GARY KILDALL
===========
Tools that made a small computer useful

Gary Kildall was my hero. These notes cover PL/M, CP/M, and some
recollections from people who worked with him.

START WITH THE TOOLS

Kildall taught computer science at the Naval Postgraduate School in
Monterey and consulted for Intel. In the early 1970s, a microprocessor
was an intriguing component, but programming it still depended on
tools running on larger computers. Kildall developed PL/M, a
high-level language for microprocessors. That gave programmers a way
to express their work above the level of individual machine
instructions. He brought the habits of a computer scientist to
hardware that was only beginning to find its place. [1]

CP/M grew out of an equally practical need: better storage. Paper tape
made an awkward home for programs. A floppy disk offered fast access
and the possibility of keeping useful files beside the computer.
Kildall developed his early system with the help of a simulator before
the disk hardware was ready. John Torode helped make the controller
work, and in 1974 they brought CP/M to life on an Intel Intellec-8
system. They could now load programs from disk. [2]

WHAT WAS BEHIND THE PROMPT?

The working parts of CP/M were small enough to explain. The Console
Command Processor handled commands. The Basic Disk Operating System
provided file services. The Basic I/O System dealt with the machine's
devices. Programs loaded into the Transient Program Area. Together,
these pieces let someone store, edit, assemble, debug and run programs
on a small machine. The surviving documentation makes the system
approachable: its boundaries are visible, and each part has a job. [2]

The BIOS boundary was particularly important. Different computers had
different disk controllers and other hardware. When Glenn Ewing wanted
CP/M adapted for IMSAI equipment, Kildall separated the hardware-
dependent routines. A manufacturer could supply that part while
retaining the rest of the system. Software could run on more machines
within the compatible processor family. This was a practical kind of
portability, with real limits; it did not make every processor or
peripheral interchangeable. [1]

A COMPANY, AND A CLASSROOM

Gary and Dorothy McEwen built Digital Research around CP/M. A common
operating-system interface gave independent software developers a
market stretching beyond one maker's computer. Their children later
described a deliberate belief that operating systems and applications
should be separate businesses. That account helps explain the shape of
the company: make a useful foundation and leave room for other people
to build upon it. [3]

There is another useful view of Kildall in the Computer History
Museum's 2014 anniversary discussion. Gordon Eubanks remembers him
handing out CP/M source code in an operating-systems class. Students
could study the actual machinery. Eubanks also recalls Kildall
suggesting a microcomputer word processor as a thesis project, an
opportunity he passed over. [4]

Tom Rolander remembers an architect thinking ahead to multitasking and
networking. Kildall's advice was to let that longer view guide
apparently arbitrary choices in a small project. Rolander went on to
develop Digital Research's multitasking system. [4]

ON THE SCREEN, ON THE SHELF

The Computer Chronicles made a memorial episode after Gary died. It's
very nicely done, with contributions from Eubanks, Rolander, Lee
Lorenzen and Alan Cooper. The 1995 program is available at the
Internet Archive. [5]

For his own account, the Computer History Museum offers part of
Kildall's unfinished memoir, Computer Connections, written in 1993.
His children supplied an introduction and authorized the release. It
is a personal recollection, so it deserves to be read alongside the
documents and other participants' accounts. The museum asks readers to
link to its page rather than reproduce the manuscript; that is the
route provided here. [3]

The references below offer several ways in: the system itself,
accounts from people who worked with Kildall, and his own writing. The
CP/M documentation is a good place to start for a closer look at how
the software worked.

REFERENCES AND FURTHER READING

[1] David Laws, Computer History Museum (2014).
    Gary Kildall and the 40th Anniversary of the Birth of the
    PC Operating System. Historical account with source notes.
    [web]
https://computerhistory.org/blog/gary-kildall-40th-anniversary-of-the-birth-of-the-pc-operating-system/

[2] David Laws, Computer History Museum (2014).
    Early Digital Research CP/M Source Code. System history,
    descriptions of preserved versions, manuals and source links.
    [web]
https://computerhistory.org/blog/early-digital-research-cpm-source-code/

[3] Leonard J. Shustek, Computer History Museum (2016).
    In His Own Words: Gary Kildall. Includes Scott and Kristin
    Kildall's introduction and access to the memoir excerpt.
    [web; memoir PDF available through page]
https://computerhistory.org/blog/in-his-own-words-gary-kildall/

[4] Computer History Museum (2014).
    Legacy of Gary Kildall: The CP/M IEEE Milestone Dedication.
    Discussion transcript; especially printed pages 10-14.
    [PDF]
https://archive.computerhistory.org/resources/access/text/2014/06/102746909-05-01-acc.pdf

[5] Computer Chronicles / Stewart Cheifet Productions (1995).
    Gary Kildall Special. Preserved television program.
    [web/video]
https://archive.org/details/GaryKild