Hi, just for completeness I post my answer to Henrik's message. (I accidently missed to CC to the list). My question whether the hidden partition is overwritten by the outer volume is now clear to me as well, since it is answered in the truecrypt documentation. Begin forwarded message: > > > > > My idea for self-destruct would be the following: > > .... > > Only an idiot would let you type in the password yourself. And what if > they backup your data first before trying to "hack" into it? A known > from-the-shelf solution will probably not be safe enough because your > attackers will know about it. You're completely right, though that solution does not require myself to type in the password ;) Anyway, as I already stated at the end of the last message (right before the glorious overlapping partitions idea), an independant self-destruct mode might not be implementable (since it always requires the "right" software to work). (As you said, the attacker will know about it.) > > A more feasible solution might be some kind of "time lock feature" > (multiple locks, automatic password change...) but there are lots of > scenarios to take into consideration here. Sounds pretty circumstantially to me. I also guess that this solution would not protect the user from forced password disclosure. > > > > > Maybe it would be better to just store 2 data-partitions in one (real) > > partition ? > > This is in the lines what True Crypt 4.x implements. Check it out > http://www.truecrypt.org/ Truecrypt's two levels of plausible deniability sound rather good to me, thank you for the tip. I have to RTFM first, but up to now I found one pretty bad disadvantage: * currently no full linux-support (no support to create volumes on Linux) So, I guess I'll have to wait and/or port it myself. Currently it is not completely clear to me, if the hidden partition is overwritten by the outer volume. Thank you for your fast answer, sincerly, stefan > > Regards, > Henrik Holst --------------------------------------------------------------------- - http://www.saout.de/misc/dm-crypt/ To unsubscribe, e-mail: dm-crypt-unsubscribe@xxxxxxxx For additional commands, e-mail: dm-crypt-help@xxxxxxxx
