In 1974 I learned my first programming language, BASIC, while at a NSF summer science
training program at FAU in Boca Raton, Florida. There I was also exposed to Fortran and
APL. That same year I built a computer around Intel's 8008 microprocessor. My largest
application was a two player chase game played on a three inch Heath Kit oscilloscope. It
used every byte of the computer's 0.000244MB memory, a pair of 8 bit D-to-A converters
built from hand picked resistors and two crude homemade joysticks. (Historical footnote:
This was the same year that Bill Gates formed Microsoft to market his MITS Basic
interpreter.)
In my first year (1975) at
Stanford University,
I learned Algol/W and wrote a "cellular router" for a graduate student, Bob Rau.
Bill vanCleemput, an associate professor in the school of Electrical Engineering, oversaw
the research that year. From the summer of 1976 until I graduated with a BSEE in 1979, I
worked part-time at the Stanford Linear
Accelerator Center (SLAC) on a Computer Aided Design system for designing printed
circuit boards at SLAC. During that time I wrote a program for graphically editing
component placement and wiring on a circuit board. I also wrote utilities for manipulating
and viewing the circuit description database. Development tools on SLAC's IBM 370
mainframe included a line editor called Wilbur, a language called Mortran and core-dumps.
In the fall of 1979, Bill vanCleemput, Kim Stevens, David Coehlo and I founded
Silicon Valley Research, Inc. to develop and market CAD
tools for designing gate-array integrated circuits. From 1979 to 1983 I worked on
placement and routing editors, graphics libraries, and programming tools. Most of my
programming during this time was in Mortran and Pascal. All the software produced at SVR
ran on Prime, Apollo, VAX and IBM computers. Writing portable software was a major
concern.
In the summer of 1983 I left California for Alaska. My wife grew up in
Anchorage and she felt that Alaska had
some distinct advantages over California. That winter I purchased a Radio Shack Model 100.
It came with three integrated programs: TELCOM - a terminal emulator, TEXT - a simple page
editor, and BASIC - a simple BASIC interpreter. What intrigued me most about the Model 100
was its vacant ROM slot in the back. Unfortunately, the ROM slot's pinout did not match
the JEDEC EPROM standard. I built a cable to rearrange the pinout to match that of a 27C64
EPROM. The pinout problem looked like a hopeless obstacle until Craig Milroy, a product
designer, suggested using a flexible circuit board to wrap the EPROM and reroute its pins
to match that of the ROM socket. By the summer of 1984 I was making an 8080 assembler/debugger
on EPROMs and selling them in Portable 100 magazine.
Interest in an assembler was high and sales were good. Since, the Model 100 didn't have
a disk drive, it used part of its 32K of RAM for the file system and a cassette tape
recorder for backup. Having an assembler/debugger in ROM allowed power users to design
machine code to augment their BASIC programs. But, many callers wondered if Polar
Engineering could put their BASIC program into a ROM.
Put BASIC into a ROM? Was it possible? The Model 100 was built around a 8080 with 32K
of ROM and 32K of RAM. The vacant ROM slot in the back was bank selected against the 32K
ROM which held the BASIC interpreter. That meant that the BASIC program code in the ROM
would have to be temporarily moved to RAM so that the BASIC interpreter in the other ROM
could run it. Enough callers seemed to want to put their BASIC programs in a ROM that I
decided to try to solve the problem.
In the spring of 1985 (after two thousand hours of hard work) Polar Engineering began
marketing its Guardian ROM service. Each customer provided Polar Engineering with a Model
100 BASIC program on cassette tape. That BASIC program was sent through a series of
transformations on an IBM PC/AT which produced a ROM image. The order was filled by
burning the requested number of EPROMs and assembling them with the flexible circuit
board. Sales were very good. Customers like the Associated Press and McDonalds bought lots
of EPROMs. I spent many hours each day on the phone, answering questions, helping
customers and filling orders. So much time that I decided it was getting out of hand.
