unforum

[hturan]

I think this 5 cows processor has been mentioned before, but I'm not sure weather 'longhorn' has been tried as the key .

[Ashin]

First off: to the last 3 of you!

I have a feeling when we do find out what the keyphrase is, we're all going to slapping our foreheads. This is, however, assuming that the 13th Labour part is actually some sort of keyphrase hint.

It's been mentioned before, but there are 12 RSA challenges http://www.rsasecurity.com/rsalabs/node.asp?id=2103, and this would be the 13th one. All of which revolve around RC5. Most likely the card title and the 64/12/8 are both hints to point us to RC5-64 and nothing else.

Right now this card is only going to be solved by time (like the rest of the RSA challenges), so keep your computers running people. When you go to sleep at night, kick off the programs from previous pages and let it run. If everyone chips in we might have a solution in 5 or 10 years

[hamatoyoshi]

I haven't been tracking the progress of Perplex City for about a year and decided to drop in. This thread kind of caught my interest.

I've read and searched, and I'm by no means qualified to discuss the topic of encryption, but here's an idea I don't think was thrown out after I read the thread and the Wikipedia entry on RC5 for a little background.

I couldn't help but notice this in the article:

[Jakeo]

[quote="hamatoyoshi"]
So my question is, is it easier to mount an attack using 2^44 known plaintexts or 2^64 keys? And if 2^44 known plaintexts is easier, could we do it, and if so, how fast?
[quote]

The attack is based on the assumption that if you have 2^44 plaintexts and you can see how they output, then you can work out the key that was used to seed the output stream.

So in short, it isn't possible to mount a chosen plaintext attack on this cipher. Sorry.

[fretty]

Sorry if mensioned before but there appears to be a little black dot at the end of the third line up from the bottom. It appears on both the card catalogue and the perplex city site picture, don't know if it's a coincidence or not. Could this contain any info?
Also has longhorn been tried as a keyword?

[groovygirl20]

again i am stepping in (groovygirl20)

the little black dot and or lines on the cards are like printing errors and not necessarily on all the cards there! i dont have my cards with me now as i am at school ... but shal have a look later when i am home

[hamatoyoshi]

[fretty]

Has anyone recieved a Von's clue for this?

[CherokeeParkes]

Back near the start of this thread there was a post from Oliver.
His post his on page 7.

The clue is rccrypt.

As perviously mentioned in other posts rccrypt is a rc-5 decrypting program. It also appears to only be for linux or windows with a c compiler. Of course all this has been googled, discussed and disected at length.

[hairysocks]

Do we know yet if the decrypted data is readable text? It could be an image file, such as a BMP. Given the size of the data its unlikely, but it may be one of those square barcodes, such as the one on the Barcode Recipe card, which may take you to a site with a question.

Another thing - I've noticed people suggesting four letter words and trying them in upper and lower case, but that may not be enough because of all the combinations of case a 8-letter word can have 256 possible variations. If you start including the possibility of number substituting letters the number of combinations rises even higher.

For example, if you allow 5 instead of S and 2 instead of Z and 1 instead of I then an 8-letter word with S Z and I will have 2x2x2x2x2x3x3x3 = 864 possible variations.

I have written some Java code that generates all combinations of letters and number-substitutions for 8-letter words, the number-substitutions I allow are:

'A' = '4'
'B' = '8'
'D' = '0'
'E' = '3'
'G' = '9'
'G' = '6'
'I' = '1'
'L' = '1'
'O' = '0'
'P' = '9'
'S' = '5'
'T' = '7'
'Z' = '2'

Please let me know if there are any other obvious ones I've missed out.

I also have a Java application that searches a file and reports on any sequence of 10 characters that are "printable" (i.e. may be the decrypted message. I'm thinking of modifying it to look for image files for the reason explained above.

These Java apps are wrapped up with some shell-scripting and run on my Linux boxes.

Is there anyone who would be interested in having this Java code and shell script?

[Jakeo]

[c1023]

In my build of rccrypt (bottom of page 12) I made it test combinations of the following:
Written forwards and backwards - "password" and "drowssap"
Padded to the left and the right - " passwd" and "passwd "
Padded with spaces and nulls - " passwd" and "\0\0passwd"
As well as all combinations of upper and lower case

You may want to add them as well.

[Rob_Riv]

[CherokeeParkes]

If is is a date then it is waaaaayyyyyy past.

64 defintley would be the year as for 12 and 8 highly interchangable. Also the style of the date yy/mm/dd

My question if it is a date which country uses it and is it a clue?

[unrealt87]

Ok guys, it's time for a quick outta-the-box think.

Can everyone who has this card verify that it has NO pixel markings along it's silver edge?

And if so, what if the cows are a bluff?
Wouldn't it be ever so convenient for the passcode to be a variant of :

nopixels?

Come on Ashin, get ready to slap your forehead, cos this HAS to be right

P.s. I Dare you to trout this