I remember we did look at skytale some time ago. It is the same as wrapping around the cipher instead of incrementing start positions (1,2,3,4,5…). With our algorithm, the period cannot share a factor with the cipher length or else only a part of the cipher will be repeated over and over again. That is interesting because it means that 19 can be used since it is prime and it does not divide the cipher length in a whole number. But 15 cannot, since it shares the factor 5 with 340.
I have ran the basic skytale again through my solver considering all start positions with the normal and reversed 340. The reversed 340 returns quite a bit better than the normal. Here’s the best solve. This shows that very likely, the 340 does not have a clean period 19 skytale transposition.
268 Score: 20681.95 Ioc: 0.07178553 Chi2: 52.41551 H: 3.881533 M: 0.1852941 C: 340 S: 63 hehissthedesimars tolearchisffornit atehowcaparteraft moutstattracoinas pandayomarterownt hemtheadinesstdca ularlalisranineis handtasthmaliissh iscondistreseculc insenclarianniawa tticsseameachoitw illdrillionwhoutd ayssoundslisundem anceanmcleelmostr umidtrohadaroomsh anchesalsoteairid emithinthedoutrie rhcoachdiatinfice iteairschetrolthe celleofhenothernl f>4+RR-L.k5*+BEc( W^C5ycU4+Gddlcp+- TFO4^92yAYc-t7ydK BS<K*FzKFc_2l+p)G AYP6zXlBT7FO%^9JK LOBWL.y8+pORjK82z <|z7CT|+(MYJ+J5/R LYPkW)R-4BT|++G(4 +(D^Nk+GZc5*tU<|2 +pGOpD|)MVbNJ/z9z FFVU(R>)B.E2#l+K@ VC|8M+HC/lp9fS<-6 zX*j^:N;*H+(:N6OB yNUOzpB2|OOCB1GFM <B+;FMSfT8_c13BR& bp2#qR_H*1Ztz+c+; .B+Zf+pK#.kS<Wc+t c4D^b2#k+)WVpd+25 VF5EVM(D&OW%lHFLq U.||5ldL>P3Z#>Mp|
I’m preparing a very big and exciting skytale AZdecrypt test for the normal, reversed, mirrored and flipped 340. I’m guessing a few million ciphers each.
At each position of the 340 a string of up to 16 characters will be removed and ran through the skytale routine considering all starting positions. In other words, it will be checked if there exists any single continues fragment or string in the 340 from 1 to 16 characters that is a null (for instance the "ZODAIK" fragment). If the answer is yes then the routine should get the solution.
Hang on, could take a while!
Edit: removal of 1 to 16 characters for the period 19 skytale but for period 15 only the compatible message lengths (338, 337, 334, 332, 331, 329, 328, 326) as explained in previous post. Some small sample tests showed that the vertically flipped 340 with the period 15 skytale shows the strongest response (highest score and average).
looking forward to your work jarlve.. I am presuming it will end up being tested by column’s and rows.
the only other method that i can think of that may fit the construction to suit period 19s and 15s is a rail fence type cipher. probably already been tested.
I’m preparing a very big and exciting skytale AZdecrypt test for the normal, reversed, mirrored and flipped 340. I’m guessing a few million ciphers each.
Nice! How many ciphers can you process per minute on average?
Thanks Mr lowe,
@doranchak, I’m running the test on my refurbished Dell T5500. 4 instances of AZdecrypt, one for each orientation (normal, reversed, mirrored, flipped) averaging about 50 ciphers per minute each so that’s 200 ciphers per minute @ 500.000 iterations and 10 random restarts per (about 20.000.000 iterations per second). It will take weeks. Right now the highest scoring result (20895) is from the reversed 340. I’m also sorting different message lengths by directories.
This test ofcourse is not the ultimate answer but when it is finished we may be able to narrow down things a bit for the skytale scheme.
That’s great! I look forward to the results. Please come up with the solution because we need to finish off this massive thread once and for all. 
If he used skytale then I wonder how he did it. Did he wrap the entire plaintext around something? Or did he make it more managable in parts? I think the latter is interesting.
Mr lowe, you have some physical experience with making skytale ciphers, what do you think? Could you come up with some malconstruction hypotheses?
Just found some 2008 skytale discussion!
http://zodiackiller.fr.yuku.com/topic/2 … pOKNf_Smpo
I have a mini-theory about the photocopied image. If there is a clue (if it’s not just a Groucho image), then I think it lies in the hexagonal shape of the magnet. Z may have been hinting that the message contained in the 340-cipher is encrypted by transposition. Specifically, Z might have used the first ever military cryptographic device: the Spartan scytale. Here’s a link:
http://en.wikipedia.org/wiki/Scytale
Just another wild theory, I suppose.
not bad at all of a theory!I understand the number of message rows depends on the cylindrical shape of the cipher tool.
Since it would be limited to 6 and less rows with a hexagonal cipher device,there is still the possibility to use multiple skytale plaintext wraps!
In case of a 20 row plaintext and a hexagonal cipher tool for example you would use 3 hexagonal wraps plus 2 more rows on the fourth one.Depending on how small or narrow your letters are composed or how narrow the wrap is,you might need to duplicate or triplicate the effort to fit a 17 letter row or line e.g..Never thought of that.Interesting idea.
