Welcome to the Prefix wars [WAS:RE: Weekly Classic Computer Trivia Question (20141205)]
t.gardner at computer.org
Mon Dec 8 20:27:37 CST 2014
>On Sunday, December 07, 2014 5:10 PM Fred Cisin
[mailto:cisin at xenosoft.com] wrote
>On Sun, 7 Dec 2014, Chuck Guzis wrote:
>> Oh, it started before the Mac. Consider, for example, the Technical
>> Reference for the IBM PC (first edition). Clear as day, it talks
> >about memory expansion options of "256K". Even so, consider older
> >CP/M machines--many of them used the memory available in the signon as
>I remember some booth staff at West Coast Computer Faire (1979?) trying to
tell me that their computer with 65.5K of RAM was better than the
>It was certainly well established by the time of TRS80/Apple][/Pet.
(4K,16K,48K, . . . )
>> It probably goes back well into the 60s talking about so many "kilowords"
of memory or mass storage.
Thanks for the comments but I still content the mess really started with
Apple's Macintosh OS which REPORTED MEMORY AND DISK CAPACITY USING K in a
binary sense without any qualification. There is no doubt that K and M were
used in a binary sense before Macintosh, but AFAIK not by any OS for disk
MSDOS and PCDOS AFAIK did not use any prefixes - the utilities just reported
the values in a string of decimal digits without commas. Interesting but
not surprising that CPM used binary K on memory but I doubt it reported on
disk capacity using any prefixes, but I could be wrong. Don't know about
TRS80 and PET but I doubt the OSes used prefixes of any sort.
The real mess started when an OS reported a disk drive advertised in
conventional MB using binary MB e.g. an ST225 advertised by Seagate as
having 21.4 million bytes but reported by an OS as having a 20910 KB
capacity (41,820 512 byte sectors per Seagate spec = 21.411.840 bytes = 21.4
MB). Where is the missing 490 KB? Macintosh OS System 1 is the first OS
that I know about.
More information about the cctalk