unforum

[Sh1ft]

I think the following (see quote from arnezami) is a great idea and so I went ahead and modified the sol.pl found on Schneier's site to allow me to try different deck orders. No less than one perl wrapper and one shell wrapper scripts were created to allow me to try keywords.

Nothing promising yet, there is this, passphrase "star":
MATQPJTMHPWVKOWASZWYESYEORMITJWIKPDSBSDYBRMTLJPSMJWOUZJGOWCLFSHGWLVKOVBNUZKMXECVCVQDYJWWKKDHVMEJKMGEIVARVON


Doesn't make sense I know, there has been only one other interesting decode, but only interesting in that contains the word worm, passphrase "hacker":
EQDCDNRUBPUGFXCDQAUFEGPYDVOWHSMQCOZRFKDMWORMPKGEYIUKVXVDJAZGEISFOTOCDVFAMGTJSBHHETJVEYTMDCNATBADEIHYKBGPPUL

PM me if you want the collection of scripts, keep in mind they were created in a hurry and I am not a programmer, also they are not very efficient. (Also contact me in pxc irc, username Shaftoe.)

[Dranioth]

[themandotcom]

[ramsfan]

count ME in

[CherokeeParkes]

Here's one for Ultimate Frisbee fans.

EQDCDNRUBPUGFXCDQAUFEGPYDVOWHSMQCOZRFKDMWORMPKGEYIUKVXVDJAZGEISFOTOCDVFAMGTJSBHHETJVEYTMDCNATBADEIHYKBGPPUL

Though i'm not sure if there's anyting to this.

[Bendover]

Does anyone know of another site or a program that is non programmer friendly for checking pass phrases. Bruce Schneier's Solitaire V1.2 Decrypter seems to be on the fritz or gone.

[Joe_the_OOF]

solitaire site

[AndyHall]

I'm trying to bend my brain around creating a php script to do this, so far if you have a keystream and cypher text it'll work it out, I'm currently working on creating the deck from a password.

http://www.ukinetmedia.com/perplexcity/

Andy.

[kontan]

Program for Palm

[justdig]

[arnezami]

Hi.

I've been busy trying to think of a new approach. No solution yet but just wanted to vent my latest idea. Maybe somebody else can do something with it.

What has been bugging me for a long time is the 3 "loose" characters (instead of 5). I'm beginning to think it almost has to be significant. I find it hard to believe the creators didn't stick to protocol by not making the length of the ciphertext a multitude of 5.

But if so the 3 characters most likely do not belong to the ciphertext. This either means the 2x10 characters are the ciphertext (but this is quite short for a message) OR only part of the 108 is the ciphertext. In other words: only the first n x 5 characters of the 108 letters belong to the ciphertext. This would mean we would have to remove n x 5 + 3 characters at the end of it (and use it for something else).

As for how large n would be. After looking at many numbers/lengths on the card its clear there IS a text that is n x 5 + 3 characters long. Its the "If entropy wins..." text. Its 43 characters long (= 8 x 5 + 3).

So my idea is the following: if you remove last 43 characters of the 108 characters then the resulting 65 characters would be the ciphertext. The last 43 characters should then probably somehow be used to shuffle the deck. The question is how of course.

What i've tried is adding (or substracting) each letter in the "if entropy win" text to/from the last 43 characters of the 108 characters (leaving 65 characters for the ciphertext) and use the (possibly meaningful) result as a keyphrase. I haven't tried all possibilities though. It may require including the 2 x 10 characters too (leaving up to 85 characters for the ciphertext). There is still a lot to be tried out.

I'm still in the experimental phase of this approach and I don't have much time atm. But maybe somebody else can pick this up or use it somehow.

Nice things about this approach:

• It would explain the need for the 3 "loose" characters.
  
• If the right method is used it may result in a sensible 43 character sentence which would clearly indicate we would be on the right track.
  
• The 65 leftover characters would be long enough to contain a sentence and would be n x 5 long (according to protocol).
  
• "If entropy wins..." could mean: first the deck has to be shuffled. Then the outer 10 characters should be removed ("outward looks should leave you cold"). So it kinda fits.
  
• The "if entropy wins..." sentence is used as cryptic hint (including the fact that its 43 characters long) aswell as a requirement to find the right keyphrase. Which is kinda cool.

Just my 2 cents...

arnezami

PS. I've also been experimenting with entropy encoding but without much success. This highly depends on having exactly the same letter frequencies list used for encoding in order to decode (which we don't have).

[AndyHall]

[BBuck]

[arnezami]

Thanks BBuck for that info on the 43 characters adding/substraction etc.

I'm all for a systematic approach. So we can keep track of what we have all tried. Since there are SO many possibilities it would be a shame if we were to do thing double...

And I think its a really good idea to look at every little detail on the card (didn't somebody give us the hint to examine the card more closely?).

We could definetely use a high resolution scan of the entire card. So everybody could take a good look at it.

arnezami

[gurupidal]

I've been following this thread for a while now. I've uploaded the solitaire.php file on my site.
www.alpop.co.uk/solitaire.php
I have noticed that on this version on the software, it makes a difference if it is spaced as 5's or compacted to one long line of letters.
I've been trying to randomly insert passwords, but to no luck, but I seem to think that there is more clues in the card.

This will probably repeat someone.
eg. "if entropy wins, outward looks should leave you cold"
Temperature clues lead to the thermal printing.
Entropy (something to do with randomness (the cipher), but also thermal dynamics).
But why is the line at the Five of Cups bar one night openly spaced. Other lines on this card do not spread the full length
FInally this is probably nothing, but an anagram maker on the web turned Djinn Worm into Dr Jinn Mow.