[ale] Story of Geek as Magician

[Ed Landa]

> circumstances).  Because similar locks had been used to secure other
> doors in the building, I had already been wondering about them and how
> much protection they actually afforded.  

This reminds me of Richard Feynman's safe-cracker escapades at Los Alamos
during the Manhattan Project.

He first discovered that you could be off by 2 numbers in either direction when
dialing in the combination.  This reduced the number of possibilities from
1,000,000 to 8,000.  By only re-dialing the last digit without disturbing the
first two of each he got it down to where he could crack dial all of the
combinations on the safe in a max of about 8 hours, but usually in about 4
hours.

He took it even further when he discovered that when the safe was open, he
could slowly move the dial and watch for it to lock.  When it did, the previous
number was the 3rd digit in the combination.  He could repeat this process for
the other 2 numbers, but capturing the first was a lot of work.

He honed his technique to the point that he would stand in a cow-orkers office
and idly play with their safe dial and pick off the last 2 digits of their
combination.

When he was later asked to crack their safe, which he often was called to do as
his reputation as a safe-cracker grew, all he had to do was try the 20 possible
first digits (remember the off-by-2 discovery).  He usually did it in a closed
office in just a minute or so and then he would read for 20-30 minutes before
coming out to announce his success.  A master has to protect his secrets...

Ed


 PGP signature