9 Dec
2014
9 Dec
'14
6:14 p.m.
On Monday, December 08, 2014 8:12 PM Fred Cisin [mailto:cisin at xenosoft.com]
<snip>
In MS-DOS/PC-DOS 2.00 and above, FDISK reported hard
disk capacity in
honest binary Megabytes of 1,048,576
I looked for but could not find a screenshot of very early PCDOS FDISKs, the
later ones I've found all include the statement 1 MB = 1,048,576 bytes so by
that time Microsoft and IBM were indeed "honest" unlike Apple. FWIW, The
MS-DOS Encyclopedia (c) 1988 which covers thru DOS 3.3 shows the "Display
Partition Data" option as giving capacity in the number of cylinders, not a
particularly useful measurement but not misleading. FORMAT and CHKDSK all
report in decimal digits without commas. Would like to see a screenshot for
an early FDISK Option 1 using binary prefixes before I concede that
Microsoft/IBM DOS used but did not disclose binary prefixes before Apple
<snip>
WHAT missing 490 KB??!?
I recall Seagate referring to the ST225 as "20 Megabyte" (~1985?) Unlike
some other manufacturers, they rounded DOWN, not UP, AND they did not use
excessive "significant digits".
(That also gave them a lot more slack than their less
scrupulous
competitors to cover bad sectors) ((In apthecary measures, there are
THREE
Scruples in every DRAM!))
The published Seagate specification for ST225 that I have states the
formatted capacity = 21.4 MB comprising exactly 41,820 x 512 byte sectors =
21.411.840 bytes = 21.4 MB
BTW the reason I picked the ST225 as an example rather than the ST412 is
that Seagate specified the formatted ST412 with 32 x 256 byte sectors per
track giving precisely 10.0 MB (10.027008 MB). IBM in the PC chose to go
to 17 x 512 Bytes sectors per track thereby increasing the formatted
capacity to 10.7 MB or 10.4 MiB. So a Seagate ST412 used with any of the
several IBM PC compatible controllers actually had a dram more than Seagate
specified, but in both cases more than 10.0; confusing perhaps but no basis
for complaining. But they soon changed to 512 byte sectors and that dram
went away. I really don't think Seagate was any more scrupulous than its
competition in specifying its drives, as near as I can tell they rounded
off, but even if they rounded down it was in the third or higher digit so we
are talking about a relatively small number when compared to the confusion
caused by binary prefixes.
Tom