From Newsgroup: comp.os.ms-windows.misc

I'm about to do a fresh reinstall of my computer. Will FORMAT in dos do, or is there some additional
stuff i need to do to completely wipe the sucker clean?


--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

Format will "wipe the sucker clean", unless you have files that would compromise national security on your hard drive.

Regards, hawk

innocent bystander wrote:

I'm about to do a fresh reinstall of my computer. Will FORMAT in dos do, or is there some additional
stuff i need to do to completely wipe the sucker clean?

--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

"hawk" <hawk@spamex.com> wrote in message news:vht7ulddjr5rc0@corp.supernews.com

Format will "wipe the sucker clean", unless you have files that would compromise national security on your hard drive.

Regards, hawk

And what if i do?

--
Listening to silence


--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


Do a low level format or use one of the many utilities available to wipe the disk clean.

innocent bystander wrote:

"hawk" <hawk@spamex.com> wrote in message news:vht7ulddjr5rc0@corp.supernews.com

Format will "wipe the sucker clean", unless you have files that would compromise national security on your hard drive.

Regards, hawk

And what if i do?

--
Mike Walsh
West Palm Beach, Florida, U.S.A.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

innocent bystander wrote:

"Mike Walsh" <mikew137@sbcglobal.net> wrote in message news:3F1EB358.B7D4AB74@sbcglobal.net

Do a low level format or use one of the many utilities available to
wipe the disk clean.

What does a low level format mean?

What utilities do you mean?

Thanks.

I suppose you could try BCWipe.
http://www.jetico.com/index.htm#/bcwipe.htm


--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

Format will "wipe the sucker clean", unless you have files that would compromise national security on your hard drive.

And what if i do?

You wouldn`t be asking - you would be taking it to the local security
office to be destroyed properly.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

In alt.msdos

innocent bystander <spam@someguy.com> wrote:
: "hawk" <hawk@spamex.com> wrote in message
: news:vht7ulddjr5rc0@corp.supernews.com

Format will "wipe the sucker clean", unless you have files that would compromise national security on your hard drive.

:
: And what if i do?

Check the size of the file, erase it, then
copy a bigger file to that file name.

It's gone.

Joe Fischer

--
3
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

A low level format will (among other things) overwrite all data on the

disc. Most format utilities will do a read test and remove the file table
but will leave the data intact.

Some people have reported that a disk was made unusable by attempting a low level format. Get a low level format utility that was designed for your

drive from the hard drive manufacturer if possible. I have done low level formats on IDE drives with vender supplied utilities without any problems.
If you are using SCSI disks you can use the adapter utility.



Are you sure a low level format does the job?
It's a much faster process than using a utility that overwrites everything several times.

Keu


--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On 24 Jul 2003 01:59:31 -0400, Joe Fischer <gravity1@shell1.iglou.com>
wrote:

Check the size of the file, erase it, then
copy a bigger file to that file name.

It's gone.

How do you know that the "bigger" file has been copied to the same
location (on the hard disk) as the one you just deleted?


Jim.

--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


"James Egan" <jegan@jegan.com> wrote in message news:4ddvhvkq3mda3jcsun6napogtvprllin0u@4ax.com...

On 24 Jul 2003 01:59:31 -0400, Joe Fischer <gravity1@shell1.iglou.com>
wrote:

Check the size of the file, erase it, then
copy a bigger file to that file name.

It's gone.

How do you know that the "bigger" file has been copied to the same
location (on the hard disk) as the one you just deleted?

If the file is erased first, it wont. However, even if a file wiping utility
is used, you are not guaranteed that all traces of the file are actually
gone. Depending on the program that created the file, there very well may be many deleted temporary copies scattered across the drive.

If one really needs to be concerned with removing old data, then delete whatever files and use a free space wiping utility. Of course, there could still be remnants left in the slack space of any files kept. OTOH, as long
as the drive has not been repartitioned, then a plain FORMAT/U is sufficient
to remove all data from the partition.

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On 24 Jul 2003 17:49:52 -0400, Joe Fischer
<gravity-central@shell1.iglou.com> wrote:

In alt.msdos

James Egan <jegan@jegan.com> wrote:
: Joe Fischer <gravity1@shell1.iglou.com> wrote:
:

Check the size of the file, erase it, then
copy a bigger file to that file name.

It's gone.

:
: How do you know that the "bigger" file has been copied to the same
: location (on the hard disk) as the one you just deleted?
: Jim.

Then don't delete it, just copy the bigger
file to that filename and tell it to overwrite it. :-)

How many guesses are you allowed, Joe?

I don't think that will work either. Maybe a resident expert can
explain how "copy" works. Does it create a file in its own (formerly
free) space and then delete the reference to the original or is the
original filespace used?


Jim.

--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


"James Egan" <jegan@jegan.com> wrote

I don't think that will work either. Maybe a resident expert can
explain how "copy" works. Does it create a file in its own (formerly
free) space and then delete the reference to the original or is the
original filespace used?

That answer may be version specific... OTOH, one could use a programming language like BASIC to open the file in binary mode and simply write garbage data directly to the file where it resides.

(See my other post to this thread too)

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Thu, 24 Jul 2003 18:31:30 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<snip>

If one really needs to be concerned with removing old data, then delete >whatever files and use a free space wiping utility. Of course, there could >still be remnants left in the slack space of any files kept. OTOH, as long
as the drive has not been repartitioned, then a plain FORMAT/U is sufficient >to remove all data from the partition.

I don't think that /U does any overwriting. It just doesn't check for
any existing partition information, and then writes the FAT's and the
root directory. Any data on the drive should still be there. But you
would need a disk sector editor to see it.

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

Joe Fischer wrote:

In alt.msdos

innocent bystander <spam@someguy.com> wrote:
: "hawk" <hawk@spamex.com> wrote:

Format will "wipe the sucker clean", unless you have files that would compromise national security on your hard drive.

:
: And what if i do?

Check the size of the file, erase it, then
copy a bigger file to that file name.

It's gone.

Joe Fischer

*** Not necessarily. COPY will not always place a new file directly over
the old. As well, copying a larger file will not necessarily erase all
parts of some previous files because of the way a file is written to the
disc.

Think of memory locations on a disc as a series of shoe boxes. If you
place an object in a box, it may not fill the box completely. The space
left over will have remnants of whatever was in the box previously. This
really shows up when you have an object which cannot fit in one box. It
must be broken apart so that one part fills the first box and the remnant
goes into only a small corner of the next available box. That means a significant portion of the previous object which occupied the second box
will be able to be seen.

