Re: Converting from 8 to 3.5 inch floppy diskettes

Herbert Johnson <herbrjohnson@xxxxxxxxx> wrote:

Glen, I was going to reply to your posts about "standards" and
"conventions" and so forth, but it would take a lot of explaining to
say why there were MANY "standards" and "conventions" in use, to read
write and format 8-inch disks, during the 1970's and 80's when they
were in use on CP/M systems.

Well, there are two separate questions. One is the low-level
format of the blocks on the disk. For that there are the IBM
SD and DD standards, used by the Intel and WD floppy controller
chips among others. Apple and Commodore used non-IBM formats.
DEC used some IBM and some non-IBM low level formats.

You've gone on to post about other
standards, yourself. I think talking about "standards", as though
there was one way to support 8-inch floppy drives, is confusing. This
may not be your intention, Glen.

Now, separate from the low level format is the high level,
or file system format. I include in that sector interleave,
sector size, sectors/track, tracks/cylinder, track ordering.

I do like the DEC system of using separate programs for low
level and high level formatting.

It's funny to have to argue the point, as it was an obvious issue in
the era, and anyone who worked with such systems back then ALWAYS had
to be aware of that. Any transfer of data between two systems by
diskette involved figuring out the format and schemes in use. Everyone
had DU, or programs by Sybex or others, to do "disk conversions" or
"disk analysis". Much the same occurred in the 5.25-inch drive era.
The greatest difficulty was transferring between "hard sectored"
systems and "soft sectored" systems. I think some of the work Mike S
is referring to, is about systems with hard sectored floppy
controllers - even less "standard" although common enough then and
still available now.

As far as this discussion goes, 8 inch drives and disks are
pretty rare, and expensive when you find them. For every day
use, 3.5in disks and drives are much easier and cheaper to find.

Anyway, it's more confusing to argue about the lack of standards and
to describe a decade's plus history of floppy disk controllers. It's
simpler to say there were many standards, which is to say that amounts
to NO standards. For specific systems, it's all in the details - which
"standards" those systems IN PARTICULAR followed. Glen, at that point
your clear knowledge of methods and designs is very useful.

IBM did a pretty good job of documenting their standards.
Apple and Commodore did not.

This goes to the other point, raised by Barry Watzman. He seemed to
suggest that if you could get some kind of diskette scheme working on
a system, in some format however unique, that's fine. I disagree, for
the same circumstances as I've mentioned above. How do you read your
"unique" disks on someone else's system? And you *will*, or someone
else will, someday. How many posts have there been in comp.os.cpm
which amount to "...so I have these disks from X years ago, that I
can't read. Please help me"? Again, a common enough issue in the 80's
and later, still an issue today in this newsgroup.

Yes, good point. To me, if you use 3.5in or 5.25in disks as
direct replacements for 8in disks, with a similar low level
and high level format, you are as close as you can be to the
old formats, yet the hardware to read/write will be easier to
find in the future.

A technical reason for some kind of "standards", is to avoid writing
data under marginal magnetic media conditions. To be brief about it:
if you cram too many data bits onto a small circle of a track on a
diskette, they may not be reliable. Or they may be OK today, or not
tomorrow. Or one drive may read them, but another may not. So, any
scheme of drive conversion has to work within standards of disk data
density, bits per inch. There are people who claim they can "beat"
such standards - but why risk losing data? 'Nuff said?

Well, marginal magnetics is one. The appropriate gap sizes
are needed to allow for speed tolerance on the drives. IBM
was pretty conservative, others shrunk the gaps to get more
data on the track. That is separate from marginal magnetics.

In any event, general discussions miss a key point of this discussion
thread. By finding 3.5" drives which can rotate at 360RPM, the same
rotation rate as 8-inch drives, you can AVOID MANY PROBLEMS.

But it doesn't solve the most important problem, which is
availability of drives. There are enough 3.5in drives in
the world now that a reasonable number of them should be around
for a long time. That doesn't apply to 3.5in 360RPM drives.

(My 5.25in drives always had a speed adjustment and strobe
pattern to be viewed under 60Hz arc lamps. I don't know that
3.5in drives have that.)

Use high-
density diskettes, and the 8-inch data rates will be OK for that media
- briefly put. And, the old computer's hardware and software will not
"see" a drive which is rotating too "slowly", wont' see a track that
is too "long", an index pulse rate too slow, etc. etc. That means you
don't have to change controller hardware or software.

With 300 RPM drives, the only difference is the time between
index pulses. If that is filled with unused (by the high level
format) sectors or gaps the disk is low level compatible with
other hardware. The bit timing and gap lengths of a 360RPM
disk are likely way too far off to be read/written (especially
written) at 300RPM.

The index pulse is used on formatting to know when to start/stop
writing format data, and in read/write to time out access to
a sector that doesn't exist. I believe that 300RPM is a better
choice for 3.5in disks emulating 8in.

-- glen
.

Relevant Pages

  • Re: Mac / DVD-RAM / FAT 32 Question
    ... My Macbook will not format or write to DVD-RAM disks even though the "super" ... use in our outboard drives. ...
    (rec.arts.movies.production.sound)
  • Re: DVD-R/DVD+R...whats the difference
    ... A corollary to that is don't use packet writing; burn only "ISO" format ... Fair comment Barry, but in reality, the RW disks were only ever meant for ... though X10 drives are available). ...
    (comp.sys.laptops)
  • Re: REQ: OLD SCSI driver for Panasonic LS3700 optical drive
    ... these drives on the net, and no OEM drivers to be found anywhere. ... Trying to help someone out who needs to retrieve old lab data off disks. ... thing and has nothing to do with the SCSI driver. ... What format was ...
    (comp.periphs.scsi)
  • Re: Any way to uninstall Operating system
    ... You can boot with your Win98 startup disk and issue the command ... The set of four install disks can be created ... | I have one hard disc with three drives namely C:,D:,E:. ... Is it good practice to format before deleting not useful data. ...
    (microsoft.public.win2000.general)
  • Re: conflict of software Nero Recordnow Drivers - or ??
    ... disk package anywhere. ... If Not when Would I format a disk? ... For most normal disc burning projects, music CDs, playable DVDs, Photo CDs, ... I know I did have DLA, On both drives and I was using when asked if I ...
    (microsoft.public.windowsxp.hardware)