it, see above.
Understand that it is impractical to render some theory delineating something which you don't understand.

Prove me wrong.

 

[heliomphalodon]

Sadly, it is you who do not understand.
You appear to challenge conventional wisdom regarding random number generators (and quantum physics) but all you do is make unsupported statements. Your unwillingness to provide any reasoning to support your position fails to disguise your ignorance. There is no need to prove you wrong -- all the literature on the subject does that.
I can see that I'm wasting my time here, arguing with a crackpot.

 

[Shalmanese]

Hmm... its interesting that very little publicity has been made about this in-built RNG. Do any of the Socket A boards have them? Sis? Via? nVidia?

How do you go about activating said generator? Is it something as simple as downloading a special compiler? Does PGP or any other crypto program use it?

 

[ProviaFan]

Originally posted by: Shalmanese
Hmm... its interesting that very little publicity has been made about this in-built RNG. Do any of the Socket A boards have them? Sis? Via? nVidia?

How do you go about activating said generator? Is it something as simple as downloading a special compiler? Does PGP or any other crypto program use it?

When I was installing drivers for an intel chipset based system, I noticed one driver on the Intel website that supposedly enabled this built in random number generator thing. Whether that driver automatically forces applications that generate random numbers to use itself I have no idea, however.

 

[Thraxen]

I challenge you to display a physical theory that is coextensive with QED but which allows the prediction of, say, the exact moment when an isolated atom of Carbon-14 will decay.

Just a question...since the rate of decay of a group of C14 atoms is predictable, the decay of a single atom can't be totally random can it? Seems to me that if is was truly random then something like half life wouldn't be predictable. What am I missing here?

 

[dejitaru]

Sadly, it is you who do not understand.
You appear to challenge conventional wisdom regarding random number generators (and quantum physics) but all you do is make unsupported statements. Your unwillingness to provide any reasoning to support your position fails to disguise your ignorance. There is no need to prove you wrong -- all the literature on the subject does that.
I can see that I'm wasting my time here, arguing with a crackpot.

You keep telling that to yourself.

Perhaps read up on new technology, rather than your outdated quantum-style theories, and you may understand that the Earth truly is not flat.

You make personal attacks when asked to provide some sort of argument? Excellent judgment!

 

[Smilin]

Originally posted by: Thraxen

I challenge you to display a physical theory that is coextensive with QED but which allows the prediction of, say, the exact moment when an isolated atom of Carbon-14 will decay.

Just a question...since the rate of decay of a group of C14 atoms is predictable, the decay of a single atom can't be totally random can it? Seems to me that if is was truly random then something like half life wouldn't be predictable. What am I missing here?

The thing you're missing is that the decay of a group of C14 atoms is NOT predictable - if you get a large enough group of them you can make some really accurate estimates but that's about it. In actuality half life is NOT predictable - but we can estimate darn close...just not down to the moment when that one atom at the halfway point decays.

As far as I know atomic decay on a single atom basis is truly random - as in it will always be unpredictable.

Dejitaru: I'm not following you either here man. Please explain a bit further - I may have more questions afterwards but I promise I won't flame. I'm not getting the concept of how atomic decay can be predicted.

 

[Mark R]

Originally posted by: Shalmanese
Hmm... its interesting that very little publicity has been made about this in-built RNG. Do any of the Socket A boards have them? Sis? Via? nVidia?

How do you go about activating said generator? Is it something as simple as downloading a special compiler? Does PGP or any other crypto program use it?

I don't know if any other boards have such a device integrated.

Intel give techincal specifications on their web site for instructions on how to access it. So if you really wanted to write drivers for it you could.

Under Windows, if you install the intel driver, then the Windows encryption engine will automatically have access to it. This means that the hardware RNG will be used for connections to secure web sites, encrypting files on EFS, etc. Without the drivers installed, Windows will use its own internal software PRNG.

If you want to access it from your own program, then probably the easiest way (on Windows) is to use the cryptoAPI to generate a random number - I've not used it much, but I'm fairly sure that there is provision for this. This approach won't work under Linux though, no doubt, similar drivers exist.

 

[Shalmanese]

dejitaru: If you are going to make allegations like that, please point to some papers, either online or offline which support your thesis or STFU. You have been semi-trolling here in the past so either back up your rants or dont rant at all.

 

[drag]

Nothing is truly random. Every object in the universe affects every other object in the universe no matter how small. I do not beleive that it is possible to reach a small enough scale or a large enough scale in which the natural law of cause and effect do not apply. If one was able to quantify and fomulize everything in the entire universe from the smallest sub atomic particle to the largest galatic clusters then you be able to predict the flip of a quarter with 100.00000... % effectivness as well predict the path of a supercharged electron as it dances around the core of the sun.

However since this is not possible to construct a all-ecompassing formula of the unviverse since it will be as complicated and large as the universe it's self, then pretty much everything we cannot control or experiance directly or thru math is going to be pretty much random, relatively speaking of course.

So entopy of a nuclear particle might as well be random, but predictable in massive enough amounts that we can experiance it thru testing and measuring.