In the fall of 1986 I licensed the entire BASIC on a ROM business to
Traveling Software (now call LapLink.com), the biggest
supplier of Model 100 software and also Polar Engineering's biggest customer.
During the summer of 1987 I received my first (and last) Macintosh, a MacII. That fall
I developed the Macintosh half of LapLink Mac for Traveling Software. For the next two
years I learned how to write GUI applications in Pascal for the Mac and developed LapLink
Mac III which supported file transfer to other Macs or PCs via Modem, AppleTalk and direct
serial link. Traveling Software brought LapLink Mac III to market in January of 1990.
Starting in late 1990 I began to learn C++. I found the process of learning C++ both
enlightening and arduous. Never had I encountered a language with so much potential and so
much detail. I love it.
In the summer of 1991 I began to write a BASIC interpreter in C++. After two years of
development Polar Engineering and Consulting began marketing WinWrap(R) Basic for Windows.
In the fall of 1994
Sax Software began marketing
the Sax Basic Engine<, a Visual Basic extension
(VBX) which adds Visual Basic compatible scripting to VB applications. Polar Engineering
and Consulting still services the scripting needs of C/C++ application designers.
In November of 1995, Sax Software started shipping OCX versions of the Sax Basic Engine
for Visual Basic version 4. Both the 16 bit and 32 bit VB compilers are supported.
In October of 1996, Bill vanCleemput asked me to help on a technical investigation for
Texas Instruments' lawsuit against Samsung. It was an exciting experience and a lot of
work. Our investigation into the operation of IC manufacturing tools provided technical
support to the testifying expert witnesses in the case. Seven weeks into the investigation
TI and Samsung settled the dispute. Each side agreed to share their intellectual property
portfolios. However, since TI had substantially more IP, Samsung agreed to pay TI an
estimated $1 Billion over the 10 year life of the agreement. My first exposure to
technical investigation was over, but not for long...
In June of 1997 Louis Touton of Jones, Day, Reavis and Pogue (the same law firm that
did the TI v Samsung cases) asked me to write a program to provide a court-room
illustration of the algorithm described in a technical paper. The technical paper was part
of a prior art defense in a patent litigation case. The program was written in VB. It
showed flow throw flowcharts, activity on hardware buses and actual displayed text.
In April of 1998 I began working on TI's lawsuit against Hyundai. The investigation
continued on and off until the trial in March of 1999. At the trial I testified as an
expert witness with regard to the testing and data gathering that I did during the
investigation. My testimony provided the eight person jury with a clear picture of how
data is gathered in a Patent infringement case. Based on this data the two expert witness
for TI, Costa Spanos and Raphel Reif, provided their infringement analysis for the IC
manufacturing tools that I investigated. After a two week trial and one day of
deliberation the Jury found in favor of TI. After that I continued to work for TI in
another one of the TI v Hyundai cases. In June of 1999 TI and Hyundai settled the dispute.
As in the Samsung case, both sides agreed to share their intellectual property portfolios.
Again, TI had a larger portfolio, so Hyundai agreed to pay TI an estimated $1.2B over the
10 year life of the agreement.
In October of 1999, Carl Oppedahl of Oppedahl & Larson LLP (
http://www.oppedahl.com/) asked me to write several web
data mining applications (Feathers, Partridge and Raptor). These applications are written in
Microsoft's ATL (a C++ template library for ActiveX). Monitoring is done via HTTP using
WinSock and E-Mail notifications are sent via SMTP also using WinSock. The applications
are self-installing and digitally signed for easy and secure installation over the web.
September 2000: In TI v LTC. Technical investigation for TI. Case settled out of court in 2002.
January 2001: Digcom v Nokia. Technical investigation for Nokia. Case settled out of court in 2002.
January 2003: MLR v Nokia. Technical investigation for Nokia. Case settled out of court in 2003.
Who knows what will happen next...
For more information contact:
Tom Bennett
Polar Engineering and Consulting
Nikiski, AK 99635
Phone: (907) 776-5509
Email:
