Code Discussion

Drogue · May 23, 2003, 08:13

This thread is for the discussion of codes, binary of otherwise.

So far we use Mani Alpha-3's "add four if the 3rd character is 0, add five if it is 1". Other options include Corelli Omega-9's "add a character after the fourth", although I am not sure how well the translates, could be able to get it on a long message. Another option could be to "add one if it is a 0, or two if it is a 1, after the fourth character".

Any others please suggest. I think we need to change, since others may have discovered the code and not told us. I don't put that pass Archaic et al.

Cedayon · May 23, 2003, 08:50

If you really want encryption, I can come up with a rather complex (ie essentially unbreakable without computers) code, and write a program that converts Ascii strings into our code and back, and distribute that program only to CyCon members (and Googlie, if he wishes).

Another method of deployment is on my cs account website (which isn't linked off the normal web to my knowledge), and put it behind a password prompt (just use our game password, unless you want another one).

The program could be modified to use a different code without affecting the users, particularly if it's web deployed... I could have it generate different encodings based on the day of the week, actually, though that would add complexity to my task.

Maniac · May 23, 2003, 08:53

Don't forget this proposal by Corelli Omega-9:

Quote:

Choose a number and convert it to binary through straight number conversions, not the binary converter. Make it a number less than or equal to 7, and keep adding extra zeros to the beginning until it's three digits long. This means the first three digits of each message will be the binary equivilent of a decimal number between 1 and 7, and the binary converters won't work on it. Then add x random digits after the number, where x is the number you chose. Finally, add a random digit after the xth digit of the actual translation.

In order to decrypt, just figure out what number the first three digits represent. Then, figure out the number, delete it, and delete the first x digits of the message after that, where x was the number you just decoded. After that, delete the digit after the xth digit in the message. This should thwart the people who delete numbers progressively.

Here's an example... Let's choose the message akizeta. Choose a number between 1 and 7, like 5. In binary, this comes out as 101. akizeta in binary is 01100001011010110110100101111010011001010111010001
100001
So, add 101 plus five random digits after and it becomes
10110100011000010110101101101001011110100110010101
11010001100001.
Then, add another random digit after the fifth digit of akizeta normally, and you get

10110100011001001011010110110100101111010011001010
111010001100001

This comes out as complete garbage in any translator, and randomly deleting digits won't work. They won't get any sense out of it, and because the inserted random digit is less than 8, they won't get even a single proper character.

As he proposed it, it might be too long for the average lazy cyborg like me, but perhaps we can shorten it to
"Three first bits are a number "x", and you have to delete digit "x" in the real message" and skip the "x" extra random bits.

We could also use different codes, and the last two bits of the message could tell us which code it is. For example, when it's my code, add 00 at the end. If it's Corelli Omega-9's code, add 01, etcetera. That might of course become too complex and tiresome again.

As for the suggestion made of using another code every day of the week (I at least think that was the proposal - I can't find the original post back), that would become too annoying with our different time zones, with posters posting before, and others reading past midnight etcetera.

Cedayon · May 23, 2003, 11:41

Doing encryption by hand is a pain, as you've noticed, so any such scheme is balanced between ease of decoding with the "key" and difficulty to decode without the key, almost always very heavily towards ease of decoding with. That's why I proposed a program, because a computer doesn't mind performing incredibly tedious encoding or decoding algorithms

I've got an hour or so before I have to go to the lab, I'll see if I can put together a web page for this encryption.

Edit by Maniac: Great idea!

(I edited your post for this short message. That way we can't be accused of spamming in our private forum.

)

Cedayon · May 23, 2003, 13:40

Well, it took 2 hours instead of 1, but I put together an encryption page that implements (I think) Corelli Omega-9's proposal. You can find it at:

http://www.cs.uga.edu/~lamothe/encryptor.html

or

http://peck.cs.uga.edu/~lamothe/encryptor.html

let me know if either/both don't work, as they're on university machines with a possibly large amount of protection/firewalling.

