behind the towers

may 7, 2026

ted turner died yesterday. he started cnn in 1980 while simultaneously attempting to defend his winning the america’s cup. of course, i have a ted turner story. i wrote the navigation software for courageous that year.

my officemate at dec, david schanin, was a hardware engineer who had somehow inherited the job of putting a dec lsi-11 microprocessor system on courageous, turner’s america’s cup 12-meter yacht. he asked me if i wanted to write the software. i said that i would love to, but i knew nothing about racing sailboats. he said not to worry, the navigator for courageous was an aerospace engineer for grumman aircraft on long island, he would help me learn. his name was bill jorch.

i met jorch a few weeks later when david and i went to newport, rhode island to install the lsi-11 hardware in a waterproof plastic enclosure that david had designed. he drew up the course on a paper pad, added some illustrations describing the forces acting on the boat, the definitions of “true” and “apparent wind”, and identified the sensors that would provide the real-time data to the lsi-11 and my software.

david and i decided that we would design the software to run on rt-11, dec’s popular real-time operating system for the pdp-11 family. we used dec’s tu-58 tape drive and media to load the operating system and the navigation software over one of two rs-232c serial communcation lines. the other serial line was connected to a handheld 2 x 24 character termiflex terminal. this would provide the interface to jorch, the navigator, for issuing commands to the software, and providing him with real-time data on the boat’s speed, bearing to the next course mark, and time to initiation of the next turn.

i decided to write the software in fortran for a variety of reasons. it was well-known by engineers and scientists, floating point arithmetic was computed efficiently by a full-function library, and it was relatively easy to debug. this meant that i would write, edit, compile, and test the code on a separate development computer, employing some fortran-coded “stubs” to simulate the data coming from the boat’s sensors. i figured i could create fully functional code in the shortest amount of time in this way.

i spent practically every weekend in newport in the summer of 1980. whenever coursgeous was docked, i was testing every wire, sensor, connector, interface, and board in our seafaring syatem. i did not go out on the boat with the crew whenever they were practicing for the actual race.

newport is a very lively place in the summer, especially among the sailing set. courageous berth was on the same dock as a popular watering hole, the black pearl. when i was finished for the night, usually well after 10pm, i would indulge at the bar of the black pearl. i never stayed long because the work was tiring, so rack time was before midnight.

once bill jorch and i decided that we were ready to put the system to the test during a practice race, i was nervous and excited. this was ted turner’s racing yacht! he was the reining champion. i did not want to let him, bill, ted’s tactician gary jopson, nor any of the crew down. i was tense. since the boat was going to be on the water most of that weekend, i did not go to newport.

on monday, i received the bad news from bill. he said that the system crashed after the first few minutes that they were underway. every time. he knew how to reboot the machine after a crash, but was still unable to kerp it alive for any substantial period. i was truly worried and embarassed, but what could i do? bill asked me to come to newport the following weekend. ted was scheduled to make a brief appearance then fly to atlanta to attend to cnn.

i drove to newport on friday afternoon, and at bill’s invitation, stayed with him and the entire courageous crew at their dormitory on the csmpus of salve regina college. bill invited me to have dinner with the crew. that is where i met ted turner.

bill had briefed ted on my role and background before walking me over to ted’s table. right away, turner sized me up and said, in his characteristic drawl, “so, you’re the computer guy. thanks for coming down.” i stumbled over my reply, something to the effect that “we will get this under control.”

the next day i went out in courageous, standing in the “navigator’s hole” closest to the lsi-11 system. there are two navigator’s holes on a 12-meter america’s cup yacht. the navigator selects the one that affords him the best view based on the boat’s direction. i was to stay in the one closest to our system which was bolted to the underside of the deck. that meant that the electronics were only accessible from below. i wasn’t too thrilled about this because i had a long history of motion sickness. but this was a sailing vessel, how bad can it get?

on the way out to the middle of long island sound where the practice races are held, the machine crashed several times on me. i rebooted it every time. hoping to get some clue from the machine’s state as to what caused the crash. the lsi-11 had a rudimentary debugger implemented in the processor’s microcode called “microodt” for “micro octal debugging tool”. it allowed a programmer to examine the cpu registers, memory locations, and i/o device registers. so in theory, one could gain sufficient insight into what might have caused the machine to go astray.

which i might have been able to do but for the sudden onset of projectile vomiting into the bottom of my navigator’s hole. at that point i was absolutely worthless to the cause. worse, i was now a liability to the crew, to bill, and to gary jopson who was the acting captain of courageous. they seriously needed me off the boat. they signalled for the motor yacht tender to come alongside and get my sorry ass off courageous. which they did.

i curled up on the aft deck of the tender, breathing in diesel fumes which had the effect of causing me to wretch continuously until we returned to the dock in newport. not a feeling of success.

i did not see ted turner again. courageous lost in an elimination round that decided which boat was to represent the united states v the international team decided in the same way.

courageous’ berth mate was clipper, the yacht supported by data general (dg), a dec competitor. dg’s board of directors had budgeted $50,000 for providing timing instrumentation for the 1980 olympics in moscow. the soviet union invaded afghanistan in late 1979. president carter decided that the usa would boycott the olympics in moscow. so dg was out of the olympics instrumentation business. instead, the company steered the budget to supporting clipper’s entry in the america’s cup.

i was competing with a dozen engineers and technicians who were working on company time. i was reminded several times by my management that i was not to work on the courageous project on dec’s time. which directive of course i ignored on several occasions. nonetheless, it was more than a bit discouraging that dg provided a solid team to support clipper, while i was expected to rely on myself.

the reason that i became seasick so suddenly was in no small part due to courageous getting into a “tacking duel” with clipper in rough long island sound waters. a “tacking duel” is a contest between racing yachts attempting to gain advantage by “stealing wind” from an opponent. this has the effect of causing both boats to undergo rapid changes in direction – tacking – in order to maximize the amount of force exerted by the wind to provide additional boat speed.

12-meter yachts have enormous sail area. changes in diection are performed quickly, and therefore violently. as a passenger, a racing yacht in a tacking duel feels like driving a formula 1 car flat out, making 30 degree turns every few hundred feet. it is not for the feint of heart. or stomach.

footnote – after wondering about this problem for years, i have concluded that the “crashing problem” was not caused by software, neither mine nor rt-11. if application software running on rt-11 were to misbehave – an illegal memory reference, attempting to execute an illegal instruction, dynamic memory exhaustion – rt-11 would field the fault, terminate the process, and return the console terminal to the control of the command line interpreter. however, a number of hardware conditions can cause the processor to drop into microodt. excessive device interrupts, corruption of the stack pointer are two examples. one more stands out in my mind though – creating a “break” signal on either of the system’s two serial lines will force the uart device on the serial interface board to generate a reset of the microprocessor, forcing it into uodt. i believe that there was sufficient stray electrostatic charge present inside courageous to cause the uart to sense a “break” signal (logic 0 for more than one character time).

electrical noise, in other words, was the proximate cause of the coursgeous crashes. in my humble opinion.

that’s my ted turner story. he was indeed “captain outrageous”.

rest in peace, ted.