by Brian Dear
TICCIT interested the committee at first. “The video aspect of it then was revolutionary within a computer,” Hofstetter recalls. But they found TICCIT’s built-in instructional design structure too rigid and restrictive, and the video capabilities that made TICCIT so attractive also meant they could not use it over long-distance phone lines, so they decided to continue searching.
They visited CERL and got the full PLATO demo. PLATO seemed to fit the bill: it already had lots of courseware, which meant not only might there be material Delaware could put to use with its own students right away, but also that many professors at Illinois had already bet on the PLATO horse and had invested great amounts of time and money in creating that courseware. It also seemed to fit the bill because the system was capable of supporting terminals at remote locations via regular telephone lines. Quite impressive was the fact that, in terms of keyboard and display speed, you couldn’t distinguish any difference between a terminal connected ten feet from the mainframe and one dialed up from a thousand miles away. For Hofstetter there were other appealing aspects as well. PLATO was the only system with touch-sensitive screens. There were other systems with screen-marking capability, but those required light pens. With a touch-sensitive screen, the only hardware a student need know how to operate was his or her finger. Hofstetter was impressed. Among the many applications for music instruction that Hofstetter could foresee was a graphical representation of a keyboard on the screen, on which students could touch the “notes” to “play” a tune.
Hofstetter’s enthusiasm warmed the interest of the UD administration, and the committee decided to green-light PLATO. In the fall of 1974, UD established the PLATO Project with Hofstetter appointed as director. By the following spring, the project had received its first PLATO terminal, connected by a long-distance telephone line to CERL.
Hofstetter had wanted to create a series of lessons to help music students learn fundamental concepts of music, including chords, harmonies, and intervals. He hired a student named Bill Lynch, who had entered UD in 1972 as a freshman, drifted from majoring in math to majoring in art, and casually taken an electronic music for nonmusicians course Hofstetter offered one semester. It was during that course that Hofstetter asked the class if anyone knew how to program computers, and Lynch volunteered. He began to code Hofstetter’s lessons, first writing them in ALGOL on a Burroughs computer using a Tektronix 4010 graphics display. Hofstetter had managed to get one of Seymour Papert’s music boxes from MIT, a crude machine that when connected to a computer could be told to make simple monophonic musical sounds.
Hofstetter grew frustrated at the Tektronix display because it was not selectively erasable. “If you wanted to erase anything on the screen you had to erase the full screen,” he says. “PLATO had a selectively erasable screen, and PLATO had touch. I thought touch was the greatest thing for a musician because you could get away from typing codes and just have the students touching musical symbols.” He went away to Minneapolis for two weeks in the winter of 1975. “I froze my tail off in January learning TUTOR!” he says. No one at Delaware knew TUTOR. Lynch picked it up quickly when in March 1975 UD’s first PLATO terminal arrived from CERL, enabling him to move from a slow and unproductive ALGOL environment on the Burroughs to the far more productive, interactive environment on PLATO.
During the spring semester Hofstetter taught what was the first TUTOR programming course on the campus. A number of Delaware’s first “PLATO veterans,” as Hofstetter affectionately terms them, attended this course. “Of course they ended up learning far more than I ever knew about it. But I knew enough to help them get started.”
Hofstetter’s nascent Delaware PLATO project started soliciting proposals from various academic UD departments. Ten departments responded, indicating that they were interested in developing new courseware within their subject areas. In what was a near repeat of the widespread adoption of PLATO by the faculty at UI, the diversity of interests represented by those ten departments revealed much about the perceived potential for PLATO at Delaware: agriculture, art, computer science, continuing education, education, home economics, music, nursing, physical education, sociology. Not surprisingly, all ten proposals were approved in the summer of 1975. Eight part-time programmers were hired.