Naturally I would use more involved encryption, in fact I tried, but that's work for another time

edit- btw, it's not perfect, I've seen glitches, but I think it works pretty well.

another edit - I forgot to mention: right now it only works in IE, not netscape, but I'll probably fix that soon

Drogue · May 23, 2003, 14:33

HongHu has broken our code. I suggest we use Cedayon's coded translator ASAP. I have posted a reply to him in the code with 4 on the top (as per usual) and for on the bottom.

Until Cedayon has got his translator working with something heavily coded, I suggest we use Function Corelli's code.

Maniac · May 23, 2003, 17:14

Quote:

HongHu has broken our code.

Perhaps I'm a bit slow, but what makes you think that?

I haven't seen him post any decent answer on our binary.

Corellion · May 23, 2003, 17:19

Cedayon's translator works like a charm! Although I am curious as to the exact encryption you used. The first three letters are a binary number, and the next x digits (where x is the binary number) are deleted, but there's an extra bit of encryption in there. A single extra digit is added, but I don't know where, or what it is. Can you please explain oh great encoder?

BTW, this is perfect... not only is the encoding difficult, but it gives a different number ever time! Let's see them crack that!

Edit: Is there any way I can engage word wrap for the decoder?

Maniac · May 23, 2003, 17:25

I've done a bit of testing (with 0 = 00110000 and 1 = 00110001 and see what's the difference when encoded), and to me it seems like Corelli's suggestion, but without the extra nonsense bit on the 'x'th spot of the real code. So I can't discover any extra encription.

Corellion · May 23, 2003, 17:47

Hmm... I've discovered a potential problem.

Since they have made a rough guess at our previous code, they could try the same methods on this code. If they delete digits from the beginning of the message for long enough, they should get the first two or three characters of the message right. These will form a coherent set of letters. They can then determine how many letters are right, multiply by 8, and delete the next character after this numbered character. This will get them the proper message. I propose everybody adds three nonsense characters, followed by a space, the the beginning of every message. This should foil their attempts.

For example, instead of typing:
How are you
type:
%y9 How are you

That shoud do it.

Maniac · May 23, 2003, 17:57

Corelli Omega-9, by using your encryption as I posted in this thread (so not the one Cedayon has used for his en/decoder) I got the following message out of the binary you posted in the general forum:

Quote:

PAread Googlies message. However, lets see them try to crack our new code.

Now we are (almost) uncrackable!!! Bwahahahaha

I saw the first three digits were three, so I deleted those three, the next three and the third bit after that. Did I do something wrong to get "PAread"?

Quote:

If they delete digits from the beginning of the message for long enough, they should get the first two or three characters of the message right. These will form a coherent set of letters. They can then determine how many letters are right, multiply by 8, and delete the next character after this numbered character.

Am I the only one who doesn't understand this?

Cedayon · May 23, 2003, 18:03

Re: word wrap, I know it just keeps going and doesn't wrap, which might be a problem if your browser doesn't like really "wide" pages... but as far as selecting it, just double click the line, that should select it, and copy it (ctrl-c on windows).

Re: that extra digit, it's in there as per:
"Finally, add a random digit after the xth digit of the actual translation."

I'm not sure if it's added after the xth digit, or before, but the encrypter and decrypter do both the same, I think.

Re: adding 3 nonsense characters at the beginning, I can have the translator do that

If you guys give me what kind of code you want, I can probably make the translator do it, but I'll let you know if it's too hard... I tried inserting random digits after every x digits, but I couldn't get the decoder to pull the right ones out.

btw, if I change the coder/decoder, I'll leave the older version(s) available via links, so we're not stuck with messages we can't read

Method · May 23, 2003, 18:08

cedayon, if your translator only works in IE, it's clearly the devil. surely googliegod would not approve of the use of such a horrific browser

Cedayon · May 23, 2003, 18:26

I've figured out what I need to do to get it to work with netscape, but when I change it I get errors... I'll keep at it, this shouldn't take long.

Googlie · May 23, 2003, 18:27

Quote:

Originally posted by Drogue

HongHu has broken our code.

Yeah - they managed to decipher my message saying I was incorrectly using the 'day of the week" protocol

'Cos of impartiality, I (obviously) couldn't warn you

I won't post in code - just read it to ensure no hanky panky is going on (similarly with other factions' codes)

G.