Then the best we can hope for a universal mathmatical law that we can use for the universe is thru quantum physics, when we can hopefully use ALL possibilitys based on there likelyhood of happening to quantify everything. (which I beleive that like the fabeled gun and cat in the box, that there is about 50/50 chance that quantum physics will end up being BS with some significant ramifications kinda like alchamy and the quest to turn iron to gold as side effects)


Am I right?

 

[Lithium381]

Originally posted by: Shalmanese
dejitaru: If you are going to make allegations like that, please point to some papers, either online or offline which support your thesis or STFU. You have been semi-trolling here in the past so either back up your rants or dont rant at all.

i don't know about the trolling stuff recently, but i haven't seen you back up a single statement(speaking to dejitaru).....just some claims.....i want links, references, SOMETHING to back up what you're saying...

 

[Agent004]

Originally posted by: drag
Nothing is truly random. Every object in the universe affects every other object in the universe no matter how small. I do not beleive that it is possible to reach a small enough scale or a large enough scale in which the natural law of cause and effect do not apply. If one was able to quantify and fomulize everything in the entire universe from the smallest sub atomic particle to the largest galatic clusters then you be able to predict the flip of a quarter with 100.00000... % effectivness as well predict the path of a supercharged electron as it dances around the core of the sun.

However since this is not possible to construct a all-ecompassing formula of the unviverse since it will be as complicated and large as the universe it's self, then pretty much everything we cannot control or experiance directly or thru math is going to be pretty much random, relatively speaking of course.

So entopy of a nuclear particle might as well be random, but predictable in massive enough amounts that we can experiance it thru testing and measuring.

Then the best we can hope for a universal mathmatical law that we can use for the universe is thru quantum physics, when we can hopefully use ALL possibilitys based on there likelyhood of happening to quantify everything. (which I beleive that like the fabeled gun and cat in the box, that there is about 50/50 chance that quantum physics will end up being BS with some significant ramifications kinda like alchamy and the quest to turn iron to gold as side effects)


Am I right?

I see where you are going, but consider this:


If Every object in the universe affects every other object in the universe no matter how small ---------> chaotic nature <---------> it is due to this nature, we can't not predict it precisely(hence random). ( I am not about to go into how observing or not observing will affect a paticle)

Also, prediction using mass amounts will only ever going to give you a very good approxmation, which is...... not exact prediction to 'randomness' (I mean, if you have exact prediction, randomness don't exists)

Consider this scenario, what do you think about a universe which only 1 fundamental paticle exists, is it predictable in the path it takes?

Originally posted by: Lithium381

Originally posted by: Shalmanese
dejitaru: If you are going to make allegations like that, please point to some papers, either online or offline which support your thesis or STFU. You have been semi-trolling here in the past so either back up your rants or dont rant at all.

i don't know about the trolling stuff recently, but i haven't seen you back up a single statement(speaking to dejitaru).....just some claims.....i want links, references, SOMETHING to back up what you're saying...

Like you are adding anything postive to this topic.

 

[ndee]

Women are normally unpredictable.






j/k but now a serious question. Why is the "mouse movement" thingie not good? That is kinda not predictable, is it?

 

[drag]

Just a guess:
Mouse movement is pretty good, and combined with other factors it is probably good enough. It's just concevable that you depend on that soly for randomness you could create a program that would emulate the mouse movement and be able to repeat it exactly to factor out it's effect and crack something.

Think about how wardrivers crack wireless ethernet security. 128bit security, heh? sniff out 50000 packets and do and/or sequences on the password section of the packets you can determine the pattern used by the badly designed encrypter and nail the password in a couple minutes and get acccess to their network. if you depended on the mouse you could do repetative type stuff or record mouse movements and figure out patterns and encrypted passwords given enough time. My understanding of the dubject is not that deep, but that makes sense to me. You must defend yourself from outside attackers but if they manage to crack a user account you still want to be able to trap them in a user account.


BTW:

With the one particle thing. It would be easy, because if you are abserving the one particle in the universe, that one particle would be yourself, because if a entity is there to observe something the entity must exist right? so you must be observing yourself. And since relativity tells that we can only percieve movement by comparing the different movements of items around us as compared to ourselves then with nothing else to compare we would be standing still!!! There would be no movement! No path! I can predict it 100% Good riddle!

 

[Agent004]

Good reasoning, but the universe also exists since the particle exists in it. So wouldn't the universe have some effect on the particle?

 

[Shalmanese]

Originally posted by: Lithium381

Originally posted by: Shalmanese
dejitaru: If you are going to make allegations like that, please point to some papers, either online or offline which support your thesis or STFU. You have been semi-trolling here in the past so either back up your rants or dont rant at all.

i don't know about the trolling stuff recently, but i haven't seen you back up a single statement(speaking to dejitaru).....just some claims.....i want links, references, SOMETHING to back up what you're saying...

However, I always qualify my statements with an IIRC or an AFAIK unless I am absolutely certain about it. If pressed, I WILL find a cite but I don't do it as a matter of course since it is not neccesary. If I find out that I was wrong and I cannot back up my claim, then I freely retract it.

 

[Whitedog]

My Wife... Now THERE'S a random number generator.

I never know what my bank acount balance is! :|