| Welcome to Crypto. We hope you enjoy your visit. You're currently viewing our forum as a guest. This means you are limited to certain areas of the board and there are some features you can't use. If you join our community, you'll be able to access member-only sections, and use many member-only features such as customizing your profile, sending personal messages, and voting in polls. Registration is simple, fast, and completely free. Join our community! If you're already a member please log in to your account to access all of our features: |
- Pages:
- 1
- 2
| The NSA May Read My Mail; or not? | |
|---|---|
| Topic Started: Sep 2 2005, 07:59 PM (1,217 Views) | |
| insecure | Oct 11 2005, 03:22 PM Post #16 |
|
Elite member
  ![]](http://209.85.122.85/static/1/pip_r.png)
|
No, I haven't read "Digital Fortress". Judging by what I've heard from others here, the guy doesn't know much about crypto. If your email can be decrypted without the proper key by fewer than 10^50 computers working at 10^20 brute-force decryptions per second in less than 1000 years, you're probably not trying hard enough to encrypt it. Before a computer can do it, a human has to work out - at least in principle - how to do it first. A human can work out how to crack a Vigenere cipher. (This has been done.) The computer can't unless it is told how to do it - by a human. That is why a cryptographer's most fearsome adversary is not the computer, but the cryptanalyst. As for email encryption meaning you have something to hide - let's just say you wanted to send something through snailmail - the details of your new invention to the Patent Office; or perhaps a specimen signature to your bank for your new bank account; or even just a somewhat personal letter to your girlfriend. Would you send the information on a postcard? Or would you put it in an envelope? It is not the government's job to read my emails, or your emails. It is the government's job to protect the people from harm by outsiders (i.e. an army), and from harm by other people (i.e. a police force), and to arbitrate in case of dispute (i.e. a court system), and that's about it. The rest is stuff they've taken upon themselves, and nobody has got around to stopping them yet. In any case, any terrorist clever enough to accomplish a terrorist act is certainly clever enough not to use easily-cracked lamebrain encryption systems. |
 |
 
|
| Donald | Oct 11 2005, 03:49 PM Post #17 |
|
Elite member
|
It's a bit difficult for a human brain to wrap itself around the actual SIZE of a 128 bit keyspace, let alone 256. Even with a building full of computers, brute force is just not a practicle attack on AES or just about any modern cipher with a large key. Also, as I said once before in this thread, NSA recommends AES for our entire banking system. Even for our goverments use. Now if the NSA thought that AES could be decrypted by a building full of computers, then they would have a problem on their hands. LOTS of people can afford a building full of computers. Certainly most major governments could. The NSA's job isn't just to spy on people, it's also to protect the US's security from outside attack. If the NSA can break AES, someone else could. And that is a risk to the security of the US that simply could not be made up for by the advantage of being able to spy on a few people. A hostile outside force with access to the keys to the banking system of the US would be devistating, let alone the military codes. So, I think you can feel fairly confident that anything encrypted with AES is not likely to be read without access to the key. And it would be MUCH easier to come to your house and beat you until you tell them the key then it would be to brute force it. So, to echo Insecure, I think we would all be better off if everyone encrypted all of their email. While admitting, that until everyone does, encrypting your email DOES draw attention to you. If I really had something to hide, I would be afraid that encrypting my email might bring some guys to the door with a rubber hose and a mission to bring home my keys... Donald |
|
|
|
| cows | Oct 11 2005, 04:21 PM Post #18 |
|
Unregistered
|
Firstly sorry for the username confusion - i am the same user as --cows-- but i am getting it sorted. You guys all make a valid point and i bow to the superiority of you all especially as you are all older than me and have a hell of a lot of more knowledge than me. But, how do we actually KNOW 100% that there is no super computer that IS told by criptographers how to break ceaser ciphers and Vigenere ciphers and all the others - why wouldit not be possible. Sure other businesses COULD buy hundreds of computers and set their purpose to deciphering email codes, but the mere fact that there is no limit on the amount of computers that a company can have or searches to make sure that what they say they are doing, is true. Then they may be put off from trying to work out a way to get a computer to do it for them. Because nobody is trying to stop them, they will not try. Just like if somebody says "Do not watch T.V. after midnight," or "Don't eat the glue" your first reaction is to want to watch the t.v. after midnight and eat the glue whilst doing so. But, if nobody said anything about the t.v. or glue. The chances are that you wouldn't watch the t.v. after midnight and you definately woudn't eat the glue. Same applies, because companies are not told not to buy lots of computers to break email ciphers, they won't. I know it's reverse phsichology and all but it is true to an extent, obviously as an adult you wouldn't eat the glue, you would do something else but you get the point. There is a joint operation thing going on at the moment where you donate your Processor speed to a company while you are not using it (say from 12:00 am to 6:00am) and they send the computer 'missions' to complete in that time or to have a go at. Surely if you had the processor power of over 1 million computers it would be possible to break a 128 bit code or even a 256 bit code. Why not. Sure it may be hard to do, but most things in life are. Sure there may be reports that say that it is not true - maybe they are lies. This is very argumenattive i know but why is it so hard to believe that this is not possible. The world of computers is sofisticated enough to send an email in under a second and download gigabytes of files in under 5 minutes. Why shouldn't this be possible? Also - just to bug the hell out of everyone and prove the point a little further PROVE IT --cows-- |
|
|
|
| Donald | Oct 11 2005, 06:11 PM Post #19 |
|
Elite member
|
Take the following quote and replace "Space" with "128bit KeySpace"
Another quote I found that attempts to explain the how big 128 bit keyspace is.
