1 |
begin quote |
2 |
On Wed, 5 May 2004 20:04:44 -0400 |
3 |
Ryan Voots <simcop2387@×××××.com> wrote: |
4 |
|
5 |
> On Tue, 4 May 2004 09:54:36 +0200 |
6 |
> "Marc Ballarin" <Ballarin.Marc@×××.de> wrote: |
7 |
> > shred is not reliable on modern filesystems. Clever block allocation |
8 |
> > algorithms, journalling and on-disk write-caches make secure |
9 |
> > deletion from |
10 |
> > userspace unreliable (shred's man page lists even more examples). |
11 |
> |
12 |
> i've actually thought about that problem and was wondering, since the |
13 |
> clever file systems and such would probably not want to move the |
14 |
> blocks around on every write, why couldn't you zero out the file |
15 |
> backwards (i understand this could end up a buffering issue but i |
16 |
> think that can be worked around) after that you just delete it like |
17 |
> normal? |
18 |
|
19 |
one word: |
20 |
blockreallocation |
21 |
(well.. *cough* ;) |
22 |
|
23 |
modern harddrives are actually larger than specified, and use something |
24 |
around 1Gb (perhaps more?) to shuffle around "bad blocks" and damaged |
25 |
sectors and do other black majjik. |
26 |
|
27 |
you cant even reach theese from software. |
28 |
|
29 |
put simply: You can't delete things from your harddrive. You can only |
30 |
reuse the space and hope for the best. |
31 |
|
32 |
//Spider |
33 |
|
34 |
|
35 |
-- |
36 |
begin .signature |
37 |
Tortured users / Laughing in pain |
38 |
See Microsoft KB Article Q265230 for more information. |
39 |
end |