> >If you want to do cryptography, be careful. Do a search on Phil Zimmerman
> >and see all the crock the government put him through. By the way, DES is
> >very easily cracked by the (USA) government. In fact, they have the golden
> >key that would allow them to see any DES encryption. I heard Phil Zimmerman
> >got into such trouble, because they (the government) couldn't crack his
> >code which was based on RSA.
> >
> >Of course, I often had dreams of cracking the RSA using hundreds of
> >thousands of PICs in parallel. So, anyone else want to try?
>
> Lehigh University did this last year.  All the PC's on campus (during
> summer) figured it out in a few days.  But you probably won't read about it
> in your local newspaper.

"THE MAGIC WORDS ARE SQUEAMISH OSSIFRAGE" if I recall, was encoded as a
really big number (taking digits pairs for letters, hence

200805001301070903002215180419...
 T H E   M A G I C   W O R D S

and this number was then encrypted using a fairly short key (something like
400 bits).  It took quite awhile, but that particular key was broken.  On
the other hand, the difficulty of breaking an RSA key increases exponentially
with the key's length.  Keys under 512 keys are considered "toys", but even
if every electron in the known universe could be harnessed to perform a
billion computations per second, it would still take millions of years to
break 2048-bit keys (btw, producing a new 2048 key takes about 256 times as
long as producing a new 512 bit key; public-key operations on the longer key
will take about 16 times as long as with a 512-bit key, and private-key oper-
ations will take about 64 times as long.  Note that while the longer keys are
a computational pain to use they are hardly impossible, especially on faster
machines.)