7 Oct
2005
7 Oct
'05
9:46 a.m.
Subject: 8" floppy system needed to recover old game data
From: "Barry Watzman" <Watzman at neo.rr.com>
Date: Fri, 07 Oct 2005 11:46:29 -0400
To: <cctech at classiccmp.org>
Let's simplify this ... A LOT.
The most common disk format (BY FAR) for CP/M was 8-inch, single sided,
single density.
Already stated but it's very important to repeat it.
If they are in that format, then virtually any CP/M
system with 8" floppy
disk drives can read the disks (assuming that they are still readable). It
is very unlikely to matter what disk controller was used to write the disks,
it is very unlikely to matter what system originally created the disks. The
8" SSSD format is "gold standard interchangeable" among virtually all CP/M
systems with 8" floppy drives. And it was, BY FAR, the most common CP/M
disk format.
Most 8" softsector systems that did "other formats" also maintained
the 8" SSSD standard.
Finding a working CP/M system with 8" drives
should not be difficult. I
have 3 of them here (I'm in Ohio, USA). There are still many thousands of
running CP/M systems around. A slightly greater problem is getting the
files moved from the 8" CP/M floppy disks to another [more useful] form of
media. What you really want is a system with both 8" and 5.25" floppy
drives, and a PC with 5.25" floppy drives. Such a system would allow you to
move the files from 8" CP/M to DOS/Windows, which is probably the objective.
The latter, moving the files is trivial. Procomm on a PC and any of the
modem programs on the CP/M system is easily done. Some systems like mine
runs 3.5" floppies as well plus a program called dos to write 720k format
DOS floppies which is very handy! Other do it by writing 8" to 360k 5.25
and relying on a pc with same disk and using one of many programs to read
the disk at the file level.
My experience in reading 8" diskettes written in
the 1970's has been
excellent. I have hundreds, probably a couple of thousand such disks, and I
have only had a single digit number of diskettes that I have been unable to
read out of all of those. Others who have actually had experience with
large numbers of 8" disks also report very high success rates with disks
that have been stored in climate-controlled environments. Now if the disks
were outside in a garage for 20 years, that would be another matter of
course, but for well stored disks, readability is not usually an issue.
My experience is mixed, I generally assume readable but occasional bad media
show up. I've had more likely due to poor storage they were subjected to
beforehand. A side note is that most of the bad media was mid 80s 3M
manufacture [remember the "blackwatch" audio tape that binder crapped after
a few years] or poorly stored. My PDP-11 (RX01) disks are typically 35
years old for some masters. Yet some of the early 80s cheap (third party)
8" media has failed. Maxell Dysan and BASF were always the best and I still
have an unopened box or two.
It is possible that the disks are not SSSD (single
sided, single density),
and if that is the case, then the problem becomes far more complex. Once
you get away from 8" SSSD, there is no standard disk format, rather there
are hundreds of different formats specific to each implementation of CP/M.
This could [greatly] complicate recovery, however, under no circumstances
would this make recovery impossible. CP/M is highly adaptable, and with
some effort (programming) it would be possible, first, to determine the
format, and, second, to adapt any CP/M system to be able to read any disks
for which it had the correct type of drive.
The only other format that shows up is Intel MDS using the double density
M2FM(oddball), I must have a dozen of these that I can't read. I point
this out as some games developers used MDS as their development platform
and debug platoform using the MDS ICE. Usually thier budget was smaller
and they used generic systems.
In general, unless the format is ***VERY***
non-standard, it doesn't matter
what the controller was that wrote the disks. Most systems used Western
Digital 17xx controllers, some did use NEC 765 controllers (which existed
for a couple of years before the PC came out), and a few used discreet logic
controllers. But for the most part (and in particular between the WD and
NEC based controllers), disks were interchangeable with regard to being
READ.
FYI: the NEC 765 was sampled in early 1980, production 1980 4thQ, IBM PC was
Q3 1981. It wasnt a few years. The 1771/1793 was the defacto CP/M FDC
through 1981 and dominent after that with only new designs after '81 picking
up on the 765 (compupro and some of the system on a single board vendors).
In any case for 8" SSSD it was not an issue as both NEC and WD controllers
were fully capable and compatable.
A few people mentioned hard-sectored media. While hard
sector 8" disk
systems did exist (I have two of them here), it was very, very rare for
hard-sectored media or systems to run CP/M. Invariably, systems using
hard-sector 8" media ran proprietary operating systems. You can assume that
any 8" CP/M system was soft-sectored and be correct at about the 99% level.
Hard sector 8" system were rare, unusual and many of the host systems
were difficult to program (new bios) or unsuitable for CP/M operation.
Another watch for thing. I have media (Maxel 2Dxd, two sided) that was
usually used for two sided and the sector hole is different and some older
drives do not see index as a result. This occured to me using an old
SA800 and the media was actually used and formatted as as single sided
SD on a Compupro Disk1 using QUME drives. It was a simple matter of using
an 8" drive that could sense index in either location (sa860). I mention
this as users and integrators of CP/M systems often did an occasional
strange thing and if a disk doesnt read it can be an innocent error thats
easily fixed.
As been repeated else where and by myself the 5.25" world was not as orderly.
Allison