On Oct 17, 6:45*am, klem kedidelhopper
wrote:
On Oct 16, 8:22*pm, "William Sommerwerck"

wrote:

I was going to keep my nose out of this, but I'm so thoroughly
confused that
I have to butt in.
What are you trying to do? Test the drives? If so, this seems
about the
worst possible way. It would make more sense to set these drives
as slaves
and boot them on a known-good computer (as someone else
suggested). Even
better, if they're IDE-ATA drives, you can mount them in a box
with a USB
interface, and simply attach them to a running machine.

What is the question you're asking? I suspect it's the wrong one.

William
Thank you for "butting in". I appreciate everyone's input, yours
included. I thought that I was being very concise. However I have
to
ask seriously did you read the OP? And if I did follow what you
just
suggested how would that have prevented a CMOS virus, if that's
what
this problem is? But to address another question I have why would
these engineers in their infinite wisdom write a bios to anything
other than a non writable eprom. It would seem like some things
should
be sacred. Lenny

What William suggests is something I and many others have done many
times without problems. I saw a computer (not mine) that was so
infected the processor was 100% busy and would do nothing useful,
including running a virus scan. Of course they needed to salvage the
apps and data and couldn't just re-format and start over. The drive
was first slaved into a good machine and subjected to a virus scan.
Remember those pesky viruses have to execute to become active. The
slave drive executes nothing during boot so activates nothing (unless
the boot drive has its own viruses). That's when I became convinced
about Norton utilities. Norton wouldn't dump a virus because it was
running. AHAH, I'll boot into safe mode and kill it before it's
running. Norton (at least that version) will not run in safe mode. I
don't have Norton.

As far as booting from unknown drives, I wouldn't even attempt that.
The OS installation gets tweaked during install for the hardware on
the machine. Who knows what the original hardware was?

I also had a machine with a BIOS bug that was fixed with an update. It
was FLASH so I didn't have to get a new BIOS chip and install it
though one time I didn't follow the procedure EXACTLY and corrupted
the BIOS. That machine was old enough to still have a socketed chip
and it 'only' took $30 and a few days wait. My current Gigabyte boards
all have dual flash BIOS chips. If you foul up a BIOS update (and I
did THAT once too) it defaults back to the known good one and boots
back up and yes, you can copy the current BIOS to the backup. The
Gigabyte boards will now update the BIOS online while running Windows.
That is about the easiest. I would think that a BIOS virus would be
harder to write as it would be specific to a board model. Or do all
the BIOS writes behave identically?

G²