Date   : Wed, 07 Jun 1989 20:59:29 GMT
From   : amelia!pioneer.arc.nasa.gov!samlb@ames.arc.nasa.gov (Sam Bassett RCD)
Subject: CP/M versions

       In the beginning there was Gary Kildall, who was a professor at
the U.S. Naval Postgraduate School in Monterey, California, and working
on the side (the Navy doesn't pay spectacularly well) as a consultant for
Intel, writing software to make their chips useful.
       Dr. Kildall became mightily (and rightly) incensed with the brain-
damaged excuse for an operating system (ISIS) that Intel perpetrated on
their development systems.  He therefore went out and wrote a simple
program loader cum hardware interface that he could use to do the work
that Intel was paying him for -- write programs quickly and easily.
       This first incarnation was written in PL/M, Intel's quirky subset
of Big Blue's PL/1 -- which stood for "Programming Language / One", among
other things -- thus, "Programming Language/Microcomputer".  Intel had
just decided to call the 8080 a microcomputer (this was 1979 or 80).
       "Hmmm", he said, "three letter acronyms with a slash seem to be
the in thing, so I guess I'll call this 'Control Program/Microcomputer'!"
Thus "CP/M".
       I don't know what the _very_ first version was called, but the
first version to get any commercial distribution (on Tarbell magnetic
tape, if my memory serves me, was CP/M 1.4.  This was found to be a bit
primitive, and was followed by CP/M 2.0 and 2.2. 
       In the midst of all of all of this, Dr. Kildall found he needed a
company to keep the Internal Revenue (Inland Revenu to you Brits) busies
out of his personal pocketbook.  He therefore incorporated as
"Intergalactic Digital Research, Inc."  A couple of years later, when he
needed some outside funding, he was persuaded to drop the "intergalactic"
part, so as not to freak out bankers.  "Kentucky Fried Computers" was
also persusded to change their name at about this time to "North Star
Computers" by a gang of bankers & lawyers -- the Col. was not amused.
       Also, about this time, DRI introduced MP/M for the 8080/8085/Z80
-- an operating system that permitted limited multi-programming, in the
sense that several programs could run in a round-robin scheduling
fashion, and look like they were running simultaneiously.  Very much like
UNIX, but limited to 8-bit machines.  This OS became moderately
successful on high-end micro hardware, like Bill Godbout's CompuPro
boat-anchors, but never among hobbyists or the general user.
       The last 8080/8085/Z80 product from DRI was CP/M 3.0, aka CP/M+
or CP/M Plus -- which added bank-switching to the amentities of CP/M 2.2.
The dreaded Intel 808x entered the scene at this point.

       Naturally, the leading (then) purveyor of OSs for micros could
not be left out in the cold, and a good deal of money and personnel were
thrown at re-writing CP/M for the 8086 & 8088 chips.  Therefrom issued,
in good time, CP/M-86 (CP/M for the 80x86 family).  The good old cash cow
was then renamed CP/M-80.
       While all of this research was going on (and eating money), a
couple of smart technical types in Seattle WA took the sources to CP/M
1.4 (which they or someone had bought) and ran them through an early 8086
Assembler, and came up with "Seattle DOS" or DOS86.
       There are a number of apocryphal stories about why IBM did not
sign up with DRI to provide the OS for the PC, and Dr. Kildall has told a
couple of the versions himself.  The one I like best was that Dr, Kildall
was out of town, and a couple of goons in dark suits showed up at the
offices in Pacific Grove, and demanded to see Kildall instantly.  they
were so rude and uncivilized that they were told to pack it in and take a
hike.  They did -- right to the airport and on to Seattle, to sign up
with Billy Gates' Microsoft (who went out and bought the rights to
Seattle DOS, changed its name to MS-DOS, and made pots of money).

       The research bore fruit, and CP/M-86 was followed by Concurrent
CP/M-86 (renamed from MP/M-86), GEM, DRI-DOS, and several other packages
and applications (including Ventura Publisher, done by DRI defectors,
reportedly because DRI would not fund it . . .).

- - - - - -

       There are as many varieties of CP/M-80 as there were hardware
companies or hackers who were willing to port it to their boxes, because
what DRI sold was the Basic Disk Operating System (BDOS) and the Console
Command Processor (CCP) -- each developer has to create a custom Basic
I/O System (BIOS) from hints in the DRI documentation (the less said about
which, the better!)  In other words, talking to the hardware was _your_
problem, said DRI, we provide the high-level stuff.
       MS/PC-DOS provided a set of device drivers for the standard PC
hardware (CP/M hardware was in _NO WAY_ standard), and the ability to
load custom dirivers -- vi CONFIG.SYS.  {UNIX hasn't learned that lesson,
yet!!!}


Sam'l Bassett, Sterling Software @ NASA Ames Research Center, 
Moffett Field CA 94035 Work: (415) 694-4792;  Home: (415) 969-2644
samlb%well@lll-crg.ARPA                 samlb@pioneer.arc.nasa.gov 
<Standard Disclaimer> := 'Sterling doesn't _have_ opinions -- much less NASA!'