R: Re: 3.5 floppy imaging

[Fred Cisin]

On Sun, 17 May 2015, supervinx wrote:
> The only reason I guess is that they're bootable disks. Otherwise a 
> simple filecopy is more than enough :)

If they are bootable DOS disks (and normal, not weird), then there is one 
part that won't come across with a filecopy.  That is Track 0, side A, 
sector 1. If they are some version of DOS for which you don't have a copy 
of the boot-sector, that might be worth saving. Using DEBUG, you can copy 
that into RAM, and save it as a 512 byte file.

DEBUG
L 100 0 0 1        ; load sector from A:    L 100 1 0 1 for B:
R CX               ; set byte count
200                ; 512
N BOOT.DAT         ; filename
W                  ; write
Q                  ; quit out of DEBUG  "most important command of DEBUG"
            NOTE: do NOT type in the comments!

There were a few boot-sector viruses, such as Alameda/Yale, Stoned, and a 
Stoned derivative that was passed off as Michelangelo.   Copying the 
sector using DEBUG does not expose your machine, but attempting to boot 
with them does.


BTW, if they are bootable, then there will be two hidden files:
IO.SYS & MSDOS.SYS for MS-DOS, IBMBIO.COM & IBMDOS.COM for PC-DOS.
When recreating the disks, those must be the first files on the disk,
in the first contiguous sectors.
Plus COMMAND.COM. The filedate of COMMAND.COM is 
normally sufficient to identify the DOS version.

On PC-DOS and "generic" MS-DOS, 720K started with V3.20
But some OEMs started using 720K as early as 2.11, often uniquely 
non-standard  (Gavilan, etc.)
1.4M started with V3.30

NEC had their own version of an MS-DOS HD format.

and, there were some 3.5" CP/M machines.  THOSE should be imaged.


--
Grumpy Ol' Fred     		cisin at xenosoft.com