On Sun, Aug 10, 2014 at 5:20 AM, AW <debian.list.trac...@1024bits.com> wrote: > On Sat, 09 Aug 2014 16:08:41 -0400 > Gary Dale <garyd...@torfree.net> wrote: > > > Whatever for? There are better checksums and md5 doesn't provide error > > correction? Even the MD5 man page advises using sha checksums instead. > > md5sum provides a relatively quick check... if it fails, then use the "real" > check, i.e. pars. This saves [or seems to save] computing resources... > > However, it was just a suggestion...
Interesting suggestion. Write the error-correction files, but only test them if the md5 checksum fails. Admittedly, this is not a situation where we are worried about attackers, but I'd feel more comfortable, myself, with actually checking the full error correction data. Both have holes, but the error correction codes provide deeper checks. (Take more space, too, but since you want to recover when you can, without referring to backups, the space consumption is part of the price you expect to pay.) If I were being paranoid about errors, I think I'd use full backup with grandfathering and periodic complete snapshots on permanent file, and with both error correction and sha256 checks, to catch as much corruption as possible before it gets into the backup files. -- Joel Rees Be careful where you see conspiracy. Look first in your own heart. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/caar43imvnbobfy9mlkj2enm8skefz+jsuojqxq9ttbqnt63...@mail.gmail.com
