unforum

[SoCalSteve]

Cypher This...

[Rogi Ocnorb]

I'd be really surprised if the first thing to do isn't reversing the ones and zeroes.
After converting that result to ASCII, you get what looks a lot like Dvorak created from ASCII, But it's not.
I can't make anything out as a simple substitution, either.
Applying a -4 or +27 ASCII ROT to the result gives a string that's all letters, but would need anagramming or is the input to yet another layer.

[pancito]

Just looking at big picture here. There are 344 individual characters, which is neatly divisible into 43 groups of eight. I thought we might want to look at a smaller set, like groups of 5, 6, or 7 but none off those divide evenly, and 4 only gives 16 possible combinations, so that means it was converted to binary as one of the last steps. You could do some bitwise operations, like rotating the binary or something, but that really gives too large a solution set, doesn't it? I mean inventing an uncrackable cipher is not that difficult. The trick is to keep the message short enough that the breakers don't have enough to work with. So you always have a trade off between security and usefulness, at least in the way we normally talk about ciphers here. Obviously there are ciphers out there that are both useful and functionally uncrackable. Try PGP for a starter. The question is, is it a method that's easy to remember and work? All that aside...

I'm going to start by agreeing with Rogi about the reversing ones and zeros and see where that leads. Doing that yields

OITXS QYQKP QM\^R [EZSH JVLYG ^^TU^ ]\QHM P\M]P M[U

And here's the ROT-4

KEPTO MUMGL MIXZN WAVOX JRHUC ZZPQZ YXMDI LXIYL IWQ

Frequency analysis shows B, F, S =0; A,C,D,E,G,H,J,K,N,R,T,V=1; O,P,Q,U,W,Y,=2; L=3; I,M,X,Z=4
The only double letter is ZZ. Repeating two letter combos = none.
The Vigenere Guesser came up empty.

So my guess is yeah, if it gets cracked it's because someone lucks into it. But you haven't proven it's uncrackable. Send us a 500 character message [or 12 43 character messages], assure us that the method is actually usable ["a series of well-defined steps that can be followed as a procedure"] and that special knowledge isn't required [i.e. that it's really a cipher and not a code. Link].

BTW, anything on the right side of the diagram below is pretty much beyond our usual purview. Even the rotor machines require that we get some clue as to the decryption key. We do classical ciphers here almost exclusively.

[TheIllustratedMan]

Yeah, an uncrackable code? Try this on for size:

300416444444456140360456468444412

[SoCalSteve]

Good Point