11 Apr
2007
11 Apr
'07
7:40 p.m.
Dave Dunfield wrote (in reply to me):
Lee Davison studied the pattern and showed pretty clearly it was a
mostly repetitive pattern, other than track/sector counters.
...
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 agree with Chuck ... this is probably as test pattern. Dave
PS: Thanks for the N* TurboDos images, I've just put them up on the site,
along with several other N* images that I have had pending.
Once it was determined that the disk had nothing useful, I booted HDOS
from a floppy, reformatted the drive, then went through
the turbodos
install procedure. It installed with no hitches, and now I have a
multiprocessor Z80 machine with a 30 MB hard drive.
The install disks I used were the same ones I imaged and sent to you,
Dave, so now we know they are good.
I have no real use for this machine and no ideas for tinkering with it,
so for now it will go back on the shelf until I think of some project
that is interesting enough to displace other projects I have in the
pipeline.
One project is to take my IMS 5000 system, hook up a logic analyzer to
the disk controller, and recover the login/password pairs from the disk.
This system uses Turbodos too, and from reading the NS turbodos
manual, I can confirm what someone else recalled, which is that this
information is stored as plaintext in a simple file.