by Brian Dear
In addition to Regency, Tenczar, Ron Klass, and Ray Ozzie cofounded an independent software company called USE—short for Urbana Software Enterprises. USE’s mission had been to create a microcomputer version of the TUTOR language. The company operated out of offices of Regency, which would eventually subsume the USE project’s TUTOR-like language, and call the new language USE as well. Regency would compete in the computer-based training industry for the next fifteen years.
In 1981 Tenczar left Regency and began Computer Teaching Corporation, shipping a new TUTOR-like language called EnCORE in January 1984. Trademark issues forced the renaming of the product to TenCORE later that year. TenCORE evolved into a powerful authoring language for PCs that attracted many of CDC’s key corporate PLATO clients.
“Through until 1981 or so, I’d say PLATO was state of the art,” Tenczar once explained in a 1987 interview. “Since then it’s worse than that….It’s always been a history of fighting, of the current staff fighting the new changes. It’s a real fight to have people change. And the use of micros—there’s still the fight going on over at CERL. Bitzer truly believes that it’s not the way. My company’s success shows that it’s a very viable way.”
Michael Allen
Once Michael Allen’s team at CDC had released the PLATO Learning Management (PLM) feature to customers, which he believed would address PLATO’s lack of tools for computer-managed instruction, he embarked on a new project that would ultimately be called PCD3, for PLATO Courseware Design, Development, and Delivery. “PCD3 started as an effort to find a more cost-effective way to develop good courseware,” says Allen, “because TUTOR programming was just too expensive. Looking at all the courseware and how much Norris [and CDC] was investing in courseware, it was just astronomical.” And he wanted that cost brought down. Allen’s expectation was that the result of this new project would be a better authoring language with more structure. “I started this because I wanted the quality to go up, I thought we needed better material, and I decided it was too hard to build complex courseware.” He hoped that whatever emerged out of this new project would mean that “you wouldn’t have to have years of TUTOR experience before you could build really good stuff. And what happened is that that message kind of got up to Norris that here’s this guy that thinks that he can reduce the cost of developing courseware. And so these things kind of came together. I ended up getting my money. They’re thinking that I’m trying to reduce the cost, and my thinking—that my primary objective—was to increase the quality.”
Norris was generous with funding, and Allen formed a team to do serious research on authoring PLATO lessons. They wanted to understand what was holding back the development of a lot more CBE courseware. Why did it take so long to create courseware, and what could they do to make it easier to create? “We realized that TUTOR was the culprit,” he says. “We also found out that interpreting educational or instructional designs into TUTOR was hard. You couldn’t have an educator do it in most cases because they just couldn’t master TUTOR enough in a short enough amount of time. So you ended up having teams all the time, you had the programmers, and you had the subject-matter expert, and you had the designers, and so on.” After interviewing courseware developers and subject-matter experts, he concluded that both groups were talking past each other and not communicating their intents clearly enough. So he wanted to build a tool that would increase involvement of each side for a project without having the nontechnical people having to learn TUTOR. The result was PCD3.
Separately, another group led by Jim Glish developed a simpler, more form-based PCD1. Oddly, this came out after PCD3. To make matters more confusing for customers, CDC also had yet another team, out in California, create yet another product, PCD2, which so baffled Mike Allen he still doesn’t know what to make of it.
“PCD3 I always considered to be a prototype of a product,” Allen says. “As it turned out, it really wasn’t another language like another TUTOR, it was more of an iconic-based approach. I thought it had some really good concepts in it and I wanted to go forward with it.”
Unfortunately, Control Data’s mid-1980s financial crisis prevented Allen from taking the PCD3 product any further. “It was the wrong time because Control Data was having so much trouble with its disk drive business and you know lots of its divisions ran into trouble and we were losing money and here I wanted to do something new.” Right at this time, Bill Norris gave Allen a “best of PLATO award,” which Allen described as “a great, big, huge, heavy, bust of Plato.” “I was invited to speak to the board of directors…and I said we’ve got something here that I’m really excited about, it’s a new way to develop software, and it has big implications for PLATO, and it can go more broadly than that, and I appreciate your giving me the millions of dollars that I spent by CDC’s calculations, it was something over $4 million that I spent researching how to develop what a better tool might look like, a better authoring tool, and I expressed my appreciation for that and said we’ve got something here that’s really going to take the market by storm, and lots of applause and so forth, and almost immediately after that, I talked with somebody in marketing and he said, ‘You aren’t seriously thinking about putting this out as a product are you?’ ” It turned out marketing had no money for another product—they had just spent another large sum on new four-color brochures and training materials for PLATO.
