[Oatka]

If now there is some algorithm that can predict where those primes exist, them it would be possible to use that algorithm to break the public key. Good-bye security.

I am mathematically challenged but if it can be used to break a current key, could not the decoders go back to previous messages and break them also? If so, lots of new worriesd in the intelligence community.

[stripes1776]

I am mathematically challenged but if it can be used to break a current key, could not the decoders go back to previous messages and break them also? If so, lots of new worriesd in the intelligence community.

Yes, I think you are correct. If you could use an algorithm to break a key, it would not matter when that key was generated--now, in the past, in the future.

[Myrddin]

I am mathematically challenged but if it can be used to break a current key, could not the decoders go back to previous messages and break them also? If so, lots of new worries in the intelligence community.

The common practice is to use of level of encryption that is unbreakable before the value of the information expires. If a message details troop movements tomorrow morning, but takes a week to crack, that is good enough. Just because you figured out the key for one message, you haven't accomplished much. The keys will be changed before the next message goes out.

[Brellium]

I am mathematically challenged but if it can be used to break a current key, could not the decoders go back to previous messages and break them also? If so, lots of new worriesd in the intelligence community.

We were doing that with the Soviets after the second world war.

http://en.wikipedia.org/wiki/VENONA