[ale] little math

[Jim Popovitch]

On Fri, Feb 12, 2010 at 18:25, JK <jknapka at kneuro.net> wrote:
> Wow, you seem to have seriously mis-interpreted my words. Sorry about
> that.

I don't think that I misunderstood you, I just don't think your
analogy (spurious fluff data in a comms line) applies to file based
encryption.

> All I was saying is that the goal is to hide genuine encrypted
> data in an ocean of random data, so as to make an attack more difficult.

I don't think that was the OPs original goal (but I'm too lazy to try
and find his original email).  Certainly the person who forked this
thread didn't have that goal in his post.  Yet this conversation keeps
getting pulled into that direction.

> Obviously any repetition in the fake data is going to be a problem,
> right?  Since it makes the location of the non-fake data obvious?

I don't believe that.  The whole point behind file based encryption is
such that the known file isn't readable.  Nothing more, nothing less.
 The notion of hiding encrypted files, in random data, may seem more
secure, but so would doubly encrypting the single file.. and a
thousand other things (hopefully that won't spawn a thousand thread
forks).

-Jim P.