Soon another terminal arrived, and then two more. The number grew to eight, then sixteen. Professors began tinkering with the system, some hiring programmers to develop lessons in TUTOR for them. A number of terminals were set aside just for authoring, and were in heavy use to the point that it was hard to find an available one. A classroom full of PLATO terminals was opened in Willard Hall, home of the university’s department of education. Not all the terminals there were hooked into Illinois. By 1977, the Willard site had a handful connected to Control Data Corporation’s PLATO system running in Minneapolis. If a terminal wasn’t in use, it just sat there, with “Press NEXT to begin” on its display. Normally a user couldn’t tell just by walking up to a terminal what system it was connected to. So in order for people to know, little flags cut from construction paper were “flown” from the top of each terminal.
Soon the number of terminals expanded from sixteen to thirty-two, but that was not the end of the growth. “When it went to forty-eight,” recalls Hofstetter, “we were paying more money to the phone company for leasing lines across the country than it cost to buy our own system.” Armed with that fact, Hofstetter convinced UD that it was cheaper to buy a Control Data CYBER mainframe and start their own PLATO system. UD agreed and bought the system, including a CYBER 174 mainframe, which was delivered on January 31, 1978. By March all testing had been completed, and on St. Patrick’s Day, 1978, CDC turned over the keys to Delaware. Another PLATO system was born. Over the next six years the number of terminals connected to the UD system would grow to over three hundred.
—
The lessons Lynch was programming comprised an ambitious new series that Hofstetter called GUIDO, an acronym for Graded Units for Interactive Dictation Operations. Named after Guido D’Arezzo, an eleventh-century Italian known as the first real music educator, the GUIDO package consisted of two main segments: ear-training skills and musical theory written skills. A unit in the ear-training skills series typically consisted of three parts: first, PLATO would display an “answer form” on the screen; second, it would send signals through a cable behind the terminal to a simple sound synthesizer that would then play a sound; and third, it would ask questions about how the student perceived the sound. Every week, the system would print out for the instructor a summary of each student’s progress.
Lynch’s art background and his strong personal sense of design gave the GUIDO lessons a distinct look and feel that, over many years, continued to make the lessons stand out. Many PLATO lessons developed at CERL had rather boring designs, full of text, with inconsistent navigation instructions and layout. GUIDO stood out with crisp displays, dramatic typography for the page headings (words like “QUIZ” and “PLAY”), and featured clean, easy-to-figure-out touch interfaces with a musical keyboard at the bottom of the screen, and a dashboard of buttons for students to interact with. The GUIDO lessons were the first TUTOR lessons Lynch had ever programmed, and he had no experience in interaction design or user interface design. But the results spoke otherwise.
GUIDO lesson Credit 41
Back at CERL, Sherwin Gooch was meanwhile building a new music synthesizer box, the Gooch Synthetic Woodwind, which was exponentially more capable and better-sounding—with four-voice polyphony—compared to Papert’s music box from MIT. Hofstetter would adopt this as the standard for GUIDO instruction. The UD music building had a small classroom equipped with a number of PLATO IV terminals with Gooch boxes connected to them, and a pair of headphones for each student. (Gooch would eventually build an even more powerful box, the Gooch Cybernetic Synthesizer, but Hofstetter decided that Delaware should build its own, the result being the University of Delaware Sound Synthesizer.) Walking by the often dark room in the music buildi
ng, one might mistake the scene inside for an air traffic control center, with headphone-wearing students intently working at their GUIDO lessons, their faces dimly lit from the Orange Glow of the plasma panels, their fingers occasionally reaching out to touch the screen.
Music was not the only subject that found acceptance among faculty at Delaware. In just a few years, there would be hundreds of new lessons developed in accounting, student advisement, agriculture, anthropology, art, biology, business, chemical engineering and chemistry, civil engineering, counseling, economics, education, English, geography, geology, nutrition, family studies, Latin and foreign languages, library science, math, nursing, physical education, physics, political science, psychology, statistics, and urban affairs. The UD PLATO team developed more rigorous quality control standards for its lessons than were in place at the time for CERL-authored lessons.
