In <v02140b07aeaed1f1f746@[10.0.2.15]>, on 11/12/96 at 05:20 PM, azur@netcom.com (Steve Schear) said: . Attila said: . .> I wonder if anyone told them the data recovery teams can pull .> multiple layers of data off a disc after it has been stripped? .I've often wondered how effective disk encryption programs, like DiskCrypt, .are at hiding data from a Class III adversary (e.g., NSA). These programs .offer a means of 'cleaning' the data and directory space using randomized .patterns applied one or more times, but this might be insufficient for .removing data written slightly off-track by a previous write cycle. . I don't have any direct experience with 'serious' error recovery either; however, over a few years span listening to some my friends who have not come in from the cold, I received the very definite reading the key is a differential measurement on flux and organization which expands the 'hearing' range exponentially -even if it is multiple destructive formatting. I written software programs for C2 and B1 clearance, including primitives which wiped as they went. I used a pattern which shifted and repeated itself --not too ridiculous on unix if you are using type 2 read/write primitives. messy? yes, in its own simple way. The software was certified by multiple fairly sensitive regula- tory agencies, plus the usual few for whom there is no name. you don't expect, and you do not receive, anything more than a yes/no. but it was in these circumstances I 'heard' the comments on recovery. and their ability to measure that differential gets better every year! No, I do not consider 'erased' disks safe from a determined federal agency. I keep sensitive stuff on zip drives which are removable, and easily crippled forever. that and our principal [a]vocation: stronger cryptography than the hardware can bruteforce or trick. -- Cyberspace and Information are Freedom! FUCK your WIPO, too. -attila