18 Aug 2004 Panic 2?
"Panic! UNIX System Crash Dump Analysis" By Chris Drake and Kimberley Brown is a great book that any UNIX or Solaris geek would find very useful.

Authors hint at a possible second version of this book on the panic website http://www.piaffe.org/panic/. It has a quick survey to gauge the interest about its 2nd version. Take a moment to give your feedback.

Link |

18 Aug 2004 Breaking a Mysterious Code
While reading the news about breaking hashes in crypto 2004 reminds me of my own little code breaking experience years ago.

I was in high school and was simply tuning radio on a lazy afternoon. (more like channel switching on TV). I heard a faint voice speaking numbers at one particular frequency on meter wave. The voice was so faint at radio's normal volume, that you would easily miss it while tuning. I stopped, increased the volume and listened. It was a female voice speaking a stream of numbers in a sequence of 3 digits - small pause - 2 digits - pause - and repeat..like 134 22 .. 445 37 .. 662 22 ...

It seemed endless and so increased my curiosity. Grabbed a pencil and paper and started noting it down. after about 10 minutes it said "repeat". That tripled my curiosity. The sequence of numbers continued for another 20mins. Then silence. 5 min passed and nothing was heard. I noted down the frequency and time. I Switched off the radio. Checked it once in every hour. Still it was all silence. Guessed that broadcast might start at same time tomorrow.

Next day switched on the radio much before an estimated time. Same voice started at an exact clock time - may be 4:00pm in the afternoon. Some four digits spelled at slow pace and 4 ... 6 ... 7 ... 3. and the regular three-two .. sequence started. It continued for about 20 minutes before the word "repeat" and I could see it was to double check previously spoken sequence. Silence at the end as before.

It happened daily. I collected the coded messages 3 more times and tried to understand what it might be. Clearly it is some one sending out an encoded message. How could they have coded it? At first I tried to apply frequency analysis by hand (I did not have access to a computer). I would take the first number, scan the list till I find a match.. and so on. But how do I know what constitutes a single character? is it the sequence of three digits, two digits or divided into sets of four digits? What could the language be? Would they be using alphabets or syllables? Have they padded it with random noise numbers?

There were some assumptions I could make. Some one is using them to communicate to someone who is on the move. May be with out a telephone. Mobile phones weren't in use those days. They repeat the transmission to reduce errors. Since it is radio, and there is a possibility that you will get a random noise in the repeated message, the coding algorithm must be able to withstand corruption of data. Since it would be humans who are getting it and decoding it, decoding algorithm has to be simple enough. May be just looking up against a table or just doing some mental arithmetic.

Well, the difficulty was I did not know the language they used and the algorithm used. It might be Russian, English or Tamil or any of the three hundred languages spoken in India. I can easily guess the native language of an Indian speaking English. For eg., Biharis speak English with an accent different from Bengalis. But the radio voice mechanically spoke only digits. I could say that it was atleast from the subcontinent - but who? What organization could be doing it? I don't know the answer. India had always been full of spies. The Great Game that started in 1800s is still being played - just that there are more participants! LTTE was active in TamilNadu and SriLanka those days and were known to use radio and could be one possible guess. CIA might be another. May it was the Indian secret service which no one knows the name or may be Pakistani ISI. There were naxalities in Hyderabad region and many other organized groups..

Those messages are still a mystery to me. I don't know if you can still tune to get them these days. I haven't preserved them. I had then concluded that breaking the code by hand where you do not know who the sender and receiver are and the language they speak and algorithm they use, is nearly impossible. But it did instill an interest in me about cryptography and cryptanalysis.

Link | Comments [3]