Now let's translate this to a computer disc. Each of the shoe boxes is called an "allocation unit", or more commonly a "cluster". Because of the methods employed to handle disc space, two different files cannot occupy
the same allocation unit, even if they could fit. Thus on a larger hard
drive, a 1k batch file must reside in a 32k allocation unit (*), meaning
that 32k of space is wasted. Not only that, but 31k of previous file
remnants are still there and may be read with the proper software.

So, copying a different, but same-size file to the hard drive might end
up in the same allocation unit, but there is no guarantee it will and
no guarantee it will directly overwrite all of the previous file.

Richard Bonner
http://www.chebucto.ns.ca/~ak621/DOS

* Different hard drives have different allocation unit (cluster) sizes.
The larger the drive, the larger these units are and the more space they
waste. Thus, if one uses a larger hard drive, one must actually have a
drive much larger than needed because of the greater amount of wasted
space. The same programs will actually take up more room on a larger
drive.

One way to overcome this is to partition the drive. Smaller partitions
mean smaller clusters, and thus less wasted space. Partitioning can also
give a slight security/anti-virus advantage, and it contributes to organisation. I always partition my drives.

R.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

In alt.msdos

Richard Bonner <ak621@chebucto.ns.ca> wrote:
: Joe Fischer wrote:

In alt.msdos
Check the size of the file, (do not delete) then
copy a bigger file to that file name.

:

It's gone.
Joe Fischer

:
: *** Not necessarily. COPY will not always place a new file directly over
: the old. As well, copying a larger file will not necessarily erase all
: parts of some previous files because of the way a file is written to the
: disc.

I think it will if no other files are written
first.

: So, copying a different, but same-size file to the hard drive might end
: up in the same allocation unit, but there is no guarantee it will and
: no guarantee it will directly overwrite all of the previous file.

I suggested a bigger file.

: * Different hard drives have different allocation unit (cluster) sizes.
: The larger the drive, the larger these units are and the more space they
: waste. Thus, if one uses a larger hard drive, one must actually have a
: drive much larger than needed because of the greater amount of wasted
: space. The same programs will actually take up more room on a larger
: drive.
:
: One way to overcome this is to partition the drive. Smaller partitions
: mean smaller clusters, and thus less wasted space. Partitioning can also
: give a slight security/anti-virus advantage, and it contributes to
: organisation. I always partition my drives.
: R.

I don't any more, because I use multiple
drives, and I don't want to have to think about
how DOS assigns drive letters.

If a bigger file is written to the same
filename it should fill all prior allocation units
and then start additional ones.

When I had high security files I experimented,
I want to be sure, and not take anybody's word for it.

But then I used 5.25 inch floppies, a scissors
will erase all data, as will a match.


Joe Fischer

--
3
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


<arargh307NOSPAM@NOW.AT.arargh.com> wrote in message news:jl81ivgekcrhsceuo2m2dej694s3hgni31@4ax.com...

On Thu, 24 Jul 2003 18:31:30 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<snip>

If one really needs to be concerned with removing old data, then delete >whatever files and use a free space wiping utility. Of course, there

could

still be remnants left in the slack space of any files kept. OTOH, as

long

as the drive has not been repartitioned, then a plain FORMAT/U is

sufficient

to remove all data from the partition.

I don't think that /U does any overwriting. It just doesn't check for
any existing partition information, and then writes the FAT's and the
root directory. Any data on the drive should still be there. But you
would need a disk sector editor to see it.

The /U means unconditional overwrite. The default action (/Q) does what you suggest above. If interested, copy some files to a floppy and format it with the /U switch. Try reading it with a disk editor. If you don't have a disk editor, you could also use a program that creates a diskimage file and just
use a file editor to view it.

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


"Joe Fischer" <gravity1@shell1.iglou.com> wrote:

Richard Bonner <ak621@chebucto.ns.ca> wrote:
: Joe Fischer wrote:

In alt.msdos
Check the size of the file, (do not delete) then
copy a bigger file to that file name.

:

It's gone.
Joe Fischer

:
: *** Not necessarily. COPY will not always place a new file directly

over

: the old. As well, copying a larger file will not necessarily erase all
: parts of some previous files because of the way a file is written to the
: disc.

I think it will if no other files are written
first.

The way DOS/COPY works is this... The pointer of the starting cluster is changed to point to the first available cluster. Then it fills each
available cluster from that point on.

Writing other files first would fill free clusters. Deleting other files
makes more free.

: So, copying a different, but same-size file to the hard drive might

end

: up in the same allocation unit, but there is no guarantee it will and
: no guarantee it will directly overwrite all of the previous file.

I suggested a bigger file.

Yes, however, you did not indicate how big or why. It must be big enough to first fill in *all* previous free clusters before it can begin to overwrite
the original file.

If a bigger file is written to the same
filename it should fill all prior allocation units
and then start additional ones.

Correct, however, without knowing where the file is stored and exactly how
many free clusters preceed it, how does the *average* person know how big of
a file is required? (You didn't quite specify with details.)

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Fri, 25 Jul 2003 19:40:42 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<arargh307NOSPAM@NOW.AT.arargh.com> wrote in message >news:jl81ivgekcrhsceuo2m2dej694s3hgni31@4ax.com...

On Thu, 24 Jul 2003 18:31:30 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<snip>

If one really needs to be concerned with removing old data, then delete
whatever files and use a free space wiping utility. Of course, there

could

still be remnants left in the slack space of any files kept. OTOH, as

long

as the drive has not been repartitioned, then a plain FORMAT/U is >sufficient
to remove all data from the partition.

I don't think that /U does any overwriting. It just doesn't check for
any existing partition information, and then writes the FAT's and the
root directory. Any data on the drive should still be there. But you
would need a disk sector editor to see it.

The /U means unconditional overwrite. The default action (/Q) does what you >suggest above. If interested, copy some files to a floppy and format it with >the /U switch. Try reading it with a disk editor. If you don't have a disk >editor, you could also use a program that creates a diskimage file and just >use a file editor to view it.

For floppys, I think that you are correct. Format actually does do a
hardware format.

However you said:
"as the drive has not been repartitioned, then a plain FORMAT/U is
sufficient to remove all data from the partition."

