3 ms·
Super interesting, thanks! This is also a good companion video -- https://www.youtube.com/watch?v=RzWB5jL5RX0 https://www.youtube.com/watch?v=RzWB5jL5RX0 -- abo
by gregsadetsky 5y ago
Super interesting, thanks! This is also a good companion video -- https://www.youtube.com/watch?v=RzWB5jL5RX0 https://www.youtube.com/watch?v=RzWB5jL5RX0 -- about the feasibility of cracking Enigma "in 2021" i.e. using computational power available to us today.
TLDW & AFAIK: it's crackable (just as it was back then, using a bag of approaches & heuristics), but you still couldn't plain brute force an encrypted message today if you didn't have any portion of the plain text, the Enigma was using 10 plugboard pairs, etc.
Also see https://crypto.stackexchange.com/a/33631 https://crypto.stackexchange.com/a/33631
- nullc 5y ago> but you still couldn't plain brute force an encrypted message today The analysis you link to shows that it has about 76-bits of keystate based on the most generous analysis for the 3 of 5 rotor machines. This should be within the range of a motivated nation-state attacker using custom hardware to just outright brute force. Even if you don't know the plaintext, you can do a quasi plaintext attack where you just try every long enough word in the dictionary at every non-prohibited location and apply the classic attack-- and I expect this would still be much faster than a brute force search while only having extremely limited assumptions about the plaintext.