Continuing my story of trying to recover the contents of the hard drive
in my Horizon 8/16 system, I took Dave Dunfield's advice and wrote a
small program to read each sector from the hard drive and dump it. 30
MB of disk data at 9600 baud, printed as ASCII hex. It took a while.
The mystery remains -- what is it?
There is no plain-text to be found anywhere ... not even accidental
sequences. I thought that perhaps it is a lot of binary data. Some of
it must be code. Searching the hex for "CD0500" (which is "CALL
0005",
the CP/M call vector) has no hits.
I know the disk is good because I can write a sector and read back what
I wrote. The non-destructive drive tests (basically read each sector
and see if any errors crop up) pass.
Even stripping the msb produces blocks of data that look like the
following text. "." represents unprintable characters.
........."E..(P @.....'N.:tiS'N.
9rdI.&M.6mZ4hQ"D..#G..:tiS'O.<xp
aB... A....,Y3gO.>}ztiS&M.7n]:ti
R$I.%J.)R%K.,Y2eJ.)R$H.#F..4hP @
.....1bD..$I.&L.0aC...9reJ.(Q#F.
.7n\8p`@.....-[7o_>|ysfM.7n\8qcG
..<xqbE..(P!B...&M.4hQ#G..?~}{w
o^={vmZ5jT(Q"D..%K./^<xpaB...%J.
(P!C...;vlX0aB...+W.]:thP!C...5k
W/^<xqcF..5kW.];vmZ5jU+V,Y3fM.5k
W/_>|ysfM.6lX0aC...<yreJ.)S&M.4i
S&M.4hP!C...7n];vlY3fL.2dI.$I.$I
.%J.+V,X1cF..1bE..*T(P @........
..9sgO.={wn];wo^=zthP @.......>}
{vmZ4iS&L.2eJ.*U*T(P A....9rdH.!
B...%J.)S'O.?~}ztiS'O.>|xqbE..(P
This is a typical block, one of almost 60K.
You prefer hex? OK, here is another block:
; TRACK #031A, SECTOR #000C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 suppose it might be 8086 binary, as turbodos supported both Z80 and
8086 CPUs in the system, even mixed, but then again, my machine only has
Z80s, no 8086s.
Unless a genius comes along and recognizes this, I'm going to wipe the
drive, install TurboDOS, and not look back.
Finally, I have to give kudos to Dave Dunfield for his great NST
utilities and his floppy disk archive. It allowed me to mint some new
NSDOS and CP/M images for the machine after my only CP/M boot disk got
trashed.
http://www.classiccmp.org/dunfield/img/