# true random numbers

• August 14th, 2003, 11:40 PM
johnnymier
true random numbers
There is an open source project going on to generate numbers from a random source. The process is made by using a CCD chip to capture some digital data, since the CCD responds to lighting there is a source generating random values of iluminance, which is then processed by some algorithms, generating unpredictable numbers. Could be really useful in the future of cryptography.

full story:

http://slashdot.org/articles/03/08/1...tid=172&tid=93
http://www.lavarnd.org/
• August 14th, 2003, 11:48 PM
Mephisto
Why is it that random numbers are useful in cryptography, and if the numbers become unpredictable then how would you be able decrypt it later on?
• August 15th, 2003, 12:07 AM
slarty
Random numbers are essential in cryptography to pick a key.

Keys must be picked at random. Various protocols require keys to be chosen more or less often, with a varying number of bits.

Anyway the whole point of a key is that you (an attacker) don't know what it's going to be. If it's picked non-randomly, that leaves an attacker the possibility of guessing it at a better than random chance, which means that the key can be broken more easily than if it were truly random.

Slarty
• August 15th, 2003, 12:12 AM
Mephisto
Aah, I see. Thanx for the explanation, my thoughts wandered of in a completely different direction ;)
• August 15th, 2003, 12:11 PM
valhallen
sounds a bit like the random numbers they were generating using lava lamps - lol
they also are generating them now using sound - you know the kinda static hiss you get?? I remember reading an article about it but am not to sure of the actual way in which it is used - will try and hunt it out and post it when i get back from work

v_Ln
• August 15th, 2003, 06:23 PM
GandalfTheGray
Sounds generally related to a project I worked on long ago -- generating random numbers from radioactive decay. One problem was that electronic detector biases reduced the random nature of the detected events, causing some concern that the key would be easier to "guess" than you might think for a truly random process like radioactive decay.
• October 11th, 2003, 12:58 PM
maxim_86ualb2
For now the best Crypto thing..... is the 1time pad methode........ I have read an artice about random #'s using a camera..... & watshing water flow......I dono about it... alot..... but I think... if U use a program...... dam... ther is nothing Random.....
• October 11th, 2003, 02:10 PM
darkes
Quote:

Originally posted here by maxim_86ualb2
For now the best Crypto thing..... is the 1time pad methode........ I have read an artice about random #'s using a camera..... & watshing water flow......I dono about it... alot..... but I think... if U use a program...... dam... ther is nothing Random.....
You are quite right that a 'one time pad' is the most secure method if used correctly.
What this means is that the two people exchanging information agree the key(s), and usually meet in the flesh to exchange it (them).

By definition, each key is only used once, and then destroyed.

The reason it fails, is if you use a so called random number, or a quotation from a book, for example, then this is easily cracked.
• October 11th, 2003, 03:49 PM
spools.exe
I would like to point out that this its very hard to generate a true random number. With visual baisc the random number function "int(rnd * 10)" if not actually random. Say you had a function on Form Load that was something like this mshbox int(rnd * 10). You would think that that would choose a different random number everytime the form laods. But in actuallity it does not. It will popup with the same number everytime that form is loaded.
• October 11th, 2003, 04:42 PM
souleman