unforum

[pammy-d]

Hi
has anyone noticed that FIVE COWS uses 8 letters could this be the key

[Langley Moor]

[dusty2229]

Re: Another lead

[ramsfan]

Tried to make rccrypt work, but it doesn't for me. Seems like I've got a few things missing, some of which should be stored in my brain. It's the RNA - just doesn't keep anymore. Anyway the point of this is to suggest another 8 letters which someone's undoubtedly tried already: "biryukov". If it can be worked out on a piece of paper (see above) this all seems a bit too obscure.

[neuromancer]

Hey dusty, where did you get that page from?


I've tried 'biryukov' but it doesn't yield anything I'm afraid.


EDIT:

After some more reading-around on distributed.net's bovine project, I found out that they were given the first few decoded characters, so they had something to go by.

We don't have that. Also, the bovine project, at it's peak had an equivalent processing power to approximately 46,000 2GHZ Athlon XP's working in parallel, and it still took them 4 years to break RC5-64!! Working out the calculations, it should have taken them a lot longer to crack RC5-64, so they we're lucky.

We're gonna need some major clues as to what the keyword might be because I highly doubt we can brute-force it now....unless somebody has a quantum computer??

It might be an idea to interrogate Kurt for clues, if we get desperate.

[dusty2229]

[specterz]

Seeing that Kurt is the creator of this puzzle, it would make sense if the 8-letter word was something to do with him. Just some suggestions for you to try here, then...
Allister (As in McAllister)
VonLubik
HMansoor (Senior Fellow Cryptology)
Sanguine (Algorithm created by Mansoor)
PathTime (after his blog)

I dunno... maybe something will work. Hard card, this!

[neuromancer]

[chichiri]

Has anyone tried the most elusive keyword of all time: PASSWORD

It's been used before so it's worth a try

also how about SCARLETT

maybe SKITEWAY or VKITEWAY

or on an equally puzzling note SHUFFLED

or on the lines of inventing new words PERPLEXO

none of them likely but ya never know.

no da

[c1023]

WARNING Stop using the new features of my rccrypt build (rrcrypt_1.4-pcx_c1023.zip) as I've found a major bug! The correct keys are not being tested.

I can't rebuild until I get home this evening, but if anyone wants to rebuild, and post it here the error is:

rccrypt.c line 848 should change from:
snprintf(keyString, 17, "%016llX", currentKey);
to:
snprintf(keyString, 17, "%016llX", nKey);

Sorry guys.

[ramsfan]

Looks like you cryptos are gonna spend all your time trying all those 8 letter keys in the corrected program. I got a very nice reply from the guy behind rccrypt. I wrote apologetically, assuming he'd be plagued with enquiries since the clue, but I was the first to write. He hadn't heard of the game. He made some suggestions which seemed new to me, but may be not to those who can understand all that has been said before. I've asked for his permission to paste these ideas. Also he asked for a link so he can look at the card, so may have some more ideas.
My friend who doesn't look at the forums came up with the following observations. People have talked about lines going through the cards before, but going through this trail I couldn't find such a comment for this one. The card has a thin line going right through. It divides the code vertically into two sections. It runs just to the right of the first digits in the 7th column (the first short column). On the reverse the line is the edge of "Marginal Lane". Could that help with which part of the code to use for key and message, or might it indicate the importance of that single column of numbers, ie the left half of the 7th column? Also he wondered if 13th labour could mean anything else. He wondered about Harold Wilson as he was elected in 1964. I've checked and he was the 13th Labour leader, though not the 13th different leader, which would be Kinnock (one person was elected leader 3 times, and Macdonald twice). I've checked for anything significant relating to Harold Wilson on either Dec 8th or August 12th 1964 but found nothing.

[meep]

This has probably already been thought of as you guys are way ahead of me in this, but I found that there is is a machine codenamed PURPLE by the United States, it was a diplomatic cryptographic machine used by the Japanese Foreign Office during, and just before, World War II.

Could a 8 letter key be something to do with this, maybe JAPANESE?

[neuromancer]

Yeah, I was thinking along the same lines with the 13th Labour leader, but it doesn't seem to yield anything.

Good call on asking the RC-Crypt dude, didn't think of that!


w00t!! 100'th post!!!! does crazy dance

[c1023]

I've fixed the previous bug and added another feature (no I don't mean bug ) And this time I've tested it properly

It tests all different case combinations of a single text key. It also tries it backwards as well as forward, and padded with spaces and nulls to the left and right.

In total up to 1024 combinations are tested for the entered key word.

E.g.
rccrypt perplex
tries all of the following, and many more:
"perplex ", "Perplex ", "PeRpleX ", "PERPLEX "
" perplex", " Perplex", " PeRpleX", " PERPLEX"
"xelprep ", "xelprep ", "XelpReP ", "XELPREP "
" xelprep", " xelprep", " XelpReP", " XELPREP"

Hopefully it will help make up the time people lost with the previous version.

The download includes the .exe, cygwin.dll, and modified source file, so it should be everything you need to run it under windows.

[ramsfan]

These are Richard from rccrypt's comments. Hope someone clever can use them. Sorry for misleading you about the importance of the marginal line separating or running along the numbers. It was very obvious in my friend's card, but not present in mine, so presumably a printing artefact.

It's possible that my code might work - the RC5 was 32/12/8, but used long keys - mine uses long long, so it may actually be 64/12/8 - I've always thought of mine as 128/X/X, but it might be that I'm mistaken.

The people at disibuted.net will almost certainly be able to look at my code and tell you.

If it is, then it's quite easy to use my stuff in a script to search though all the keyspace - you'd just need a way to split up the keys to search.

If this is the case, it would make sense as to why they sent you the "rccrypt" clue.

First off, I'm not at all offended or annoyed by your request - I put the form up so people can contact me

You are the first person who has contacted me about this.

I had a request some time ago about using my RC-Crypt software for encoding/decoding a "clue" in a game. I wasn't able to give the people what they wanted, and assumed they had found another way. I don't know if they were connected with Perplexcity or not.

They had asked about encoding a clue that would significantly help in a roleplaying game if people could figure it out. They expected that a group of people with computers would be able to do it, possibly using my code or code from the RC5 group at www.distributed.net.

They wanted my code to be rewritten to use a different number of blocks, but I didn't have the time to do it for them.

My guess (I know nothing about the game, and have only heard about it from you) is that they have used my RC-Crypt program to encode something.

There are 176 characters, which is an integer number (11) of 16 bytes - the block size my code does. If they did use rccrypt, then the first 16 bytes are random data to make decoding harder, then the rest are the real message, with possibly some padding added at the end.

I would guess the 64/12/8 are the parameters, though they are not entirely obvious.
I would have thought 64 was the block size (in bits), but my code does 128 bits art a time.
The 12/8 are probably the size of the key in bytes, and the number of rounds used. Which is which? No idea!

To decode this, you'd need to run through every possible 12/8 byte key and check each result for a valid output - probably in English, but depends on the game.

I suspect that is not what they expect you to do, as an 8 byte key has 64 bits, and it would take a LONG time to find the solution (see www.distributed.net). I would guess there may be another clue giving some of the key to shorten the time period.