[Cryptography] NSA versus DES etc....

John Gilmore gnu at toad.com
Thu Oct 2 05:32:03 EDT 2014

> One point for:  Suite A and friends, which remains a heavily shared secret.
> ...
> I guess the various well-funded enemies have figured out each other's
> secret algorithms by now, but out of politeness and common interest they
> cartelise the secrets.

Let's go a bit deeper into this.  Politeness?  Common interest?  Really?

Suppose Nation X reveals big Nation U's sooper secret crypto
algorithms.  Then Nation U is embarrassed -- and possibly has to go to
great trouble and expense to update all their crypto algorithms.

The only time Nation X has a real interest in keeping the algorithms
secret is when Nation X has cracked them and don't want Nation U to
know it yet, since they might change to an as-yet-uncracked system.
But if Nation U is running its spooks on crackable crypto, in these
days of gigahertz fingernail sized embedded systems, Nation U's secret
bureacracy is sounding new lows in incompetence.

It's likely that Nation X could get away with revealing the secret
algorithms without implicating themselves; they could find some
hacker, academic, activist, freedom-of-information maven, journalist,
or someone else to actually do the public posting.  They may only have
to gently steer some of these folks in the direction of asking the
question, or to finding information that has been left lying around on
some obscure user-contributed web site from some long-dormant IP
address.  Or the classic brown paper envelope that "fell off a truck".

So what's the real reason?  "It just isn't done"?  Come on, these
guys do every other *&$(#)! thing they aren't supposed to do -- why
not this one?


More information about the cryptography mailing list