Cedayon · May 23, 2003, 18:58

hmm, javascript is apparently not as cross-browser as I thought... still working...

Method · May 23, 2003, 22:01

Quote:

Originally posted by Cedayon
get it to work with netscape

opera

Cedayon · May 23, 2003, 23:46

The translator now works in both IE and Netscape 6 (Netscape 4 doesn't support enough stuff to do what I want)...

Opera? If it works, great, if not... well, I can probably figure it out, if you really need me to (can't install netscape, or does the thought of doing so cause convulsions?)

Cedayon · May 24, 2003, 01:51

I made a modified version of the translator that uses a much more involved encryption scheme (I really, really doubt a casual code-breaker could get through it), it can be found at these locations:

http://www.cs.uga.edu/~lamothe/encryptor2.html

http://peck.cs.uga.edu/~lamothe/encryptor2.html

It works rather well in IE (wordwrap and everything, just don't use line breaks or put in anything very long).

It works in Netscape, but it's apparently impossible to get word-break wordwrap in Netscape, or at least I couldn't find a way of doing it in 30-45 minutes of searching a bunch of references and stuff.

I don't know what it does in Opera

I'll tell you all the encryption scheme sometime later, first I want to see if any of you can figure it out

edit -

10011110001110110100000100001110000001000110001011 00100010110100010001101011111000101101000100000100 00101010001010010001100010101000000011101010001110 00101000101000110101000111001000110010001001000101 010100110111011100111010101

Method · May 24, 2003, 09:39

excellent. 10101101100111111100100001110101000101111010101110 00110011000010000100000010001100010110010001101000 01000011010111110001011010001000001000010010001000 10000100001010100000100001001000101100010100010100 01101010001110010001101010001010001010000010000101 01001101001001110001011100001000000011011110101001 11001101111111100100000000110100110101010100111101 10010110110011011010011010110010000000011110011010 01001101010101001101011101011111001101001001100001 00001001100101011100101111000110001010011100111001 00110100111001101111010100110110110000000100001011 10111111110011110110011000010000001110111011110010 10001001011001001100001011100000001101001101010101 01111111100100000000110001101111010100111001101111 11110011001110011110011010101010010000110010000010 00001001101010011001010011011110011001001111001000 00000111100100111100100110010100110100100110011101 11000010000101011010000111010101010010110011010010 01101010101001101011101010101001011011001110101011 10000100001001100100010100101000101110000100000111 10000000111010110110011101010011110010011001110010 101010011001110011001011010100110010011000100
opera seems to be the same as netscape for getting pages to work i guess. the no word wrap is a pain though

otherwise great

EDIT: oh, and i loaded it up in IE just to see - it doesn't even work in IE now

Maniac · May 24, 2003, 09:44

What coder are you using? I can decipher Cedayon's last message "Release the gorilla" with both his coders, but your binary I can't decipher with anything. (I'm using IE)

Method · May 24, 2003, 09:59

hmph, that's odd, i can't code it back either. let me try again

EDIT: 11110010001000010011100001110010000111000100101011 10011111110010111110011000010000001010011101110011 11111001011111001100000001100101001110111001111111 0011111110011010111011101000101110100010000111

Cedayon · May 24, 2003, 12:23

Btw, the first coder wasn't supposed to be able to decipher that, I had forgotten to change something back (the first coder was using the second algorithm)... I fixed it now, so the first coder won't work for binaries generated by the second coder (it's not supposed to).

TKG, did you put a line break in the input for your last encoded transmission? All I get is ooga ooga ooga and a bunch of stuff that's usually the result of a line break being in there.

I don't know what to do about the line break thing, I don't know why it's not parsing correctly.

Method · May 24, 2003, 14:05

the second one is just ooga ooga ooga

can't remember the first one

Cedayon · May 24, 2003, 14:21

Ahh, I see...

11011110101111001011111001010101001000011101000101 00110100110101010100110000110101111100100000100000 01010010000111010101010010101110111000010000101010 01110011101001111001001101001100011101001111001011 11100101100100100000100000011100001000100111001100 10101110011100010011101010011110010011101110111111 11001101111001011101001011001001000101001100001000 00010100111011100111111100111111100100000000111010 1001100011001011111001100111001111000110110010

btw, has anyone figured out what I'm doing to encrypt it?

