Random seed for symetric encryption

Christoph Anton Mitterer cam at mathematica.scientia.net
Thu Jan 12 01:48:09 CET 2006

Nikolaus Rath wrote:

>Now I wonder why gpg needs random data for symetric encryption. Should
>I care about the message or not? And how can I make it disappear?
As far as I know, even for symmetric encryption gnupg uses a session key
package, which is than encrypted via s2k-algorithms (your passphrase and
salt and so).

See section 5.3 of the standard

The reason for doing so is mainly that you can specify more than one

The standard would also allow to directly use the s2k-generated key, but
I don't know wheter or not gpg support this.


More information about the Gnupg-users mailing list