Floppys don't have partitions, and I think that my reply is correct
for hard disks.

The documentation I checked impiled that '/u' causes a 'format track
and verify' operation. I don't think that that command works on
current hard drives.

If I get time this weekend, I will setup a system, and find an old IDE
drive, and find out which way it actually works.

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

In alt.msdos

Todd Vargo <toddvargo@nccw.net> wrote:
: "Joe Fischer" <gravity1@shell1.iglou.com> wrote:

Check the size of the file, (do not delete) then
copy a bigger file to that file name.

:

It's gone.
Joe Fischer

I think it will if no other files are written
first.

:
: The way DOS/COPY works is this... The pointer of the starting cluster is
: changed to point to the first available cluster. Then it fills each
: available cluster from that point on.

If a file is written to the filename of an
existing file, it will use the allocated clusters
for the existing file first.

I think in the real world, if a file is
deleted and a file written to that filename, the
first available clusters were from that filename.
This is because most people do not do
much erasing, and the deleted file may very well
be the lower numbered clusters.

: Writing other files first would fill free clusters. Deleting
: other files makes more free.

I said if no other files are written first,
I said nothing about deleting.

Some of what I am saying is from experience,
which may or may not be correct.

: So, copying a different, but same-size file to the hard drive might

: end

: up in the same allocation unit, but there is no guarantee it will and
: no guarantee it will directly overwrite all of the previous file.

I suggested a bigger file.

:
: Yes, however, you did not indicate how big or why. It must be big
: enough to first fill in *all* previous free clusters before it can
: begin to overwrite the original file.

I think it rewrites the clusters of the
existing file in the same order they exist, then
writes new clusters.
It wouldn't make sense when overwriting
to not overwrite using clusters allocated to the
file being overwritten.

If a bigger file is written to the same
filename it should fill all prior allocation units
and then start additional ones.

:
: Correct, however, without knowing where the file is
: stored and exactly how many free clusters preceed it,
: how does the *average* person know how big of
: a file is required? (You didn't quite specify with details.)

DOS tells file size with a long directory,
and anybody using a computer should know 10 is
bigger than 5.
The location of the file is transparent
to the user, if DOS overwrites a file with an
identical file, it uses exactly the same clusters.
Why should it start out differently when
a file is modified to be bigger (as with word
processing and frequent saves to protect work)?

Doing image work I save files to the
same filename often, and that does not mean
empty clusters are written first each time.

But if you are using XP, ignore me.

Joe Fischer

--
3
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


<arargh307NOSPAM@NOW.AT.arargh.com> wrote

However you said:
"as the drive has not been repartitioned, then a plain FORMAT/U is
sufficient to remove all data from the partition."

Floppys don't have partitions, and I think that my reply is correct
for hard disks.

The documentation I checked impiled that '/u' causes a 'format track
and verify' operation. I don't think that that command works on
current hard drives.

I don't know what documentation you are reading, but if you read the documentation provided in HELP from MSDOS 6.xx, there is no mention about track/partition relationships. Basically, FORMAT/U treats clusters in a partition the same as it would physical tracks on a floppy.

HELP FORMAT
http://2dos.homepage.dk/batutil/help/FORMAT_S.HTM

Note, in Win95 and newer, FORMAT no longer stores unformat data at the end
of a drive, which is why the UNFORMAT utility was removed.

If I get time this weekend, I will setup a system, and find an old IDE
drive, and find out which way it actually works.

I'll be hanging on the edge of my seat!

BTW, make sure that when you type FORMAT/U that you do not in any way
include a /Q with the command. On some versions, the /Q simply disables the
/U, while others refuse to accept both (because they mean exactly the opposite).

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

For floppys, I think that you are correct. Format actually does do a hardware format.
However you said:
"as the drive has not been repartitioned, then a plain FORMAT/U is
sufficient to remove all data from the partition."
Floppys don't have partitions, and I think that my reply is correct
for hard disks.

From memory, a floppy can be /Q quick formatted if it has already been formatted

The /U option just means "unconditional" or essentially "force" a format (saves the "are you sure ?" "all data will be lost !" "get a grown up"
style of error messages)

The /U option will not remove all data from that partition - what it will
do depending on switches will...

/Q /U quick unconditional format (overwrite the FAT only)
/U skip through the drive rewriting all the blocks *

* just making sure the block can be accessed is not the same as removing
the data - depending on the size of the block, you may lose *some* data,
but much of the data would still be visible using utilities, although it
may not all be recoverable depending on the way the blocks have been
written. The only way to overwrite the data is to fill the drive
completely with 1`s or 0`s. If you wanted to "securely" wipe the drive,
you`ll need a sledgehammer, a grinder, and an oxy-acetylene blowtorch ;-)

I had one drive I was having trouble with that I reformatted with the /Q
quick switch, and couldn`t get it to do what I wanted. I then threw the
drive onto my Amiga which has been known to cure some weird and wonderful drive problems, and formatted it with that. I then put it back in the PC
and did some sort of unformat and discovered (I think my block size was
8k at the time) files smaller than 8k were fine, anything larger was
slightly corrupted.

The documentation I checked impiled that '/u' causes a 'format track
and verify' operation. I don't think that that command works on
current hard drives.

Not had to do it recently (use the /U switch), but I can`t see why not...

I`m sure i`m talking complete b*llocks though, i`ve not had much sleep !
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


"Colin Wilson" <btiruseless@btinternet.com> wrote

The /U option will not remove all data from that partition

I`m sure i`m talking complete b*llocks though, i`ve not had much sleep !

Before anyone else responds with further contradicting information, please check it for correctness before posting. At least one person intends to
verify me, so lets wait and see the result. ;-)

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

I`m sure i`m talking complete b*llocks though, i`ve not had much sleep !

Before anyone else responds with further contradicting information, please check it for correctness before posting. At least one person intends to verify me, so lets wait and see the result. ;-)

The question is, was *I* entirely right ? :-p
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Sat, 26 Jul 2003 04:16:25 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<arargh307NOSPAM@NOW.AT.arargh.com> wrote

However you said:
"as the drive has not been repartitioned, then a plain FORMAT/U is
sufficient to remove all data from the partition."

Floppys don't have partitions, and I think that my reply is correct
for hard disks.

The documentation I checked impiled that '/u' causes a 'format track
and verify' operation. I don't think that that command works on
current hard drives.

