[net.micro.pc] Reading/writing 360k diskettes on an AT

madd@bucsb.bu.edu.UUCP (Jim Frost) (10/28/86)

In article <1986Oct26.184138.24015@utcsstat.uucp> spence@utcsstat.UUCP
writes:
>I'd appreciate hearing from anyone who has used CPYAT2PC (or any similar
>program).  I use an AT as well as several PCs.  The AT has a 1.2 Mb drive 
>in addition to its hard disk.  I don't want to add a 360K drive, but I do
>want to be able to write 360K diskettes as well as read them.

>Does CPYAT2PC really do the job?  Are there other solutions? 

>Ian Spence {allegra,decvax,ihnp4,seismo,watmath}!utcs!utcsstat!spence

Um, I've been using an AT for two years or so now, and here are my
experiences:

Any 360k diskette written by any reasonably compatible computer will
operate correctly (ie read and write).  Regardless of what IBM says
about diskettes being readable only on AT's after an AT has written
on them, this hasn't been my experience.  I have used both newer and
older (but not first issue, I think) AT's.

You can force the AT to format a 360k diskette by specifying

FORMAT A:/4

I have used this extensively.  It works fine.

Now, I had people tell me that once a 1.2Mb floppy drive writes on
a 360k diskette, it would be unreadable by 360k drives.  I also found
this in one of IBM's manuals.  Like I arleady said, I haven't had a
single problem with reading/writing them.  I have ported 360k diskettes
between an AT and:

   IBM PC
   IBM XT (same thing, I know, but...)
   Tandy 1000
   Tandy 2000
   THE PC+ (your basic real cheap PCompatible)

The point of all this is that you don't NEED a special copy program.
If you want to copy from a 1.2Mb to a 360k (or vice versa),
the COPY command works fine.  If you only have one drive, it still
doesn't matter since PC/MS-DOS checks the formatting before writing.
You can also use the DISKCOPY program, provided the diskettes are
formatted by the same version format program.

Note:  I only used DOS 3.1 on the AT's.  Maybe they fixed something in
       in the routines that determine disk type; I don't think so though.
-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
                   - Jim Frost * The Madd Hacker -
UUCP:  ..!harvard!bu-cs!bucsb!madd | ARPANET: madd@bucsb.bu.edu
CSNET: madd%bucsb@bu-cs            | BITNET:  cscc71c@bostonu
-----------------------------------+-----------+------------------------
"Use the key, unlock the door                  |    o/ <- Rudolf the
 See what Fate might have in store." -- Rush   |   _O_    waving penguin

plocher@puff.wisc.edu (John Plocher) (10/28/86)

In some article Jim Frost replies to:
> spence@utcsstat.UUCP who asks:
>>           I use an AT as well as several PCs.  The AT has a 1.2 Mb drive 
>>in addition to its hard disk.  I don't want to add a 360K drive, but I do
>>want to be able to write 360K diskettes as well as read them.
>
>Any 360k diskette written by any reasonably compatible computer will
>operate correctly (ie read and write).
                                ^^^^^
 Not really!  A small illustration:

     [###]    Width of head on 360K drive
      [*]     Width of head on 1.2 Meg drive
      [ ]     Gaps between cylinders of data on a disk

1)      ### ### ### ###    cylinders of data on 360K disk
	### ### ### ###    (a top down look onto the disk surface)
	### ### ### ###    ( <- to outer edge: ### ### ### :to inner hub -> )

2)       * * * * * * *     cylinders of data on a 1.2 Meg disk
	 * * * * * * *
	 * * * * * * *

3)       *   *   *   *	   cylinders written on a BLANK 360K disk by 1.2M drive
         *   *   *   *     (Note that every OTHER cyl is used)
	 *   *   *   *

4)      #*# #*# #*# #*#    cylinders of 360K data overwritten by 1.2M drive
	#*# #*# #*# #*#    (note that not all the 360K data is erased...)
	#*# #*# #*# #*#

>                                        Regardless of what IBM says
>about diskettes being readable only on AT's after an AT has written
>on them, this hasn't been my experience.
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    Most people will not have problems when swapping disks between drives
    because the data on the disk is never overwritten (I'm talking
    about DOS reusing clusters on the free list, NOT about updating
    a file with an editor...)  Partially filling a disk with 360K stuff and
    then adding some 1.2M stuff and reading it all back on a 360K drive 
    should work.  In that case, you only cover cases 1 and 3 above.

    BUT if you write 360K, erase some of it, then write some new stuff
    on it with the 1.2M drive, you can get case 4 above.  Since
    the head in the 1.2M drive is smaller than that in the 360K drive,
    it writes a smaller path of data on the disk.  Unfortunately, it is
    right down the middle of the existing 360K swath, and not all the
    360K swath is rewritten.  Still no problem on a 1.2M drive, since
    it NEVER sees the overlap (the #'s in case 4).  But if you take this
    disk back to a 360K system, the drive sees the old 360K data mixed in
    with the new 1.2M data and doesn't know what it's seeing!