—
Early University of Delaware PLATO users, including Lynch, would experience an interesting phenomenon that PLATO users at other sites would go through in subsequent years. By starting out on the CERL system, these users had seen firsthand the feverishly active, boundlessly creative, every-day-there’s-something-new-going-on online community of CERL, with many notesfiles including technical ones where a question might be answered in minutes, if not seconds. For UD courseware developers new to TUTOR, having this support network enabled them to come up to speed far more quickly than if they were not able to tap into the community. But then, as UD grew the number of terminals on campus, it began to migrate terminals from CERL to CDC’s own Minnesota system, a system that Lynch found empty, a desert island relatively devoid of the bustling activity he’d grown used to on CERL’s system. When UD activated its own PLATO system and all of the terminals were switched over to it, the bustling community was gone, the online experience for those with author signons as quiet as a tomb. Delaware would have to start from scratch building its own community, but it had a distinct advantage: by 1978, enough UD people had been exposed to CERL before the switchover that they knew what the ingredients were for an online community. The tools were all there, the notesfiles simply needed to be fleshed out. Delaware also had the advantage of having over several years hired several people from CERL, including Bonnie Seiler, Jim Wilson, and Brand Fortner. In short order, many of the most popular notesfiles from CERL had been re-created on the UD PLATO system, and slowly but surely activity perked up.
The new UD PLATO mainframe also had the advantage of having a telecommunications link connecting it to CDC’s machines in Minnesota. In fact, by the late 1970s, CDC had so many clients on its Minnesota PLATO services that it split it into multiple systems, one dedicated to internal R&D and other machines for customers, each machine supporting several hundred simultaneous users. All of these machines were also interconnected with telecommunications links. And, of course, there was a link between them and CERL. Add to this a series of powerful new capabilities enabling pnotes to be sent to any system so interlinked, and notesfiles being able to be interlinked as well, and the online community of PLATO suddenly started looking like a small Internet. Any system connected via a link to this growing network was able to take advantage of communicating with all the other linked systems. Any system that did not have a link was truly a desert island, its online community—if any—tiny in comparison to a linked system.
—
Delaware established a policy that other universities would copy in subsequent years: Control Data employees were not allowed on campus, period. Fred Hofstetter, in a 1980s interview, described the policy this way: “Our relationship with Control Data has been very, very professional. We want to be the user interface, we don’t want Control Data to be interacting with our users. Control Data is not allowed to speak with our users. They are not allowed to solicit among the faculty here. They must contact me. And over the years a couple of times some overzealous salespeople have contacted directly some of our users and they were fairly soon after taken off our account. Because we just do not permit that. We don’t want corporate salespeople dealing with our users. They come from a different environment….The main market for PLATO is industry. And the Control Data sales force is much more industry oriented than they are education oriented. When these fellows come in with their three-piece suits and they take faculty members out for their three-martini lunch and so on, it doesn’t have the academic feel to it that it needs to have to be successful on the campus.”
Another site, the University of Maryland, Baltimore County (UMBC), installed a PLATO system of its own in 1982, and also banned CDC personnel from poking their heads in on campus unless they were there for a specific purpose such as repairing equipment. UMBC hosted PLATO service for a number of school districts, including nearby Baltimore but also Richmond, Virginia.
There was palpable antipathy to Control Data at some of these schools. In Baltimore, CDC and Commercial Credit, which was based in Baltimore, had originally run PLATO service in some schools and adult education centers. UMBC inherited that business as part of its purchase agreement, which in essence gave UMBC control of PLATO for any nonprofit, educational, or governmental body that wished to use PLATO anywhere in the state of Maryland. Schools that had interacted with CDC personnel prior to the UMBC switchover had not had uniformly positive experiences with the company or its people. At the Walbrook High School in Baltimore, the school receptionist would sternly interrogate any PLATO personnel from UMBC before allowing them any further into the building: “Are you now, or have you ever been, an employee of Control Data Corporation?” was the pointed question. A visitor asked the question had the clear understanding that to say yes meant leaving the premises immediately.
