Project Logs

Collapse

• Who needs pad addresses? Pt2

  gannon • 10/07/2014 at 21:10 • 0 comments

  I've gone through and restructured the code to be more efficient and cleaner. This was mainly to support bruteforcing the ciphers.

  The brute forcing page is currently separate from the main page and can be found at http://otp.portablesofdoom.org/brute.html

  I have only tested it with a few messages, but it works. It takes a little bit of time to respond though, so be warned.
  

• Who needs pad addresses?

  gannon • 10/07/2014 at 19:06 • 0 comments

  While trying to figure out the cipher posted in http://hackaday.com/2014/10/06/hackaday-10th-anniversary-1o57-and-the-art-of-encryption/ I decided that maybe either the pad address was wrong or it was double enciphered with a second pass somehow. While I was working on figuring this out Brian came forth mentioning he had actually just made a typo in the address.

  My method of trying to decipher the message was simple. Brute force it! There are only 18400 pad starting addresses after all. After working out some bugs in my JavaScript code I was able to get it to work...although it would return hundreds of results as I only checked to make sure the first word in the sentence was a real word.

  When I have some time I will work on cleaning this code up and getting it into a better working state. My proposed method for brute forcing the message deciphering is as follows.

  1. Split the message up into words

  2. Decipher the first ciphered word starting with each of the 18400 pad addresses

  3. For each result that is a known word continue deciphering the message

  4. Sort messages based on ratio of known words to unknown words

  Is this method slow? Currently in JavaScript, a bit. Also I seem to be hitting an issue where I can't process all of the data at once. I am unsure if this is an issue with memory consumption or something else yet. That is my main reason for not simply deciphering the entire message 18400 times. I will probably break the problem up into a more efficient method later.

• Oops...

  gannon • 09/15/2014 at 14:21 • 0 comments

  After testing the latest cipher that was posted I noticed a bug in the code!

  It was actually quite simple. My bounds check was off by one (a common issue with computing).

  Now, how long until a new bug is found?

• The next step

  gannon • 09/11/2014 at 20:13 • 0 comments

  I was all set and ready to wait for the pad I had ordered to come in so I could rip it apart, scan it, and hopefully get a more error free result from the OCR process. But that was when a little birdy dropped something...

  An initial character count shows a very not quite random distribution. I expect this might yield interesting results if people poke at it further.

  As far as the web utility is concerned...I suck had doing CSS by hand, so it will be stuck as ugly for now!

  I hope to add support for more OTP books if they are made and also for entering your own pad (although many sites already work with that method)

• A first step

  gannon • 09/11/2014 at 19:52 • 0 comments

  This is my first step into HTML5. So far it looks like things are a bit simpler since I last have done web development.

  After many attempts to figure out why a few parts of my code were not working I realized I have completely forgotten Javascript scemantics....oops

  The initial attempt had two pages of the pad entered in using an OCR program on the published images....needless to say there were typos due to this.

View all 5 project logs