    Also, a point to remember is that the data swath on the 360K disk is
    that wide for a reason:  The 360K drives *need* that much amplitude
    so that the flux transitions can RELIABLY be recovered from the disk
    surface.  The swath laid down by the 1.2M drive may not have enough
    amplitude for some 360K floppies to read!

>You can force the AT to format a 360k diskette by specifying
>FORMAT A:/4
>I have used this extensively.  It works fine.

    The format /4 command produces case 3 above; if the
    disk was orig used on a 360K system, case 4 might
    come into play.  Note:  IBM MIGHT be formatting as case 2
    to really erase the disk, then redoing as case 3 to
    generate proper 360K format...

>Now, I had people tell me that once a 1.2Mb floppy drive writes on
>a 360k diskette, it would be unreadable by 360k drives.  I also found
                     ^^^^^
>this in one of IBM's manuals.

    A better word is 'might'.  IBM is saying 'would' to keep people
    from coming back to them with  "I did it the way that you said
    should work, and now it doesn't work..."

>The point of all this is that you don't NEED a special copy program.
>... the COPY command works fine.

    Reading a 360K disk on a 1.2M drive will NEVER give problems.
    Reading on a 360K drive:
    - Using a disk only used in a 360K drive will ALWAYS work :-)
    - Using a disk formatted on a 1.2M drive and only written to
      by a 1.2M drive should give you no problems.
		      ^^^^^^
    - Using a disk formatted on a 360K drive and written to by 
      360K and 1.2M drives (but without files being erased or modified)
      should still be no problem.
      ^^^^^^
    - Using a 'working disk' (edit/compile/edit...) interchangably
      between 360K and 1.2M systems is bound to cause problems
      when you can least afford them!  ^^^^^^^^^^^^^^^^^^^^^^^

    Hope this clears up some of the confusion...
-- 
"Never trust an idea you get sitting down" - Nietzche
------------	{harvard,seismo}!uwvax!uwmacc!uwhsms!plocher        (work)
John Plocher    {harvard,seismo}!uwvax!puff!plocher                 (school)
------------	decvax!encore!vaxine!spark!121!0!John_Plocher       (FidoNet)

smvorkoetter@watmum.UUCP (Stefan M. Vorkoetter) (10/28/86)

Simply adding files to a 360K disk with a 1.2M drive (assuming the disk
was originally written with a 360K drive) can still cause problems.  Even
if the drive writes on a previously unused part of the disk, there is still
information stored in those sectors (such as the CRC) that gets changed.
In addition to this, the directory gets rewritten.

psfales@ihlpl.UUCP (Peter Fales) (10/28/86)

Subject: Re: Reading/writing 360k diskettes on an AT (long)
Newsgroups: net.micro.pc
References: <564@bucsb.bu.edu.UUCP> <273@puff.wisc.edu>

(Long article about 360K diskettes on a 1.2M AT drive deletetd)

Just to add another data point to the confusion, I have had almost no
problems swapping 360K disks between an AT&T PC-6300 and 6300+ (which
use the 360K and 1.2M drives respectively).  This is independant of
whether the diskette previously contained data written on either drive.
I have seen only a single case where data added to a 360K diskette by
a 1.2M  drive was unrecoverable on a 360K drive.

The author of the above article states

>     Most people will not have problems when swapping disks between drives
>     because the data on the disk is never overwritten (I'm talking
>     about DOS reusing clusters on the free list, NOT about updating
>     a file with an editor...)  Partially filling a disk with 360K stuff and
>     then adding some 1.2M stuff and reading it all back on a 360K drive 
>     should work.  In that case, you only cover cases 1 and 3 above.
> 

This is not correct because adding a file to the disk using a 1.2M drive
will still result in changes to the FAT and directory sectors which were
originally written with a 360K drive.  But as I said, I have had no
problems doing this.  However, for all the reasons given in this and
other articles DO NOT USE THIS METHOD WITH CRITICAL DATA.  As an easy
method of transferring data between machines, fine.  Keep in mind that
data on any part or all of the disk may become unreadable, so valuable
diskettes should not be swapped between the two types of drivess.

Peter Fales
ihnp4!ihlpl!psfales

burton@parcvax.Xerox.COM (Philip M. Burton) (10/29/86)

**

In all the discussions on this topic, no one has stated WHY it is necessary
to go trhough this kludge, which is admittedly risky.  Unless someone is 
using two full high hard disk drives, or perhaps a half-high tape backup
unit, in the AT, there is room for a second 1/2 high floppy drive.  At today's
prices, a half-high 360 KB floppy drive is about $100, with do-it-yourself
installation.  The cabling is already in the machine.  

Two installation notes.  

1.  Remove the termination pack from the floppy, and jumper the floppy as
drive 2 <------

2.  If the drive is not "AT compatible", you can make it AT-compatible
by using some tape to cover line 34 on the 34-line interface edge.

With this pricing, there's no reason to risk any valuable work to luck.

Phil Burton
Xerox Corp.

[no disclaimers ]