I don't know what documentation you are reading, but if you read the >documentation provided in HELP from MSDOS 6.xx, there is no mention about >track/partition relationships. Basically, FORMAT/U treats clusters in a >partition the same as it would physical tracks on a floppy.

HELP FORMAT
http://2dos.homepage.dk/batutil/help/FORMAT_S.HTM

Note, in Win95 and newer, FORMAT no longer stores unformat data at the end
of a drive, which is why the UNFORMAT utility was removed.

If I get time this weekend, I will setup a system, and find an old IDE
drive, and find out which way it actually works.

I'll be hanging on the edge of my seat!

BTW, make sure that when you type FORMAT/U that you do not in any way
include a /Q with the command. On some versions, the /Q simply disables the >/U, while others refuse to accept both (because they mean exactly the >opposite).

Ok here are the results of the test.

Built a system.
Used a 4 year old 4 gig seagate drive that was collecting dust.
Set the bios type to 'large' so that DOS 5 could see the whole drive.
Used DOS 5 FDISK to create a 20 meg FAT 16 partition.
Used DOS 5 FORMAT to format it.
Used QBasic to fill the drive with known data.

With each of MSDOS 5.00, 6.22, 7.00 (win95), 7.10 (win98):
Did a "format c:/u"
Used a disk editor to check if the data was still there.

IT WAS.

Format /U does NOT wipe data on hard drives.
It just zeros out the FATs and the root directory.



Also, this:

Note, in Win95 and newer, FORMAT no longer stores unformat data at the end
of a drive, which is why the UNFORMAT utility was removed.

implies that the data was still there, else there is NWIH to UNFORMAT
a drive.

About the only good thing was that the motherboard I found was one
that I had forgotten that I had. It is the kind that I use for
servers. And I need to build a new server pretty soon.


--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


<arargh307NOSPAM@NOW.AT.arargh.com> wrote

Ok here are the results of the test.

Built a system.
Used a 4 year old 4 gig seagate drive that was collecting dust.
Set the bios type to 'large' so that DOS 5 could see the whole drive.
Used DOS 5 FDISK to create a 20 meg FAT 16 partition.
Used DOS 5 FORMAT to format it.
Used QBasic to fill the drive with known data.

With each of MSDOS 5.00, 6.22, 7.00 (win95), 7.10 (win98):
Did a "format c:/u"
Used a disk editor to check if the data was still there.

IT WAS.

Format /U does NOT wipe data on hard drives.
It just zeros out the FATs and the root directory.

I stand corrected! I have not reformatted a HD in a long while, but I do
recall it wiped all data from a HD.

Now that you have a spare machine, could you try it with the /C switch?
FORMAT C: /U /C

Also, this:

Note, in Win95 and newer, FORMAT no longer stores unformat data at the

end

of a drive, which is why the UNFORMAT utility was removed.

implies that the data was still there, else there is NWIH to UNFORMAT
a drive.

IIRC (and I'm proven to be wrong on occasion), in DOS 6.xx I thought it was
the /Q that stored the unformat information.

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


"Colin Wilson" <btiruseless@btinternet.com> wrote in message news:MPG.198c6b871fad192998993a@news.cis.dfn.de...

I`m sure i`m talking complete b*llocks though, i`ve not had much sleep

!

Before anyone else responds with further contradicting information,

please

check it for correctness before posting. At least one person intends to verify me, so lets wait and see the result. ;-)

The question is, was *I* entirely right ? :-p

We haven't tested the sledgehammer, grinder, and oxy-acetylene blowtorch commands yet. :-]

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

In alt.msdos

Todd Vargo <toddvargo@nccw.net> wrote:
: I recommend you (and others) play with a disk editor before commenting
: further about what each of you *think* you know.

Like DEBUG?

Your premise that format does or ever did
anything to obliterate data seems to be wrong.

But as long as critical files are not more
than the size DEBUG can handle, or F800 bytes
(whichever is smaller) it is simple to
overwrite the file with all zeros, all FF's, or
any other string desired.

Simply load the file into DEBUG with

DEBUG filename.ext

and type

F0100 F800 00

then type

W0100


Debug will write the same number of bytes
that was in the file with all zeros, to the same
location.

Instead of 00, FF will write all FF's,
and "A N Y S T R I N G" will write any string.

Joe Fischer

--
3
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Sat, 26 Jul 2003 14:32:23 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<arargh307NOSPAM@NOW.AT.arargh.com> wrote

Ok here are the results of the test.

Built a system.
Used a 4 year old 4 gig seagate drive that was collecting dust.
Set the bios type to 'large' so that DOS 5 could see the whole drive.
Used DOS 5 FDISK to create a 20 meg FAT 16 partition.
Used DOS 5 FORMAT to format it.
Used QBasic to fill the drive with known data.

With each of MSDOS 5.00, 6.22, 7.00 (win95), 7.10 (win98):
Did a "format c:/u"
Used a disk editor to check if the data was still there.

IT WAS.

Format /U does NOT wipe data on hard drives.
It just zeros out the FATs and the root directory.

I stand corrected! I have not reformatted a HD in a long while, but I do >recall it wiped all data from a HD.

Now that you have a spare machine, could you try it with the /C switch?

It was never assembled that far. Just a MB & PS in a case, with
cables hanging out connecting to drives.

FORMAT C: /U /C

First I would have to find a drive with bad sectors:

/C Tests clusters that are currently marked "bad."

The only drives that I know that have bad sectors are either MFM,
ESDI, or SMD. No IDE or SCSI.

IIRC (and I'm proven to be wrong on occasion), in DOS 6.xx I thought it was >the /Q that stored the unformat information.

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Sat, 26 Jul 2003 22:40:01 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<arargh307NOSPAM@NOW.AT.arargh.com> wrote

"Todd Vargo" <toddvargo@nccw.net> wrote:

I stand corrected! I have not reformatted a HD in a long while, but I do
recall it wiped all data from a HD.

Now that you have a spare machine, could you try it with the /C switch?

It was never assembled that far. Just a MB & PS in a case, with
cables hanging out connecting to drives.

FORMAT C: /U /C

First I would have to find a drive with bad sectors:

/C Tests clusters that are currently marked "bad."

The only drives that I know that have bad sectors are either MFM,
ESDI, or SMD. No IDE or SCSI.

Why would a drive with bad sectors be required?

Because: /C Tests clusters that are currently marked "bad."
No bad sectors, no clusters marked "bad".

