[rescue] nuking from orbit

[G W Adkins]

> On Wed, Aug 29, 2001 at 06:08:48PM -0400, G W Adkins wrote:
> > This is ridiculous.  1 low-level format/verify pass from a real utility,
and
> > the data's GONE...  Forever.

<< Numerous other replied not quoted for brevity's sake >>

Alright, I'll take this as a challenge...

We'll start with a disk...  Let's say a Seagate ST1480N,
1.  we'll write 10 10meg Images of celebrity faces on the disk, and place a
record of the identities in the hands of a third party.
2.  Let me have it for 6 hours. (In working order when I get it, in working
order when it leaves)
3.  Then we send the disk to anyone you like.  Anyone you can get...  And if
they can properly name two out of ten of the images on that disk, I'll
_EAT_  the disk.  (we'll just break it up into swalowable sized pieces
first.)

You see, it's something like this...  if a residual image remains after you
place data on the disk, then a residual image remains after each of the 9
bit-stepped test patterns I write over the disk.  Sooo, is each residual
image weaker than the one before, allowing "layered" reconstruction of the
data?  No.  The residual left from the 420 Megs of pr0n I fill the disk with
before it leaves will be the last and therefore the strongest after-image
left on the disk.  Try to tell John Voight from Christopher Walken through a
picture of Anna Kournikova's backside.