The section of the manual discussing the secure delete function references the Gutmann article on secure deletion. From what I understand, Gutmann himself has more recently said that with modern drives the patterns mentioned in his original article are largely pointless and that the 37 passes he recommended are unnecessary. One or two passes with random data are about the best you can do with modern drives.
If multiple passes are specified for the DOpus secure delete, does it follow the Gutmann patterns for those passes or is it following some other patterns? I suppose the question of greater concern to me is if I specify a small number of passes, like, say, one, is that pass going to contain random data or will it be some other predictable pattern that may not be all that secure if it's the only pass used?
It would be nice if the manual and/or preferences provided some indication of how many passes should be considered sufficient to protect against various levels of snooping and types of recovery. For instance, one pass might be sufficient to protect against normal deleted file recovery tools. Three passes would protect against any software-based recovery. Nine passes should protect against most types of hardware-based recovery. I realize there could be no guarantee, but some general guidelines would be nice.
[quote="Patch"]The section of the manual discussing the secure delete function references the Gutmann article on secure deletion. From what I understand, Gutmann himself has more recently said that with modern drives the patterns mentioned in his original article are largely pointless and that the 37 passes he recommended are unnecessary. One or two passes with random data are about the best you can do with modern drives.
If multiple passes are specified for the DOpus secure delete, does it follow the Gutmann patterns for those passes or is it following some other patterns? I suppose the question of greater concern to me is if I specify a small number of passes, like, say, one, is that pass going to contain random data or will it be some other predictable pattern that may not be all that secure if it's the only pass used?
It would be nice if the manual and/or preferences provided some indication of how many passes should be considered sufficient to protect against various levels of snooping and types of recovery. For instance, one pass might be sufficient to protect against normal deleted file recovery tools. Three passes would protect against any software-based recovery. Nine passes should protect against most types of hardware-based recovery. I realize there could be no guarantee, but some general guidelines would be nice.[/quote]
I would say that the manual does provide this guidance (although it is disclaimed). Read page 124:
[quote="The Opus User Manual Authors"]Directory Opus 8 has the ability to securely delete files by overwriting them many times before they are actually deleted. This uses algorithms similar to those recommended by the US NSA and US Department of Defense for securely destroying the contents of a file to make recovery by forensic analysis of the magnetic medium of the disk extremely difficult if not unlikely.
Preferences - File Operations - Deleting contains an option for the number of passes for each file – files are actually deleted by overwriting with a specific random pattern algorithmically chosen according to the number of passes selected. Three passes is regarded as quite secure and using more that this makes the process much slower and would seem like overkill. Three passes should be good enough for all but NSA scrutiny – But don’t quote us![/quote]