Why not just run the command on the same drive?

No clusters marked "bad".

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

In alt.msdos

Todd Vargo <toddvargo@nccw.net> wrote:
:"Joe Fischer" <gravity1@shell1.iglou.com> wrote:

Todd Vargo <toddvargo@nccw.net> wrote:
: I recommend you (and others) play with a disk editor before commenting
: further about what each of you *think* you know.

Like DEBUG?

:
: No! Debug is not a "disk editor".

It is for people who know how to use it,
but maybe one dangerous to use and of limited capability.

Your premise that format does or ever did
anything to obliterate data seems to be wrong.

:
: Hold the salt. "FORMAT C:/U/C" may be what I had in mind.
: I just do not have a spare HD right now to verify it myself.

I would send you one but postage has become
unaffordable.

My opinion is that DOS format stopped writing
data when disk size began to grow, the time involved
is too great.

As far as the original question about file
security is concerned, there aren't many people
selling used drives with critical data on them.
And the answer for that may be multiple
drive systems with small drives for the critical
data, and destroy them with a hammer instead of
selling them.

I don't have a desire to discuss situations
where the need to obliterate data is to avoid law
enforcement from reading disks.
There may be a lot of people in the US that
are worrying about the 800 cases in court over
music sharing, belonging to a web ring could
be construed as conspiracy and theft.

The good thing about compact disk media is
the ease of destruction, especially when most of
the commercial audio and video files have codecs
that link to the copyright owner's website if
played while online.

Debug will write the same number of bytes
that was in the file with all zeros, to the same
location.

:
: Did you use a disk editor to verify this? No, but
: I did, and the pointer to the first cluster was changed
: just like it is for COPY.
: As I said, use a "disk editor" before commenting further.

Ok.

: To do what you want (wipe a single file without moving it), you could use
: qbasic to write to the file in binary or random access mode.
:
: OPEN "filename.ext" FOR BINARY AS #1
: lLOF = LOF(1)
: wipe$ = "A N Y S T R I N G"
: DO UNTIL LOC(1) >= lLOF
: PUT #1, , wipe$
: LOOP
: CLOSE

I quit using basic when they quit using line numbers. :-)
If basic will overwrite existing clusters why won't DEBUG?

Joe Fischer

--
3
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

"arargh307NOSPAM@NOW.AT.arargh.com" typed:

First I would have to find a drive with bad sectors:

/C Tests clusters that are currently marked "bad."

The only drives that I know that have bad sectors are either MFM,
ESDI, or SMD. No IDE or SCSI.

I have an old 2.0 Gb Segate Hard-disk lying in my drawers. It is
completely filled with bad-sectors, so much that even Linux and
FreeBSD aren't able to change or write to the partition table. A month
ago I tried to remove the hard-metal case, and to have a look at the
disks inside. But I failed. The case is still intact, even after
hammering it.

--
Ayaz Ahmed Khan

Yours Forever in,
Cyberspace.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Sun, 27 Jul 2003 16:10:01 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<arargh307NOSPAM@NOW.AT.arargh.com> wrote

"Todd Vargo" <toddvargo@nccw.net> wrote:

Why would a drive with bad sectors be required?

Because: /C Tests clusters that are currently marked "bad."
No bad sectors, no clusters marked "bad".

Why not just run the command on the same drive?

No clusters marked "bad".

Never mind, it was not important anyway. :-(

All right, I plugged the drives back in, and under DOS 7.10 ran a
format c:/u/c
As I expected, it did exactly nothing about wiping data.

Then, I went in and by hand marked some clusters bad. (They weren't
really). Ran format c:/u/c again. Still didn't wipe any data. It did
clear the 'bad' cluster marks from the FAT, however.

I assume that the '/C' test means read the cluster, and if you didn't
get an error, assume that it is good.

I don't think any version of MS format will wipe data. At least none
that I have seen.

If you really need to wipe the data area of a hard disk, there are any
number of utilities floating abound the web that will do that. Some
even follow the goverment rules for security in that area.

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Sun, 27 Jul 2003 16:04:19 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

: qbasic to write to the file in binary or random access mode.
:
: OPEN "filename.ext" FOR BINARY AS #1
: lLOF = LOF(1)
: wipe$ = "A N Y S T R I N G"
: DO UNTIL LOC(1) >= lLOF
: PUT #1, , wipe$
: LOOP
: CLOSE

I quit using basic when they quit using line numbers. :-)
If basic will overwrite existing clusters why won't DEBUG?

The BASIC program above (add line numbers if preferred) accesses the disk >directly. The manner in which you used DEBUG does not. Actually DEBUG can >access specific sectors of a drive, however, you must know exactly where the >file resides and hopefully it's not out of DEBUG's reach. Its NOT worth the >effort.

It is possible to use DEBUG for a file anywhere on a drive, if the
file is not too big - about 300k to 400k limit:
do 'debug big.fil'
do 'r'
the 'r' is to find out where DEBUG ends.
use the 'f' command to fill memory up to 9000:ffff, but don't
overwrite any part of DEBUG.
do a 'w'
That should overwrite the file.

Better to use a wipe file program.




--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


<arargh307NOSPAM@NOW.AT.arargh.com> wrote in message news:7rj8ivsi0r1vhordq37ar2o9bj935h84em@4ax.com...

On Sun, 27 Jul 2003 16:04:19 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

: qbasic to write to the file in binary or random access mode.
:
: OPEN "filename.ext" FOR BINARY AS #1
: lLOF = LOF(1)
: wipe$ = "A N Y S T R I N G"
: DO UNTIL LOC(1) >= lLOF
: PUT #1, , wipe$
: LOOP
: CLOSE

I quit using basic when they quit using line numbers. :-)
If basic will overwrite existing clusters why won't DEBUG?

The BASIC program above (add line numbers if preferred) accesses the disk >directly. The manner in which you used DEBUG does not. Actually DEBUG can >access specific sectors of a drive, however, you must know exactly where

the

file resides and hopefully it's not out of DEBUG's reach. Its NOT worth

the

effort.

It is possible to use DEBUG for a file anywhere on a drive, if the
file is not too big - about 300k to 400k limit:
do 'debug big.fil'
do 'r'
the 'r' is to find out where DEBUG ends.
use the 'f' command to fill memory up to 9000:ffff, but don't
overwrite any part of DEBUG.
do a 'w'
That should overwrite the file.