If we leave the questionable quantum computers out of the equation, Modern cipher breaks don't come by brute forcing keys. They come about through weaknesses in the cipher, or more likely, in the implementation of that cipher. It is just SO much easier to break into someones house and put a keylogger on their computer than it is to brute force 128bit key space. And it would be one heck of a lot cheaper to just pay them a few million dollars to give you the key. And if you don't have much money, rubber hoses are cheap and in abundant supply. Donald |
|
|
|
| insecure | Oct 11 2005, 06:42 PM Post #20 |
|
Elite member
|
Don't worry about the user names: we guessed you folks were the same folk anyway (after all, the names were presumably similar on purpose). Computers can break Caesar and Vigenere ciphers trivially - and do. And they can break stuff more complicated than that, too. But only because a human has first worked out how to break them, and instructed the computer accordingly. You talk about "no limit" to the number of computers an adversary can have, but what you really mean is "lots". It is possible to design a cipher so that "lots" isn't enough - provided the algorithm is solid enough to hold off the cryppies. Take, for example, a 256-bit cipher. Let's pretend, for the sake of argument, that we can perform 1000000000000 trial decrypts per second on a single computer. We can't, but let's pretend we can. And let's say we want to guarantee to retrieve the plaintext in under 100 years. How many computers will we need? We have 3162400000 seconds to play with, which means we can try 3162400000000000000000 keys on a single computer during that 100 year period. (I have been generous to the attackers, and assumed that every year is a leap year.) How many possible keys are there? 115 792 089 237 316 195 423 570 985 008 687 907 853 269 984 665 640 564 039 457 584 007 913 129 639 936, that's how many. So the number of computers we need is the huge number divided by the large number. That turns out to be: 36 617 109 782 089 972 748 295 823 532 903 229 310 004 928 362 692 447 138 565 568 713 289. Let's assume (rather adventurously) that each of these computers occupies just a single millimetre cubed in volume. If we think of that as a cube shape, it's over 33000000000000000000 metres, or well over 3400 light-years, on a side. You see? If you get the algorithm right, and have a big enough key, it's simply not going to be brute-forced. And that means you need a human to figure out a weakness in the algorithm. Computers suck at that kind of thing. |
|
|
|
| Donald | Oct 11 2005, 06:51 PM Post #21 |
|
Elite member
|
Insecure beat me to the punch by just a few minutes again.
But heck, I'm going to post my numbers anyway.
