Enigma machine - encryption for the 1940s?

[Spike]

I was just looking at the "Odds and Sods" on theregister.co.uk, and they have an article telling of the sale of a genuine Enigma machine on Ebay, and I recall that there is a distibuted computing project running to decode four enigma transmissions, two of which I believe are complete.

The thought occurs to me that if the Enigma machine could so encode a message that it would require so much computing power to decipher it, why should it be only a relic of history?

Would it actually be possible to write a software equivelant that randomises the (virtual) starting positions, and gives you a reading of what those positions are, and then encrypts alpha-numeric data (which I imagine would make the binary of it pretty much undecipherable).

The same program could then be fed the correct staring positions (as per the original read out) to decrypt the data.

If this could be done, what would be to stop someone enigma encoding the same piece of data 2, 3, 5, or 10 times? (with the exception of the fact that it would then have to be decoded an equal number of times which would require the storage of each encoding start position in sequence).

I'm pretty sure it wouldn't be much use for data transmission, but if possible, would it not make an excellent method of securely storing info on your hard drive?

 

[kirock]

A quick response because I haven't read on the subject of encription science in some time now, is that this is exactly what modern encriptions do. I think it was in the early 50's when there was a break through concept on the whole theory of encription, the use of a Key. I think the rudiments are still substitution and position transfer (man I wish I has a better memory, I read a paper on this in Scientific America in the mid to late 90's) but I'm forgetting the details of how the key works.
Bottom line is todays encriptions are 8, 16 and 32 bit (last of which is top secret military and CIA stuff)(some commericial banks etc use a form of this too). If I remember correctly (not bloody likely :blackeye: ) the 32bit code would require something like the factoral of 10^32 computations to break it without the key. In todays modern computer power the universe would have long collapsed into a black hole again or drifted off into infinity and all the stars are long dead before they could complete the req'd number of calculations. Then all that would be left is this Super Computer orbiting a black hole and using its quantum singularity as a power source and suddnely the program stops and the Computer declares: Eurika! the answer is: "Bring home milk and bread".

Now you've peek my interest and I'm going to have to read up this again. Great thanks a lot Spike really appreciate it.

 

[Nodsu]

Unbreakable encryption is not a problem. There is an infinite number of transformations you can apply to data and it would take infinite amount of time to brute-force the result.

The problem is the "shared secret". Once you do your n-times Enigma thing (or any other encoding), you have to let the recipient of your message know what your starting positions are (unless you agreed on them beforehand). So you either have to transmit the starting positions (unsafe - someone may intercept) or you have to use a fixed starting positions set (unsafe - it can be figured out eventually) or you have to agree on some algorithm to change the starting positions after every transmission.