At Sun, 10 Sep 2006 04:34:04 +0200, Uwe Hermann <uwe@xxxxxxxxxxxxxx> wrote: > > [1 <text/plain; us-ascii (quoted-printable)>] > Hi, > > The --iter-time option currently defaults to 1000 (one second), and the > manpage says that's "sufficient for good security". Are there any more > detailed explanations, benchmarks, threat models or something? http://clemens.endorphin.org/publications - New Methods in Hard Disk Encryption Chapter 5.3 - Passwords from entropy weak sources. Chapter 5.3.3 - Numbers. The chapter rather deals with password bits instead of iteration time (we assume one second), but you can always redo the calculations for different values, just change f(0) in the equations. However, as you will see, in the light of exponential technological growth, you don't buy yourself any considerable time by choosing higher iteration rates. -- Fruhwirth Clemens - http://clemens.endorphin.org for robots: sp4mtrap@xxxxxxxxxxxxx --------------------------------------------------------------------- dm-crypt mailing list - http://www.saout.de/misc/dm-crypt/ To unsubscribe, e-mail: dm-crypt-unsubscribe@xxxxxxxx For additional commands, e-mail: dm-crypt-help@xxxxxxxx