“Here’s a company in desperate need,” says Allen, “and we’ve got an opportunity here to do something that will really get a lot of attention and I think make a lot of money and so on, and he said, ‘I can’t. We can’t do that. We’ve already printed our brochures!’ And so I took that to Bill Norris, when I heard that, and I said, ‘Listen, can you believe this? What kind of business strategy and thinking is this?’ And he said, ‘I don’t want to admit defeat, but I got to tell you, things are not good in the company now, and support for PLATO is at an all-time low, and I really appreciate what you’ve done, and I think what you’re doing is really exciting, but I think that as a product champion, you’re going to have to take it outside Control Data and run with it, because I think it’s going to die here and by the time we can get support for it and do something with it inside Control Data your technology will be dated, and so you got to go with it.’ ”
It took Allen over two years to get it out of CDC. Every time he tried, some VP would offer to help but then attempted to kill the product behind Allen’s back. Finally he appealed to a high-level executive who ran the Government Systems Division, who was appalled at the treatment Allen was getting from various VPs. He forced CDC’s executives to come up with a legitimate business plan for the next generation of PCD3, or stop blocking Allen from taking it to market on his own. “Of course, nobody had a business plan,” Allen says. Once he got the go-ahead to proceed independently, lawyers raised new intellectual property roadblocks. “And so the agreement came out that I could have my know-how, but I couldn’t take any documents. So that’s what we did. When we started the company, I owed Control Data $2 million [for] my know-how, but if I paid them back earlier, why, there was a reverse-decreasing schedule, so if within the first year I could buy out Control Data’s interest in my own company, for $400,000, then I would be free and clear forever, and owe CDC nothing more….I had raised enough money so that in the eleventh month I paid $400,000 to Control Data and had the rights to do it.”
The new company, formed in 1987, would be called Authorware. The product was called Course of Action, and was developed on a Macintosh. Eventually it was renamed Authorware. Before the product shipped, Allen had the team re-create Darts, How the West Was One, Titrate, and the Fruit Fly lessons from PLATO to prove that the new product could do all of those famous lessons—on a Macintosh. It was a dramatic step forward beyond what PLATO could do: here was a full mouse-controlled graphical-user-interface environment, rich with menus and icons, for developing compelling interactive learning content. The product, not cheap at $8,000 apiece for the professional version, n
evertheless grabbed 80 percent of the authoring system/computer-based training market within three years, becoming the de facto standard in the industry.
Sometime around 1988 Steve Jobs made a move to hire Allen out of Authorware to join NeXT, Inc. “We went out and demonstrated the early version of Authorware to him,” Allen says. Jobs was impressed. “He said, ‘You like this, you’re in this education stuff,’ and I said, ‘Yeah.’ He said, ‘Huge market for me, huge market for me. I need somebody who can bridge computer programming with user interface with education. You’re my man, I want you to come and head it up for me.’ And I said, ‘Whoa, I just wanted to get your funding for Authorware.’ And he said, ‘Well, you think about it, but you can’t turn this down.’ He left, and it was about lunchtime, and I was wanting to go to have lunch with him and they said, ‘Well, nobody goes to lunch with Steve in his office.’ His team, the top guys that were there, took me to lunch, and they said, ‘You got to think about this carefully, because, let me just tell you a few things. Every good idea you have is Steve’s idea. You don’t intend to win an argument with Steve. Even though you may actually feel like you’ve won it, you know, you only can sell your idea and it becomes Steve’s and actually you were probably arguing against it. So you don’t get credit for anything, except among us, you know, we all know that, and he’s very temperamental, and he may stand behind you a long time and he may not. So, there’s a big risk, but we’d love to have you.’ ” Allen turned down the offer. Steve Jobs was furious. Nobody turns me down, he told Allen.
In 1991 another hot multimedia company called MacroMind merged with Paracomp, makers of popular 3D graphics software. A year later, that combination in turn agreed to merge with Allen’s Authorware company to form a multimedia authoring powerhouse called Macromedia, which subsequently had its initial public offering and was soon worth over $1.5 billion. In 1996, a struggling NeXT was acquired for $400 million by an even more struggling Apple. In 2005, Adobe Systems, Inc., acquired Macromedia for $3.4 billion.