^^^^^^
Basically, you are using DEBUG the same way Joe was.

Here is the setup I used to test it.

Create two different files. To keep the test legitimate, make the first
file larger than DEBUG's capability (about 600k should do).
Use a disk editor to identify the starting location of each file.
Delete the first file created.
Use your debug method to wipe the remaining second file.
Now examine the starting location of the wiped file.

My tests have shown the wiped file is written in the location of the first file, and the original contents of the supposed wiped file remain intact on
the drive.

Better to use a wipe file program.

Being a fellow BASIC programmer, you should know the BASIC program I posted
is a wipe file program without any bells or whistles. And with minor modification, the same program can be used to wipe free space, but why
reinvent the wheel. ;-)

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


<arargh307NOSPAM@NOW.AT.arargh.com> wrote:

All right, I plugged the drives back in, and under DOS 7.10 ran a
format c:/u/c
As I expected, it did exactly nothing about wiping data.

Then, I went in and by hand marked some clusters bad. (They weren't
really). Ran format c:/u/c again. Still didn't wipe any data. It did
clear the 'bad' cluster marks from the FAT, however.

I assume that the '/C' test means read the cluster, and if you didn't
get an error, assume that it is good.

I don't think any version of MS format will wipe data. At least none
that I have seen.

This is very discouraging. I was sure that I had wiped a HD before using
only FORMAT. Apparently I was totally wrong (it happens) about FORMAT's /U switch operation on hard drives. Thanks for volunteering your efforts and
sorry you wasted your time humoring me.

BTW, someone asked a question that maybe you can confirm. When you used the
/U switch, was the 'Are you sure' type message displayed? I seem to recall
/U did not suppress this message as was claimed. ;-)

If you really need to wipe the data area of a hard disk, there are any
number of utilities floating abound the web that will do that. Some
even follow the goverment rules for security in that area.

Yes, this was suggested to the OP already.

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Mon, 28 Jul 2003 03:29:00 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<snip>

This is very discouraging. I was sure that I had wiped a HD before using
only FORMAT. Apparently I was totally wrong (it happens) about FORMAT's /U >switch operation on hard drives. Thanks for volunteering your efforts and >sorry you wasted your time humoring me.

BTW, someone asked a question that maybe you can confirm. When you used the >/U switch, was the 'Are you sure' type message displayed? I seem to recall
/U did not suppress this message as was claimed. ;-)

Yes, it always asked. I don't think that there is any way to suppress
that question. If there is, it is not documented.

If you really need to wipe the data area of a hard disk, there are any
number of utilities floating abound the web that will do that. Some
even follow the goverment rules for security in that area.

Yes, this was suggested to the OP already.

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Mon, 28 Jul 2003 03:28:30 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<snip>

Create two different files. To keep the test legitimate, make the first
file larger than DEBUG's capability (about 600k should do).
Use a disk editor to identify the starting location of each file.
Delete the first file created.
Use your debug method to wipe the remaining second file.
Now examine the starting location of the wiped file.

My tests have shown the wiped file is written in the location of the first >file, and the original contents of the supposed wiped file remain intact on >the drive.

Yes, you right. I forgot that using when files, debug does normal
file opens & closes, instead of leaving the file open, and rewriting
it.

Better to use a wipe file program.

Being a fellow BASIC programmer, you should know the BASIC program I posted >is a wipe file program without any bells or whistles. And with minor >modification, the same program can be used to wipe free space, but why >reinvent the wheel. ;-)

Sure, that's how I clobbered the drive in my tests.

Something like: (not tested)

dim b as string * 512

open "c:\filler" for binary as #1

' some number larger than the size of the disk
for n&=1 to 65535
b="block:"+str$(n&)
put #1,,b
next n&

Run it, wait for it to get a disk full error, and presto, your free
space now has garbage in it. Remember to delete c:\filler, or a
windows 9x system will get sick, quick.

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

<snip>

If one really needs to be concerned with removing old data, then
delete whatever files and use a free space wiping utility. Of
course, there

could

still be remnants left in the slack space of any files kept. OTOH,
as

long

as the drive has not been repartitioned, then a plain FORMAT/U is >>sufficient
to remove all data from the partition.

I don't think that /U does any overwriting. It just doesn't check
for any existing partition information, and then writes the FAT's
and the root directory. Any data on the drive should still be
there. But you would need a disk sector editor to see it.

The /U means unconditional overwrite. The default action (/Q) does
what you suggest above. If interested, copy some files to a floppy and >>format it with the /U switch. Try reading it with a disk editor. If
you don't have a disk editor, you could also use a program that
creates a diskimage file and just use a file editor to view it.

For floppys, I think that you are correct. Format actually does do a hardware format.

However you said:
"as the drive has not been repartitioned, then a plain FORMAT/U is
sufficient to remove all data from the partition."

Floppys don't have partitions, and I think that my reply is correct
for hard disks.

The documentation I checked impiled that '/u' causes a 'format track
and verify' operation. I don't think that that command works on
current hard drives.

If I get time this weekend, I will setup a system, and find an old IDE
drive, and find out which way it actually works.

Gees,
Here is an easy way try it or dont bother.
Use the google search engine look for a file called delpart.exe
Download it to a bootable dos disk. Run the delpart.exe it will show and
ask you which partitions you want to delete. Select using the keyboard commands. After you blow it off tell it to save. Then when you reboot
machine with the dos disk do a fdisk/mbr this will make sure it blows the
boot sector off the drive. Then regular fdisk the drive then format and install.

No matter even if you use Power Quest Lost and Found it will not see
anything !!!!!!!

How ever if you are doing this on a Goverment PC then you better be doing
this legitiment or they will nab you which I would say KUDOS to them for busting you!

The above format /u is just an uncoditional command to format. Nothing
more! It is useless. How ever after the procedures I mentioned is done then you can format /u /s the drive which the again /u uncoditional and /s
system files aka io.sys msdos.sys files and command.com is copied to the
drive making it bootable again.

Oh btw the /q is quick format how ever the drive has to already be
formatted or it will do a regualar format~!!!!!!!

Again the regular formats is crap!

If this doesnt make sense then RTFM!