Check me on this, I'm bad at math guys. 128 bits in decimal = 3.4*10^38, or, to write it out: 340,000,000,000,000,000,000,000,000,000,000,000,000 BIG just can't quite describe this number. And, personally, I use 256 bit encryption for many things. Talk about overkill. 256 bits in decimal is 1.2E+0077, or, to write it out: 120,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000 The total number of atoms in the universe is estimated to be between 4*10^78 and 1*10^79. So 256 bits is only 1 or 2 orders of magnitude off of that. According to wikipidia, the fastest computer as of March 25, 2005 does 135*10^12 floating operations per second. So lets assume comparing a key took ONE floating operation. (It takes a LOT more than that), todays fastest super computer would still require 2.5*10^24 seconds to explore 128 bit key space. Thats 2,518,518,518,518,518,660,000,000 seconds, or approximatly 7.9*10^16 years. To write it out: 79,220,219,535,072,368 years. So, IF you had a building with 79,220,219,535,072,368 copies of the worlds fastest computer, AND if it only took 1 floating point operation to test a key (NOT!), you could brute force a 128 bit key in one year. But my 256 bit encryption would still take 1.5*10^60 years to completely explore the key space. Remember, thats with 7.9*10^16th copies of the worlds fastest computer. 1.5*10^60 years. Keyspace is BIG. really big. Donald |
|
|
|
| insecure | Oct 11 2005, 07:31 PM Post #22 |
|
Elite member
|
The numbers look about right to me, Donald. It's interesting how people choose analogies to represent big numbers. You seem to favour the "it'll take ages" approach, but I've always liked the "how big a lump would it make?" approach myself. When the decrypting cluster is thousands of light-years across, it conjures up a wonderful picture. Such a cluster, even if it could exist in the first place, would have to be very carefully designed if it were to work - even the task of dividing up the keyspace between the machines would have to be very delicately planned. |
|
|
|
| Donald | Oct 11 2005, 07:37 PM Post #23 |
|
Elite member
|
Actually, I think the BIGGEST problem would be making certain the cluster doesn't collapse into a black hole under it's own mass. Donald |
|
|
|
| insecure | Oct 11 2005, 08:13 PM Post #24 |
|
Elite member
|
I nearly said as much in my original reply, but changed my mind. I don't think "black hole" is even remotely adequate to describe a phenomenon of that size! |
|
|
|
| cows | Oct 18 2005, 02:57 PM Post #25 |
Advanced Member
|
o.k. i stand corrected. *bow in humble way to masters and shuffle back towards door* How many processors did the worlds best computer have by the way? Also even though it is fictional - 3000 processors were said to take only 10 - 15 mins to work out a 128 bit key, and the math had to come from somewhere. (unless it was a completely random number  )
|
|
Everything is possible, The impossible just takes longer If we do not know what a particle is doing then it is allowed t do everything possible simultaneously. "Anyone who can contemplate Quantum Mechanics without getting dizzy, didn't understand it." | |
|
|
|
| insecure | Oct 18 2005, 06:47 PM Post #26 |
|
Elite member
|
Let's give them 15 minutes, which is 900 seconds - in fact, let's give them 1000 seconds - and there are 3000 of them. 340,000,000,000,000,000,000,000,000,000,000,000,000 combinations to try. Keys per processor = 340,000,000,000,000,000,000,000,000,000,000,000,000 / 3000 = approx 100,000,000,000,000,000,000,000,000,000,000,000. Keys per processor per second = 100,000,000,000,000,000,000,000,000,000,000. So all you have to do is find three thousand processors each of which is capable of testing a key against a ciphertext and analysing the resulting plaintext, 100,000,000,000,000,000,000,000,000,000,000 times a second. Assuming your processor design is custom-built and dedicated to the task (which makes it pretty well useless for anything else), and assuming it can do an entire test in a single clock cycle (BIG asssumption), you just need a processor running at 100,000,000,000,000,000,000,000 GHz - i.e. approximately a hundred thousand million million million times as fast as a typical family desktop PC. |
|
|
|
| colly | Dec 8 2005, 10:51 PM Post #27 |
|
Just registered
|
You are assuming that they will have to use brute force to crack the cipher text, there are other ways of cracking many ciphers without needing to check every possibility. |
|
|
|
| insecure | Dec 9 2005, 03:54 PM Post #28 |
|
Elite member
|
Yes, that's precisely the point we're making - that brute force is impractical if the algorithm is reasonably good. So an adversary must come up with a much, much, much better attack than brute force. |
|
|
|
| sas01 | Dec 31 2005, 09:09 PM Post #29 |
|
Just registered
|
Like ask the target politely
|
|
|
|
| cows | Jan 1 2006, 10:59 PM Post #30 |
|
Advanced Member
|
Never know - it might work Excuse me Mr Blair - I want to read all your outgoing mail but need to decrypt it first - wanna give me the key?? Or maybe you could just send it to me un-encrypted first. 
|
|
Everything is possible, The impossible just takes longer If we do not know what a particle is doing then it is allowed t do everything possible simultaneously. "Anyone who can contemplate Quantum Mechanics without getting dizzy, didn't understand it." | |
|
|
|
| 1 user reading this topic (1 Guest and 0 Anonymous) | |
| « Previous Topic · Debates · Next Topic » |
- Pages:
- 1
- 2
10:42 AM Nov 23
