Any comments? .

back to main +ORC page

Copyright: This document is copyright © 1997 David Nicholls. It may be freely linked to, and reproduced electronically or in print provided it is credited to the original author and includes the above email address and this copyright notice.

This page looks awful in Mosaic 2 or other early browsers, OK in Netscape 1.2 and MSIE 2 (transparent GIF and column colour problems), but best in Netscape 3+.

Is it a Substitution Cipher?

I was just now idly solving a cipher problem in the computer pages of one of the local newspapers (The Age : perhaps not the Neue Zürcher Zeitung or the Economist, but a sober and sensible paper). The puzzle was an enciphered quotation. The only clues were the word structure and the knowledge that each individual letter in the cipher corresponded to a single coded letter. Rudimentary code-breaking stuff, just made challenging enough by its length.

When I'd finished, I wondered whether the +ORC Riddle might not also be a substitution cipher. In other words, is he asking us to take numerical information from the text of the Riddle and simply substitute it for numbers in the URL? The process in decoding the cipher problem involved intuitively identifying patterns, trying them out, and settling on internally consistent solutions. Would the same process work here?

There are many ways that you might identify numbers from the Riddle (see "patterns") and I want to keep things as simple as possible, so I'm proposing to keep identification of the "numbers" and the consequent substitutions as simple as possible. I also propose that the appropriateness of the substitution might be indicated by the structure of the Riddle itself. In other words, if the Riddle appears to indicate a "6" and a "1" in the first line, and since the first octet of the URL includes a "1", then we should substitute the "1" in the URL with the number "6".

The first question to answer is what are the numbers that show up in the Riddle? The obvious ones are the numbers of bars. Other possibilities are the numbers of users (see the original essay), and the ranked values of gold, silver and steel. Both of these have the virtue of being available on every line of the Riddle.

If you generate the numbers from the bars, the metal ranks and the user numbers from each line, you finish up with this:

    Line 1:  6,   1,   1
    Line 2:  5,   2,   2
    Line 3:  4,   2,   3
    Line 4:  0,   3,   2
    Line 5: null, 3,   2

What can you do with these? Well, the first line suggests substituting numbers "1" and "6". There are two "1" digits in the first octet, 131. Since we can't exceed 220 odd, this suggests changing 131 to 136.

Following this line of reasoning to the second line, we have a possible substitution of "2" by "5". Since the second octet is 92, this suggests changing it to 95.

Coming to the third line, the process is not so obvious. The third URL octet is "15", and the only numbers we have from the third line of the Riddle are "2" or "3" to be substituted by "4".

Again, the fourth line has the same problem. Now we come to what I can only call a fudge. Adding the numbers from the third and fourth lines gives us "2+3" or "3+2", depending which column you choose. And there is a "5" in the third octet. Perhaps this means replacing the "5" in "15" by "4+0", to give an octet value of 14. A bit feeble, but it hasn't fallen over totally.

The final line suggests replacing a "2" or a "3" by a "null". There's no "3" in the final octet but there is a "2" So might this indicate replacing the "2" in 128 with nothing, that is, a final octet of 18?

This gives us a resulting URL of . It is getting very boring to tell you that this also doesn't work. I'm not surprised that this is so, as the method used to force an answer was rather ugly. It also says nothing about the directory suffix.

But it serves to illustrate the idea, and I'm hoping it might inspire someone out there.

'Til the next inspiration strikes.

David Nicholls
14 June 1997