Method · May 24, 2003, 14:46

11111111100110111000000011100010010001110010001110 01001101001101001101011111001010111001100000001101 01100101111100110000000111001100101100110100010100 11101010011110010011101010011000110111000010000010 01001110001011101101001110111011100000001100011001 01000100101001100100100100111000100100000100001011 10011110110011011010011111110011110110010000000011 00111000001100011110111000010001001011011001010101 00111110100111111000100000000111010110100010100111 00010011101010011001110011000010000001110000100010 11100101001101001001100011101000000001101001101010 01100101011100111111100100001101111100101111001010 100011010101011

Cedayon · May 24, 2003, 15:35

10010110101101001111001000100000000010100010100011 10001010100000010000000011100110010010011010100110 01001001001100011101000001000001011001011111001100 00100000001110101010101111111100100000100001000101 11101010011110110010000010000000110010011110010100 01001000000001110101101000011011111111001100011101 01010100100100110100001101111001100101000101110000 00011011010111100010111011010011001110111010011010 10101001110011011110101001001111001010101010100001 00000111100101010100100000000110011101111000000101 10100011110110001110100011100000010011110000110110 00101111000000101000111010001100011000100000001110 01100100010100101000101110000000111101100101101100 11101010010011010111111110011001110010000011010101 01001100011001000001000000101001000111011110101001 01101100111010100100101100100000000111100000111100 11010111010011001010011100110010100010111000000011 10101001100011001111111001101001101000001101000001 00000010100110001110101010100110101101010000000110 00110010110110011100010010000010000100010010100010 11100111001010011001000010001000000010110010111100 0001101

Drogue · May 25, 2003, 13:07

Congrats Cedayon, this get's the Prime seal of approval

Very good work

However, we have no need to use code in our forum, we have the Gorilla to protect the secrecy of this place 00010110001000110000100000100100011101001001101000 10100111000101110011101110000000111010000100000100 00100010011101010010100010011000010000100010111100 01001001001001110111011100001000011000001111111001 11101100110000100000100100111111000100011001001110 0110100010001110001111

Cedayon · May 25, 2003, 14:40

I'm glad I could help

2 caveats of this system:
1. The encrypt/decryptors are hosted on machines not under my direct control (my dorm computer is behind a ferocious firewall, else I would have the option of hosting it here), and both have occaisional downtime (and my peck account will be wiped by June 6th or so) ... the ajax one (www, rather than peck) will be available most of the time though.

2. If our enemies find the encrypt/decrypt pages, that's that

... they're not linked to from anything, so that probably won't happen unless someone goes nosing around in my directories (they have to be world-readable for web use) or someone rats out the URLs.

edit-
01000100111010011000100010010100111000010000101010 01101111001001101001110011001100001000011111001000 01100100000100001100100111100100111011000111001100 10110111010000010000000110010011110010001010011000 00001110001001111011001110011101011111001101001101 00010100110000100001100110100010100010111100000111 00010110100011100000010011110000011100010110100011 10000001001111000000011000100001000000101001010111 00111001110101111100110111100101010100110000000110 00011010100110010100110010111010010000000011100010 11111001101010011101001011001100000001111000001100 10100111100100111100100110110100110001101111111100 11110100010000010000100110010010010011100010111000 00001110001011100011001110101011111111001010001001 10000100001000101111111100110111100100000100001111 100111000100010001011101110110110001

Drogue · May 25, 2003, 15:03

Quote:

Originally posted by Cedayon
2. If our enemies find the encrypt/decrypt pages, that's that

... they're not linked to from anything, so that probably won't happen unless someone goes nosing around in my directories (they have to be world-readable for web use) or someone rats out the URLs.

If that happens, I suggest you delete the site, change the url slightly, and rehost it, if that's possible. That'd be a pretty remote chance of it happening. I think it's more likely they either write a program to crack it or use the rule of only writing in Finnish and English against us (does Binary count as another language, or is it English because that's what we write from?)