Running key 6-25-19

Here’s a randomly generated Vigenere running key cipher. I solved it without a crib using a key search program and a combination interactive solver and hill-climbing program. Both of these programs used local servers to run. The word search server was listening on port 3000, and the hill-climber was listening on port 8080. The search program ran through a list of about 600,000 words and phrases, looking for a corresponding plaintext that included a word or phrase at least 7 letters long. The key search returned four possibilities, one with a 9 letter phrase and three with 7 letters phrases. Three out of the four results actually occurred in the solution. I filled in the rest of the plaintext by switching back and forth between the hill-climber and the interactive solver.

For a crib I’ll include the (caesar-shifted) phrase whose corresponding plaintext included the 9 letter phrase.

Vigenere running key:
UUOHZ LWHJM AATPD XIEZC TVYAO XPJBZ OXFHC VNTLV XZWYD YDYZX JXMLX PMHDZ PXEOS.

caesar-shifted crib: NKSJHJXXFWD

update: I’ll add a caesar-shifted hint:

Hint: YMJ XTQZYNTS XYFWYX BNYM F SFRJ FSI JSIX BNYM F SFRJ

2 thoughts on “Running key 6-25-19”

  1. I spent about three hours, maybe more, trying to get a wedge without the crib, and came up with a good-looking pt/key combination: ATLENGTHL/FORRANSOM, but I wasn’t able to extend it. I finally gave up and used the crib, only to find that it locates right where my pair was. So that combo I found is a false solution. I have extended the crib a bit, I think, but I’m not sure my extension is right. This is a toughie.

    I just now decrypted the hint, too. The last few hours I spent writing a program to identify frequent starting words and phrases and testing them. That turned out to be wasted time.

    Edit: 9:54pm.
    I finally got it, but only because I had succeeded in extending the crib enough to look up the quote. Once I got a few more words I was able to extend it by hand to the end, but I never was able to work backward to the beginning. That initial tetragram is a killer. BION, I’m in awe of the program you created to solve it.

    1. Two of my search results overlapped and couldn’t both be correct. The wrong one was “sawithappen” and the correct part of one was (caesar-shifted) NFRLTNSL . I tried both of them with my hill-climber, and it wasn’t clear for a long time which was correct.

      Speaking of hill-climbers. I think running key ciphers are another case where scoring trial decrypts using logs of tetragraphs isn’t good enough. My hill-climber was not very successful until I switched to scoring trial decrypts using logs of pentagraphs plus word-list scoring. Also the “word list” includes a large number of two-word phrases.

Leave a Reply to The Rat Cancel reply

Your email address will not be published. Required fields are marked *