Zodiackillersite.com

glurk wrote:It's easy to show, here is a cipher: $@*%#!
I could create a key to make those 6 symbols stand for any six-letter word. Without the key, it cannot be solved. Not by the NSA, Sherlock Holmes, or Einstein...

Could you make it a four letter word? I could surely solve that one.

Instead of stirring eachother up we could discuss something more fruitful. Like bringing forward ideas on how to improve our solvers ability to tackle higher multiplicity ciphers.

One straightforward idea is to include even longer ngrams together with a larger corpus. Another idea is to include another measurement that somehow is able to differentiate further between gibberish and meaningful text.

Jarlve-

I agree with you 100%. I don't know why cipher discussions turn into contention and animosity, but they often do. And I've been around the "Zodiac Community" since 2006 or so, and have seen it happen time and again. Makes me sad.

The only possible idea I have is for there to be a semi-private "crypto-only" section of the board to work on these things, but I don't think there would be much activity, to be honest.

But I agree with you, and I personally apologize to everyone here for any animosity I have caused, which was never my intent. I just felt the need to defend my work, after spending 7 or so years on it.

I try to help on the crypto end. It's not for 'everybody' and I know that. I'm going to stick around, regardless.

-glurk

Hey guys... this thread is indeed very interesting, even though swords have been crossed by and by.

That chart you uploaded is a great help to the understanding of us cyber-dunces, glurk. Thanks. Same goes for the simple explanations/reminders of the different cipher/encoding types.

I have done a 408 for you guys to look at. The multiplicity is lower than 0.17, so it's near the "difficulty borderline" on the aforementioned chart. More difficult than 408/340, but less so than jrob's.

And for a reason I thought logical (to me anyway), I have also altered the character/symbol frequency complared to Z's 408. (Sorry if I'm not using the correct terminology here). Like I say, I have a reason, so it's not unreasonable, i think. And I haven't assigned outlandish numbers of symbols to each character.

Like Z, I have split it into 3 parts, and am not letting on about the correct order. But will do if you so require. There may be a bit of of Z-type spelling/phaseology here and there.

So, here's my 3-parted 408 winging it's way to the Glurk Chronicle, Jarive Times Herald and Doranchak Daily News...

q%2@FMGH&WK5Qh/%+
aWL4lc2JbH&#sTWT~
XoB9zxQ0!H<6dlEqR
FJ5oUYw4p6KUbam@7
kmOsxnHAaedd4@QW9
*2poTln%+NQXFcLbU
EO/FX+~qH&%Q¬5AOs
pDB>*W<K4eH9!^2CW

cR^QX</Fqr2d6Ub#D
=+7*U^n9z@WABp+X9
S&!hU#<bQ¬Ob&GDw7
xa0n4ClE/MU+9~oqQ
MPWpp5TOoUYbAnwHD
GTr^E/WlQJq%mN<#O
¬EzTG¬dOxU0FwQ065
bTk&Y*o2O+K4noxEY

hUQp@ao>c4R+%9H7s
xhQ^SMPEUKDl5Id<q
W%9O4JJA#T@!<zk&O
7/~21GH2zqoMH7#I¬
^FAGe*TX¬hRcUrnEa
xDlewa6W0X<zSW^P+
KS25OT&Ms&L¬EOJQr
K%0>T5c%s¬%2+FoHY

I (using ZKD) have already solved this.

THIS IS THE END OF THE BEGIN

Too easy. Amazing what my lousy kludge of a program can do, I guess.

(Sorry, jroberson, you got it wrong)

BE GOOD NOW.

-glurk

EDIT: I shouldn't post the complete solution, should I?

glurk wrote:I (using ZKD) have already solved this.

THIS IS THE END OF THE BEGIN

Too easy. Amazing what my lousy kludge of a program can do, I guess.

(Sorry, jroberson, you got it wrong)

BE GOOD NOW.

-glurk

Great stuff, Glurk... he did it instantly folks BTW.

Feel free to PM me the entire solution.

Like jrob, it took me ages to put together and I learned a fair bit about how mistakes can be made putting such a thing together. And how it's a pain in the neck to do.

I thought the little added-complexity I added might fox dekrypto a bit, but not so!

BTW, the multiplicity as I calculated is 0.16666-recurring.

I'll let others work on it. It made for a good getaway, but it is not so easy.

-glurk

EDIT: Funny thing is I LIKE jroberson, although he is an @$$. That's a cipher, BTW.

Thanks Glurk for the graph. That shows me that the breaking point for programs is problematic. The problem I see is that the computer programs inflexibility into taking to task idiocycronys with deciphering the code. The hardens had the human ability to work around all of the below using reasoning something that a computer program may have only if it's programmed to seek it out.
A spelling errors
B code designer mistakes
C + instead of and
D filler
E the amount of multiplicity encoded by the author.

I would expect a few of the above can take any code over the graph that Glurk has shown.

So my question is .. By reducing the size of the code into smaller sections rather than trying to holy grail the code in one hit. Has this been tried. I understand the larger a code the better..but not if it is out of a programs capabilities.
Break it into halves, thirds, 40 letters 50,60, 80',100 letters at a time.. Draw out a few words.. Surely the code must have some multiplicity weaknesses in parts. Then pen to paper and use the laptop attached to our shoulders. I did have this discussion with jarlve but don't know of any results, got a bit busy with life for a while...

And I wish not to cross swords.
Cheers.

Thank you very much RTF for this pen and paper cipher! It probably will be of good long term use for me since it is reasonably similar to the Zodiac ciphers.

Mutiplicities:
Zodiac 408: 54 / 408 = 0.132
RTF 408: 64 / 408 = 0.156
Zodiac 340: 63 / 340 = 0.185
jroberson 405 (_pi transcript.): 99 / 405 = 0.244

When reducing the length of your cipher my solver (AZdecrypt) still comes up with reasonable solves up to around a multiplicity of 0.225. Maybe with a few adjustments my solver could be in range for jroberson's cipher. Something for later.

@Mr lowe,

Reducing the length of the cipher, by excluding parts as you mention, only increases the multiplicity and therefore the difficulty of the cipher because: multiplicity = (unique number of symbols) divided by (total number of symbols).

Jarlve wrote:Thank you very much RTF for this pen and paper cipher! It probably will be of good long term use for me since it is reasonably similar to the Zodiac ciphers.

Mutiplicities:
Zodiac 408: 54 / 408 = 0.132
RTF 408: 64 / 408 = 0.156
Zodiac 340: 63 / 340 = 0.185
jroberson 405 (_pi transcript.): 99 / 405 = 0.244

Thanks, Jarlve. And to Glurk for running this thru ZDK. And apologies for getting the multiplicity wrong - i included a couple of characters that went unused (in my 408) in my calculations.

In a while I'll make a few observations and comments on the construction of my 408, and what it's brought to mind about Zodiac's ciphers. As a non-cipher guy myself, I hope they may have some validity or usefulness - esp. if it is considered that Z himself was merely dabbling in the art.