| Subject: Re: Grid Software (Close...( |
Author:
Adam
|
[
Next Thread |
Previous Thread |
Next Message |
Previous Message
]
Date Posted: Mon, Sep 01 2003, 4:04:50
In reply to:
Steven
's message, "Re: Grid Software" on Tue, Aug 26 2003, 5:11:43
O.K., so all I have to do now is finish my documentation and then link all the appropriate help buttons to the corresponding section.
A note on the "Grid" decoding scheme. My implementation of a grid scheme allows for a full range of analysis; this is actually more of a "problem" than a feature I'm proud of. I'll use the "Treasure" formula as an example.
Treasure magazine allows for a total of about 200 additional numbers (counting all the rearrangements and +-*/3's) that can be added to each column independently.
Let's look at the math for a second, shall we.
A 3x3 grid has three columns, meaning one of the 200 numbers can be chosen for each column. Let's assume for a second we had a cipher consisting of only 9 numbers, thus only one grid. That 9 digit cipher has 200 combinations the computer needs to check for the left column, 200 for the middle column and 200 for the right one. 8,000,000 combinations. (Actually there are 237 numbers, so it is more like 13,000,000.)
For each one of those translations the computer needs to cross reference the KeyText, retrieve the appropriate value, perform a translation and store it to write to a file. I can (without any serious optimization, which I just ain't gonna do) perform about 200 or so translations a second. This means, on a standard desktop computer, to run a full "Treasure" translation on _just_ 9 cipher entries will take _18 HOURS_! I'm not even going to bother calculating just 18 numbers, because the calculation time increases EXPONENTIALLY with each grid we add.
Under normal circumstances I would, as the developer, limit the range of options available to the end user in order to make sure nobody mistakingly ends up with a 6 year translation project on their hands. BUT, given that this is designed for people who (hopefully) understand a thing or two about ciphers and brute force-attacks, I'm going to leave everything as is (with an estimated time warning). The grid decoding scheme needs to be used carefully, once I finally release it. Do a small range of integers (10 or so) on a single section of a Beale Cipher, check the results, then try another ten.
Well, we'll see anyway. I just wanted to give ample warning for anybody who thinks it'll be easy to prove/disprove a grid theory.
The Treasure theory was designed in such a way, I'm afraid, that the author surely knew his theory couldn't be disproven...unless somebody here would like to help me get some computer time at the NSA. Even then, if you look at the total cipher, it could be considered darn near computationally-unfeasable when you also examine the fact that someone (more likely something) needs to check and see if the translation makes any sense...
[
Next Thread |
Previous Thread |
Next Message |
Previous Message
]
| |
Feedback: 