UMBC epitomized the desert-island type of PLATO system that popped up around the world in the 1980s. Their missions were clear-cut: deliver PLATO courseware as a service. The university administrations would buy these multimillion-dollar systems, hire technical and educational staff to deliver and support the services, and when (or even if) asked if they wanted a telecommunications link back to the rest of the PLATO network, typically would not even know what that was or why it could possibly be beneficial. And when shown the cost of such a dedicated line to Minnesota, the reaction was usually “no thanks.” The mission for later-stage PLATO systems was typically to deliver existing courseware, rather than do what Illinois and Delaware did, enlisting the academic departments to find faculty members eager to create their own lessons for their respective courses. Control Data had by the early 1980s a large library of its own material, including the Basic Skills curriculum, which these remote PLATO sites were happy to provide to adult learning centers and other nearby institutions whose mission it was to help people improve their skills to get a better job. However, the astronomical cost of the CDC courseware would usually stop these universities in their tracks. UMBC closed down its PLATO system in just a couple of years, due to the sheer costs involved in offering CDC’s courseware. The royalties for that courseware were so high, the university could not charge enough to the various educational and training institutions using PLATO to pay for it. The CYBER and all of its equipment would be disconnected and carted away. (Once asked about the UMBC debacle, Bitzer had this to say: “The reason they didn’t have any more money was ’cause CDC screwed ’em.” He says he bought the used UMBC PLATO hardware through a broker “and I put it all to good use.”)
—
CDC sold PLATO systems to a number of educational institutions between the late 1970s and mid-1980s, including the University of Quebec, the University of Hawaii, the University of Western Australia, University of Alberta, University of Massachusetts, University of Connecticut, Florida State University, University of Nebraska, Royal Institute of Technology in Sweden, University of Brussels, and the University of the Western Cape in South Africa. Perhaps the biggest sale in CDC’s history was to the California State University system in the mid-1980s for a reputed $50 million. In addition to the many CSU campuses having access, many school districts in
California tried out PLATO. One, the Sweetwater School District in the San Diego area, found that troubled students and high school dropouts enjoyed using PLATO and often came back to finish their high school degrees because of it. The feedback the school administrators received from the students was essentially the same that prison inmates given a chance to use PLATO for instruction verbalized years earlier: they felt comfortable with the ever-patient machine, enjoyed the personalized feedback, and they were free from any potential embarrassment they might experience in a conventional classroom.
PLATO was also utilized in countless corporations for training, as well as a variety of military services in the U.S., Israel, and other countries. But despite installations all over the world, and pockets of profit here and there, by the mid-1980s, CDC’s mainframe-and-terminal-based PLATO systems were rapidly becoming dinosaurs. Microcomputers were taking over.
25
The Crash Pit
They came from all over the world, to drink, party, and compete on a Waimanalo polo field on the windward side of the island of Oahu. It was February 2013, time for the three-day Kaimana Klassik: an annual Ultimate Frisbee tournament. Ultimate Frisbee is a mash-up of strategies and gameplay from basketball, rugby, and soccer, but using a Frisbee instead of a ball. The game, invented in the late 1960s, has hundreds of collegiate teams in the United States, and organized leagues and tournaments all over the world.
Thirty-two teams would compete, some having traveled thousands of miles to Oahu. Team names tended to be a bit deranged: Philthy, Beer4Breakfast, Stigmata Chelada, Freaks, Sarcastic Fringeheads, French Kisses, Skeletor, and Freshly Squeezd. One Oahu resident, a longtime Klassik spectator and Stanford alum, had come to the tournament to cheer on the Stanford Bloodthirsty team. His name was Brodie Lockard. He made his way to the field in a motorized wheelchair along with a 24/7 personal caregiver.