--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

arargh307NOSPAM@NOW.AT.arargh.com wrote in news:bcs4iv4k1jk7ktv1i5i0havdeqp77no37e@4ax.com:

On Sat, 26 Jul 2003 04:16:25 -0400, "Todd Vargo" <toddvargo@nccw.net>
wrote:

<arargh307NOSPAM@NOW.AT.arargh.com> wrote

However you said:
"as the drive has not been repartitioned, then a plain FORMAT/U is
sufficient to remove all data from the partition."

Floppys don't have partitions, and I think that my reply is correct
for hard disks.

The documentation I checked impiled that '/u' causes a 'format track
and verify' operation. I don't think that that command works on
current hard drives.

I don't know what documentation you are reading, but if you read the >>documentation provided in HELP from MSDOS 6.xx, there is no mention
about track/partition relationships. Basically, FORMAT/U treats
clusters in a partition the same as it would physical tracks on a
floppy.

HELP FORMAT
http://2dos.homepage.dk/batutil/help/FORMAT_S.HTM

Note, in Win95 and newer, FORMAT no longer stores unformat data at the
end of a drive, which is why the UNFORMAT utility was removed.

If I get time this weekend, I will setup a system, and find an old
IDE drive, and find out which way it actually works.

I'll be hanging on the edge of my seat!

BTW, make sure that when you type FORMAT/U that you do not in any way >>include a /Q with the command. On some versions, the /Q simply
disables the /U, while others refuse to accept both (because they mean >>exactly the opposite).

Ok here are the results of the test.

Built a system.
Used a 4 year old 4 gig seagate drive that was collecting dust.
Set the bios type to 'large' so that DOS 5 could see the whole drive.
Used DOS 5 FDISK to create a 20 meg FAT 16 partition.
Used DOS 5 FORMAT to format it.
Used QBasic to fill the drive with known data.

With each of MSDOS 5.00, 6.22, 7.00 (win95), 7.10 (win98):
Did a "format c:/u"
Used a disk editor to check if the data was still there.

IT WAS.

Format /U does NOT wipe data on hard drives.
It just zeros out the FATs and the root directory.

Also, this:

Note, in Win95 and newer, FORMAT no longer stores unformat data at the
end of a drive, which is why the UNFORMAT utility was removed.

implies that the data was still there, else there is NWIH to UNFORMAT
a drive.

About the only good thing was that the motherboard I found was one
that I had forgotten that I had. It is the kind that I use for
servers. And I need to build a new server pretty soon.

Ok folks once again Format /u uncoditional meaning I dont care for you to
ask me am I sure just do it. /s copy system files once you are done.
/q is a quick format it has to already be formatted for this to work no f*@king exception.

so now example

Format C: /u /s {will format C: /u "uncoditional" /s copy the system files afterwords. io.sys, msdos.sys and command com will be placed on the drive with this command which are needed to boot the system.

the Msdos.sys and io.sys are hidden.

Now if you have a already formated c: then you can do this

Format c: /u /s /q the q can be used in this case IF THE DRIVE IS PREFORMATED. Meaning it is already FORMATED as like the case above.

Also you can do this to a floppy as well.

Might I sugguest the old MSDOS book from Microsoft Press for Dos 5.0 or
even 6.0 usually still avaiable in Barnes and Nobles and Borders book
store! The 5.0 is a Blue Book and the 6.0 is a Green Book!

To wipe a drive such as a hard drive then you need to learn how to FDISK
not the same as format!!~!!!!!!!

FDISK creates a Partition to a fixed disk drive! Format makes the partition
or disk useable. Meaning you can not do this to a floppy. To make other sections on a Floppy you have to use ramdisk!

Fdisk/mbr this will wipe a partition how ever data stays! it will rebuild
the Partition table index.

Do not mix these up or you will be sorry.



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Mon, 28 Jul 2003 14:29:17 GMT, z <z@o.com> wrote:

<snip>

Gees,
Here is an easy way try it or dont bother.
Use the google search engine look for a file called delpart.exe
Download it to a bootable dos disk. Run the delpart.exe it will show and
ask you which partitions you want to delete. Select using the keyboard >commands. After you blow it off tell it to save. Then when you reboot >machine with the dos disk do a fdisk/mbr this will make sure it blows the >boot sector off the drive. Then regular fdisk the drive then format and >install.

No matter even if you use Power Quest Lost and Found it will not see >anything !!!!!!!

Unless 'delpart.exe' has explicitly wiped the data area of the
partition, all the data will still be there. Use a disk sector
editor, and look around. After the format and before the install.

<snip>

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

Use the google search engine look for a file called delpart.exe
Download it to a bootable dos disk. Run the delpart.exe it will show and >ask you which partitions you want to delete. Select using the keyboard >commands. After you blow it off tell it to save. Then when you reboot >machine with the dos disk do a fdisk/mbr this will make sure it blows the >boot sector off the drive. Then regular fdisk the drive then format and >install.

No matter even if you use Power Quest Lost and Found it will not see >anything !!!!!!!

Unless 'delpart.exe' has explicitly wiped the data area of the
partition, all the data will still be there. Use a disk sector
editor, and look around. After the format and before the install.

Earlier in this discussion it was suggested that a low level format, unlike fdisk and format, would actually wipe everything from a disk. Anyone tried that?

Keu


--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Tue, 29 Jul 2003 07:23:29 GMT, "Marc R. Kooij" <mr.kooij@chello.nl>
wrote:

<snip>

Earlier in this discussion it was suggested that a low level format, unlike >fdisk and format, would actually wipe everything from a disk. Anyone tried >that?

Almost any SCSI or IDE drive made in the last 6 or 8 years should
ignore you, if you try it. Some drives may honor it, by simply wiping
the disk. You would have to consult the tech specs for the drive, if
you can find them. But, I am not going to try it on any of my drives.

I believe that most modern drives use imbedded servo info (the servo
info is in between the data sectors) so that any true low level format
might zap the servo info, and thus the drive becomes junk.

Same applies to degaussing a drive. Wipes the servo info, now junk.

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

Earlier in this discussion it was suggested that a low level format,

unlike

fdisk and format, would actually wipe everything from a disk. Anyone

tried

that?

Almost any SCSI or IDE drive made in the last 6 or 8 years should
ignore you, if you try it. Some drives may honor it, by simply wiping
the disk. You would have to consult the tech specs for the drive, if
you can find them. But, I am not going to try it on any of my drives.

No, I suppose not.
I thought a feature like 'erase disk' (as it is called in disk manager for
IBM drives) is a low-level format utility, writing zeroes to the drive. From what you're saying here a low-level format is quite a different process then
I had in mind.
Still would like to know if this 'writing-zeroes' utility can wipe a disk (without leaving it unusable).