Jarlvie you will need to dumb it down for me .. Malconstruction.?. A word not in my vocabulary .. So I looked it up and I’m none the wiser. So here goes anyway..I have been pondering the construction process of such and my thoughts kept digressing to simplicity. It is a bloody big scytale if that’s what he has done. Could he have bothered.. The possibilities are many but here are a few..My thoughts wandered to an old builders/surveyors tape .. Flexible 100 feet+ long, (I actually own one of these) tape one end to a pipe of say 6.4 inches and wrap around and around. (That would end up being inches along the radians?) Then copy your plaintext across the lines using inch or even half inch as lines , unravel and copy out your plaintext which is now jumbled. Then turn it into a zodiac cipher with lots of symbols and multiple letter use. He may have used another process as well but if he did then it will make it tough. The problem I see with the above is the starting point of writing out the plaintext and the filler at the end as it will be dispersed through out the code which is important when breaking down a partial solve as you could end up with say..threel Ls in a row which is fine. You just have to dismiss some periods as nulls or junk.Two other products to use for a large scytale, a few dressmakers tapes or an old fashioned paper streamer, ruled up in inch or 1/2 inch blocks or maybe he just cut and pasted a sheet of graph paper. That’s an interesting observation the old post about the shape of the magnetic key ring and scytale. It does share the scytale flat sided staff you see in books. Other thoughts, a baseball bat with a conical taper down the handle would give a changing twist. All of this of course just pure conjecture, scytale does fit the alignment process if that is in fact the smoking gun, and it’s really simple and effective when in combination with other processes. Or he did something else like rail fence (zig zag) which I believe has been explored?
Sorry to bore you
Cheers all
Mr. Lowe, I think that Jarlve is saying that he wants you to think of ways that Zodiac could have screwed up a scytale transposition, on purpose or by accident, so that un-transposed it won’t solve.
One idea: to make a perfect scytale the spacing of the plaintext would have to be uniform throughout. If the spacing of the plaintext varied just a little bit there wouldn’t be skipped or added symbols but there would still be misalignments. If he taped pieced of paper together the taped joints could have caused spacing issues. I also like the baseball bat idea.
Ok smokie thanks.. Purposely making a scytale harder you could, when writing out your plain text lines On the scytale skip a row until you come back around ..basically odds and evens. That would mix it up more.. Also using two letters per space. Also starting with a row of junk and finishing with filler. Also maybe writing alternate lines out upside down gives you some letters like p as a d .. Writing it backwards..
Misalignment in the construction process as smokie pointed out Is the only way I can think of him mucking it up. It’s a pretty simple thing to do so I think if this is what he did it should be right.
Jarlvie. Making it in parts is very possible. A two or three part scytale cipher with different alignments .
Edit:Could the other z codes be scytale. Wrapped around a pencil. Still to short to break..
mI gniog ot tsuj ssot siht aedi tuo ereh, dna i ma erus uoy lla era erawa fo ti tub siht saw desu ni a eop rehpic
-krulg
It’s a good idea, thanks. I will work on a set of n-grams to accommodate.
Solves reasonably (Score: 22571) without adjusting n-grams when reversed. So I’m pretty sure something like that, or variations of it are not holding us back of getting higher scores on the 340 transposition hypothesis.
Score: 22571.45 Ioc: 0.06388509 Chi2: 21.68135 H: 3.962675 M: 0.244186 C: 86 S: 21 michercoeainusedwasthisbutitofawareareallyousurealiandhereoutideathistossjusttogoingil kclwfilnfycpoqfgzyqbwcqxobcbnuyzyifyifyaaenoqoifyscypgwfifnobcgfybwcqbnqqmoqbbnjncpjcs micher coe a in used was this but it of aware are all you sure al i and here out idea this toss just to going il
That’s great! I look forward to the results. Please come up with the solution because we need to finish off this massive thread once and for all.
Yeah it’s about time. There are other unsolved ciphers lurking out there.
I have been pondering the construction process of such and my thoughts kept digressing to simplicity. It is a bloody big scytale if that’s what he has done. Could he have bothered.. The possibilities are many but here are a few..
Simple things are often the best, and we must try them first. I really wonder if the Zodiac would have messed with a real physical skytale construction process at all and wonder what the book doranchak has ordered will reveal.
Hehe a, baseball bat cipher.
Mr lowe, if you have the time and feel up to it. You could make a skytale transposition of your own (that has something wrong with it).
Here’s a plaintext:
ilikekillingpeopl ebecauseitissomuc hfunitismorefunth ankillingwildgame intheforrestbecau semanisthemostdan gerousanimalofall tokillsomethinggi vesmethemostthril lingexperenceitis evenbetterthanget tingyourrocksoffw ithagirlthebestpa rtofitisthatwheni dieiwillbereborni nparadiceandallth eihavekilledwillb ecomemyslavesiwil lnotgiveyoumyname becauseyouwilltry
You could use this cipher generator to encode a plaintext or send me a PM and I’ll do it for you: http://zodiacrevisited.com/zodiac-kille … generator/
Ok Jarlvie This may take me some time to work out how to construct this. It’s such a simple thing to build its hard to make one wrong. Unless you want to put a twist in it.
Pondering..
Ok Jarlvie This may take me some time to work out how to construct this. It’s such a simple thing to build its hard to make one wrong. Unless you want to put a twist in it.
Pondering..
Yes, put a twist in it!
