First, Dave wrote:
> Date: Thu, 23 May 2024 15:53:53 -0400
> From: Dave Dunfield
> I've just passed on my "Mits Altair 8800" - this is a very historic system
> from the 70s - it is:
> First system Bill Gates wrote code for (long before Microsoft)
Which is on the face of it incorrect.
Then Christian Corti responded (in replying to someone else's objection to
Dave's claim of firstness:
> Date: Fri, 24 May 2024 11:44:46 +0200 (CEST)
> From: Christian Corti via cctalk <cctalk@classiccmp.org>
> >> First system Bill Gates wrote code for (long before Microsoft)
> Didn't he write code for DEC machines at his school before that?
Which is nearer the mark, but not fully correct.
Then Sellam Abraham stuck his oar in:
> Date: Fri, 24 May 2024 07:40:31 -0700
> From: Sellam Abraham via cctalk <cctalk@classiccmp.org>
> > Didn't he write code for DEC machines at his school before that?
> Yes, poorly.
Oh, FFS, Sellam.
OK. Once again, the history goes like this. I have heard it from the horses'
mouths (yes, plural).
Bill Gates and Paul Allen, along with 4 other students (out of a class of about
20), really cottoned onto programming in BASIC when a class was offered at
their school, the Lakeside School in Seattle. That class used a remote
timesharing service called GEIS (General Electric Information System), which
ran on GE 635 computers.
The six boys (it was a boys' school until the next year when it went co-ed)
were allowed to visit a new computer service bureau called CCC, because one of
their mothers was acquainted with one of the primaries. This company was using
a DEC PDP-10 timesharing system; the boys were given guest accounts under the
proviso that when the system crashed they would document what they were doing
at the time of the crash.
They were so eager to learn that the systems programmers (two MIT alums and a
Stanford alum) allowed them access to the hardware and system call reference
manuals, so that they learned assembler programming as well as BASIC, to an
expert level.
The summer between Paul's graduation and starting college, he along with Bill
and three others of the group got ACTUAL PAYING JOBS PROGRAMMING PDP-10 SYSTEMS
FOR THE BONNEVILLE POWER ADMINISTRATION, on a project called RODS (Real-time
Operational Data System) which used the systems for control purposes. (The
sixth member of their coterie got a job as a junior ranger at Mount Rainier
National Park, so wasn't interested in being indoors all day all summer.)
Paul dropped out of college after his sophomore year and moved to the Boston
area, where he worked for Honeywell's software division and hung out with Bill
and Bill's college friends, meanwhile looking for a way to have a small
computer of their own. They read the industry magazines to news of small
systems.
In the mean time, they tried to create a company to sell a traffic counting
device based on the Intel 8008 microprocessor. The prototype hardware failed
in their first demonstration to the City of Seattle traffic department, and
they shelved the idea.
When the Altair issue of Popular Electronics came out in mid-December 1974
(cover data January 1975), they were prepared for the challenge. After
ascertaining that Ed Roberts and MITS would entertain the idea of looking at a
BASIC interpreter for the new system, they sat down and created one from whole
cloth, with the division of labor as follows:
Bill Gates: the interpreter itself
Paul Allen: a simulator running on the PDP-10 for the Intel 8080 processor
Monte Davidoff: a math whiz freshman who wrote the transcendental math
routines
(My sources are Paul Allen and Bob Barnett. Bob was Paul and Bill's manager at
RODS, and the original business manager for Living Computer Museum. I have no
reason to believe that either had any reason to lie to me.)
Micro-soft incorporated in June/July 1975, so six months after they wrote their
first 8080 machine code, so Dave is wrong about "long before Microsoft".
And Sellam is simply wrong.
Rich