I believe that most modern drives use imbedded servo info (the servo
info is in between the data sectors) so that any true low level format
might zap the servo info, and thus the drive becomes junk.

Same applies to degaussing a drive. Wipes the servo info, now junk.

--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

In alt.msdos Marc R. Kooij <mr.kooij@chello.nl> wrote:
: Earlier in this discussion it was suggested that a low level
: format, unlike fdisk and format, would actually wipe everything
: from a disk. Anyone tried that?

Yes, back when hard drives were small, like 5 meg
to 20 meg (I said meg, not gigs).
But a chart of the factory bad sectors was on
the back of the drive, or a separate sheet of paper,
and they had to be entered manually in some cases.

Once the larger drives started coming out, format
gave up all attempts to write to data areas, it simply
took too long.

Spinrite could take hours just to format and
check a 40 meg drive, which would translate into
thousands of hours for a 40 gig drive today.


I have heard of at least one case where hard
drives were about to be sold as used from a state
agency when it was discovered they had critical
privacy data on them about patient health and such.

I have bought a number of used drives, most
have been fdisked and reformatted, and definitely
contain nothing worth the trouble of using a disk
editor to read.

But I did get a drive that had the original
operating system on it registered to a major public
facility, with some office correspondence on it,
which vanished when I fdisked it (for all practical
purposes).

When I had critical data on disk, the only
computer I had was a TRS-80 Model I, which was
one of very few computers costing less than $6000.

And it only had 5.25" floppy disks, single
sided (the original IBM PC had 5.25" floppies,
often one double sided and one single sided drive).
I encrypted the data with a system that could
not be broken in many man years of work, and then
slit the top of the holder and reversed the rotating
part of the disk.
And then when I was through with the files
(in 1988), I used a scissors on the disks.

I no files are ever erased on a DOS drive,
it is simple to replace the data files with all
zeros, just use debug to create a file with all
zeros, and write it to the file name needed
destroyed.

Erasing a file before doing this makes
it difficult to do anything simple to destroy
data files.

Joe Fischer

--
3
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc


"Marc R. Kooij" <mr.kooij@chello.nl> wrote:

Earlier in this discussion it was suggested that a low level format,

unlike

fdisk and format, would actually wipe everything from a disk. Anyone tried that?

Sure. But as mentioned, it should be specifically for that brand of HD.
Using another is taking a chance of turning it HD into a door stop. Check
the HD mfg. website for downloads. Some call it different names, such as LLF
or maxblast or recertification utility or something else. Some require you
to run their program before a warranty claim will be honored.

Note, some LLF utilities include wipe only and reformat features, but the actual LLF process takes a long time because it vigorously tests each sector for reliability, hence, the reason some call it recertification.

--
Todd Vargo (body of message must contain my name to reply by email)



--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

"arargh307NOSPAM@NOW.AT.arargh.com" typed:

BTW, someone asked a question that maybe you can confirm. When you used the >/U switch, was the 'Are you sure' type message displayed? I seem to recall >/U did not suppress this message as was claimed. ;-)

Yes, it always asked. I don't think that there is any way to suppress
that question. If there is, it is not documented.

I believe the _/Y_ or _/y_ switch makes the command with which it is
issued non-interactive in that it doesn't ask for confirmation any
more.


--
Ayaz Ahmed Khan

Yours Forever in,
Cyberspace.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

IN the olden days - when disk sizes were measured in Mby - not Gby -
there used to be a DOS utility - supplied with all versions I think -
probably upto DOS 5 - called PREFOR.COM - I still have q copy somewhere
- this was used to do the low-level format - and took along time. Its
what the manufactureres now do before they sell their disks. You could
do your own if reqd. This is BEFORE the FDISK operation which itself
precedes FORAMT C:
PREFOR.COM wiped the disk clean - everything went...

--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

On Wed, 30 Jul 2003 13:05:20 +0500, Ayaz Ahmed Khan <ayaz.a.khan@floccinaucinihilipilification.org.uk> wrote:

"arargh307NOSPAM@NOW.AT.arargh.com" typed:

BTW, someone asked a question that maybe you can confirm. When you used the >> >/U switch, was the 'Are you sure' type message displayed? I seem to recall >> >/U did not suppress this message as was claimed. ;-)

Yes, it always asked. I don't think that there is any way to suppress
that question. If there is, it is not documented.

I believe the _/Y_ or _/y_ switch makes the command with which it is
issued non-interactive in that it doesn't ask for confirmation any
more.

Nope. Not in DOS 6 or DOS 7. That does work for the copy command.

--
Arargh307 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html

To reply by email, remove the garbage from the reply address.
--- Synchronet 3.18b-Win32 NewsLink 1.113

From Newsgroup: comp.os.ms-windows.misc

Joe Fischer wrote:

In alt.msdos

Richard Bonner <ak621@chebucto.ns.ca> wrote:
: Joe Fischer wrote:

In alt.msdos
Check the size of the file, (do not delete) then
copy a bigger file to that file name.

:

It's gone.
Joe Fischer

:
: *** Not necessarily. COPY will not always place a new file directly over : the old. As well, copying a larger file will not necessarily erase all
: parts of some previous files because of the way a file is written to the
: disc.

I think it will if no other files are written
first.

*** To my knowledge, there is still no guarantee it will be written to
the exact same loaction on the disc.

: So, copying a different, but same-size file to the hard drive might end
: up in the same allocation unit, but there is no guarantee it will and
: no guarantee it will directly overwrite all of the previous file.

I suggested a bigger file.

*** I tend to think that may be even worse because the larger file may
require more clusters.


(Re: Drive Partitioning)

I don't any more, because I use multiple
drives, and I don't want to have to think about
how DOS assigns drive letters.

*** It is quirky, but if only one drive is used, the drive letters are
more logically assigned.

If a bigger file is written to the same
filename it should fill all prior allocation units
and then start additional ones.

*** But is that the case in reality? Does DOS always start at the first cluster and go forward?

When I had high security files I experimented,
I want to be sure, and not take anybody's word for it.

Joe Fischer

*** Did your experiments confirm that new files are always writen to old clusters first?

Richard Bonner
http://www.chebucto.ns.ca/~ak621/DOS
--- Synchronet 3.18b-Win32 NewsLink 1.113