BootMii as boot2 on new Wii's

Quote
bg4545

Quote
syncmii
On new Wiis, you cannot install bootmii as boot2. hackmii is smart and removes the option when it detects newer wiis.

I'm pretty sure we all know that.

I smell a troll.

Same person as davidcalman_32 or whatever his name is.

Quote
SifJar

Quote
elmoreas
Would they not also need a Wii and a PC too? I know that is a stupid question but I was told earlier the Wii had nothing to do with the Privated key so I was wondering what the bruteforcer ran on the PC actually bruteforced the key out of? Thanks.

Elmoreas

A Wii is not needed. Anything encrypted with the private key can then be decrypted with the common key, which is widely known. The bruteforcer would just guess the key, sign something with it, then try to decrypt it with the common key and compare to the original. It doesn't "bruteforce the key out" of anything, so to speak, it just keeps guessing till it gets it right or is stopped manually (i.e. closing the program).

Well crap. Just thinking a little simply here, so someone point out any inaccuracies , if someone was gonna write an app to do that, I was willing to run it. If you figure that each of these attempts would take 10 seconds or so. That being said since the key is RSA-2048, it is comprised of 2048 bytes, or 618 decimal numbers. Therefore, to figure out the amount of time required, you would merely need to take the factorial of 2048, to figure out how many possible combinations there are, and then, multiply that by 10, and you'll have the number of seconds required. However, every calculator I have used has computed 2048! as ?. I'm going to say that's not technically correct, since that is assigning ? to a definite value. I'm just going to say however that that is a REALLY F'ING BIG number, and as is such, would not be feasible to run an attempt in this manner. More likely to get a hit than with a brute force, but, still not feasible. Now, we could increase our chances by having a bunch of people run this at the same time, but still not highly likely. Seems to be the best chance there is right now though, is it not?

It's not 2048!, it's 2^2048. Substantially better odds, but still terrible.

RSA-2048 is 2048 bits, not bytes.

Even if it was bytes, 2048! would still be wrong. 2^16384 would be correct, and even that would be somewhat lower than 2048!.



Edited 2 time(s). Last edit at 02/16/2010 03:01AM by WikiFSX.

Quote
WikiFSX
It's not 2048!, it's 2^2048. Substantially better odds, but still terrible.

RSA-2048 is 2048 bits, not bytes.

Even if it was bytes, 2048! would still be wrong. 2^16384 would be correct, and even that would be somewhat lower than 2048!.

Sorry about that, I just misread a brief quote I read up on it, and used that in my "equations"

Still, even though it'll probably never work, I say we give it a shot anyway, what's the harm?

Quote
cactusjack901

Quote
WikiFSX
It's not 2048!, it's 2^2048. Substantially better odds, but still terrible.

RSA-2048 is 2048 bits, not bytes.

Even if it was bytes, 2048! would still be wrong. 2^16384 would be correct, and even that would be somewhat lower than 2048!.

Sorry about that, I just misread a brief quote I read up on it, and used that in my "equations"

Still, even though it'll probably never work, I say we give it a shot anyway, what's the harm?

If you decide a random atom in the visible universe to be "the right one", then randomly guess an atom in the visible universe (there are about 10^90 of them) until you find that "right one". This has pretty much (but still, kinda far away) the same likelyness to happen as to guessing the key if one guess equals finding the right atom 6 times IN A ROW!

There's really no harm but the problem is: We wont find it unless we find some better method or get an extreme (EXTREME!!!) amount of luck.



Edited 1 time(s). Last edit at 02/18/2010 11:25PM by profetylen.

Quote
cactusjack901

Well crap. Just thinking a little simply here, so someone point out any inaccuracies , if someone was gonna write an app to do that, I was willing to run it. If you figure that each of these attempts would take 10 seconds or so. That being said since the key is RSA-2048, it is comprised of 2048 bytes, or 618 decimal numbers. Therefore, to figure out the amount of time required, you would merely need to take the factorial of 2048, to figure out how many possible combinations there are, and then, multiply that by 10, and you'll have the number of seconds required. However, every calculator I have used has computed 2048! as ?. I'm going to say that's not technically correct, since that is assigning ? to a definite value. I'm just going to say however that that is a REALLY F'ING BIG number, and as is such, would not be feasible to run an attempt in this manner. More likely to get a hit than with a brute force, but, still not feasible. Now, we could increase our chances by having a bunch of people run this at the same time, but still not highly likely. Seems to be the best chance there is right now though, is it not?

That IS a bruteforce. What did you think a bruteforce was? Its where you have a program which keeps guessing every possible key until it gets it right.

Hi all!

(haven't written anything here for a while, but this interesting thread excites me)


I see 2 big problems, which could prevent us from bruteforcing the Wii's key.


To bruteforce any encryption you need to know three things:
-the algorithm
-a lot of encrypted data
-a lot of unencrypted data


Why do we need "a lot" of data?
I am no expert on RSA, but in general the following can (and will) happen, if you try to bruteforce:
You will have mutliple possible results, so you may think you have found the key, but in reality you may have just found a key, that fits the datat you have, but will not work on other data.
So every algorithm has a minium amount of data requirement.
Does anybody know how much data is required for RSA-2048? (I don't)
Since Nintendo uses RSA2048 only for the small tickets, I believe it may be possible, that we actually may not access to enough data.


And let's talk about time:
In this tread somebody said, it could take the age of the universe to bruteforce the RSA-2048-key.
Well, let's take thing down to earth :)
I once read, how long it would take to bruteforce AES-256, another encryption algorithm, which the Wii also uses.
Let's assume, that we use every bit of silicon mother earth has to give to build CPUs. If all of them would be working together, it would take them over a year to try all possible combinations.
And the earth does not have enough energy resources to power al those CPUs :)

Granted AES is a lot more complex then RSA, which is actually quite simple, but to compensate that, RSA uses longer keys.


And Nintendo could always upgrade it's encryption :)

To 'crack' RSA you only need the public key (which we have), which is the product of 2 prime numbers. You 'just' need to find those prime numbers.

[en.wikipedia.org]
[en.wikipedia.org]
[en.wikipedia.org]

RSA-100 has 330 bits (so it's really a RSA-330) and:
"It takes four hours to repeat this factorization using the program Msieve on a 2200 MHz Athlon 64 processor."
RSA-640 has 640 bits and:
"The computation took 5 months on 80 2.2 GHz AMD Opteron CPUs."

That's from 4 hours on 1 CPU to 5 months on 80 CPUs for double the amount of bits. Which is a factor ~900 in time and a factor 80 in CPUs.
Scale up to 2048 and you need to double the amount of bits 2 more times, which is 337500 years on 320 CPUs.


(Quantum computers should be able to factor numbers really fast, but we don't have those yet)



Edited 2 time(s). Last edit at 02/19/2010 03:57PM by Daid.