Hi,

> Precisely.  Now I have to use a semi-magic invocation of dd(1) to do
> what I want, because there's a fairly arbitrary restriction in
> fdisk(8).  That's lame.  It's also distinctly *un*like what the DOS
> FDISK does -- and it's actually pretty broken, in that it assumes that
> the signature being present means there's valid boot code.

I can assure you that at least all versions of DOS FDISK that I tested
 _do_ assume that the signature being present means there's valid boot code.
You can initialize the boot code on such disks with the (undocumented to
the best of my knowledge) FDISK/MBR command, but that _does not_ change the
partition table!

Ciao,
Wolfgang
-- 
ws@TooLs.DE     (Wolfgang Solfrank, TooLs GmbH) 	+49-228-985800