Changing preferences

Robert J. Hansen rjh at
Wed Sep 24 04:48:36 CEST 2008

Faramir wrote:
>   I think I will add some more algos, to avoid using 3DES (while it
> should be safe enough... I don't like the solution "lets do it 3 times")


Not to ask a dunce question here, but why not?

It's perfectly safe.  In fact, 3DES is probably the most trustworthy
algorithm on this list.  A few years ago when Schneier was asked for his
pick for "most trusted encryption algorithm," he said something like
"3DES.  Nothing else even comes close."  Sure, use AES for new crypto
software, but if you absolutely _must_ have the most overdesigned,
overbuilt thing out there...

It's been subjected to withering cryptanalysis for coming up on 30 years
now.  It's one of the standard ciphers graduate students are exposed to
in cryptography/cryptanalysis courses.  It has turned a generation of
brilliant young graduate students into burned out alcoholic wrecks.  I
have participated in bar crawls after getting beaten by 3DES.

It is big, clumsy, ungainly and slow.  It has all the aesthetic values
of the Soviet Realism school of art, and processes data about as fast as
a snail coming off a three-day scopolamine trip.

And it is still beating up every cryptanalyst out there and stealing
their lunch money.

If you don't like 3DES because it's slow, okay, fine, I can respect
that.  But objecting to "let's do it three times" is nonsense.  Do you
object to Blowfish because it does it does it 16 times?

More information about the Gnupg-users mailing list