Ray Ozzie, Tim Halvorsen, and Len Kawell
Ozzie, Halvorsen, and Kawell were three friends who worked as junior systems programmers at CERL while pursuing their undergraduate degrees at UI in the 1970s. In 1977, Halvorsen and Kawell had taken jobs at Digital Equipment Corporation (Ozzie, widely recognized by Microsoft years later as “one of the world’s best programmers,” had been turned down). Ozzie stuck around and cofounded Urbana Software Enterprises, working there for a while, but he was eager to move out and join the burgeoning computer industry like his friends had done. In 1979 he got a job offer at Data General in Massachusetts to join a small team led by a brilliant programmer named Jonathan Sachs.
“DEC had just started to develop this processor called the VAX,” says Halvorsen. “I really wanted to be involved in the VAX. It was just starting. It hadn’t really gotten anywhere yet, and I really wanted to jump on it because I was just excited about the whole idea.” But Halvorsen immediately experienced the culture shock of leaving the teeming, thriving PLATO virtual community, with all of its tools for communication and collaboration. “Leaving PLATO left a gap,” he says, “the gap being the entire experience of community. You lose that right off the bat….I’m talking about the degree to which you’re interacting with the rest of the community that’s electronically available to you, and the constant availability of an always networked system, was, you know, addictive. Obviously it was. And so Len and I both went to DEC. And the first things we did—we went into the same group—the first things we did, I mean, I wrote TERM-talk for VMS [the new VAX operating system], and he wrote pnotes, essentially. Both of us basically wrote it as a way in which to provide some of that experience in our new computing environment….That got us all started, and obviously everybody loved it because prior to that point there was no such thing as mail.”
As a result of the success they had PLATO-izing the VAX environment, Kawell decided to replicate PLATO’s notesfiles as well. “It was all written in Pascal,” says Halvorsen, “it was a simple little version, but it worked just fine.”
Halvorsen and Kawell successfully imbued the VAX VMS operating system with TERM-talk, Personal Notes, and Group Notes (the notesfiles), and VMS included them as features. “Essentially we were trying to see if some of that [PLATO] experience mapped and translated into the commercial world.”
“We were having a great old time trying to be real-life commercial programmers and whatnot,” says Halvorsen, but the intense experience of PLATO the three had gone through at UI kept giving them ideas for wanting to do something, a project, maybe a new company of their own. “Ray and I got the bug,” says Halvorsen, “because we hung around all the time together, got the bug of wanting to start our own company. We made a couple attempts to try and start a company with venture capital and whatnot, and it didn’t work out so well.” Personality conflicts, conflicting personal agendas, and a mediocre business plan for something they called Microcosm did not excite the VCs. Halvorsen is glad it failed. “I think it is really valuable experience when you’re first starting out,” he says, “is to fail right off the bat.”
Meanwhile, Kawell and Halvorsen thrived at DEC.
Ozzie tired of minicomputer maker Data General, and was eager to enter the booming business of microcomputer software. He joined Software Arts, makers of the hottest software application at the time, a revolutionary spreadsheet called VisiCalc, where he was involved in developing a version for the IBM PC. In the meantime, Jonathan Sachs had left Data General to team up with a local software entrepreneur, Mitch Kapor, to create a new spreadsheet they were calling 1-2-3. The company they formed was called Lotus Development Corporation. Lotus 1-2-3 enjoyed explosive success in the IBM PC marketplace, and the company was riding high. Sachs kept contacting Ozzie to join the team (“Come to Lotus, come to Lotus…we need more good people,” Ozzie recalls him saying). Ozzie was by now keen on doing his own new program relating to online collaboration among teams, which he initially called “MX,” the ideas of which had been floating around for a while but were now beginning to dominate his focus. MX eventually got a new code name, “Echo,” only to eventually get another, “Notes,” named intentionally after PLATO Notes. Ozzie wanted to take the ideas he had seen work so well on PLATO—tools for team collaboration and productivity—and bring them to the workplace, where it was abundantly clear by the mid-1980s that workplaces everywhere were going to be filled with networked PCs. There was an opportunity to address a software product void that few even knew was coming.
Sachs continued to pester Ozzie, urging him to come meet Mitch Kapor and at least hear him out. So he did, first meeting him at a computer convention where Lotus 1-2-3 was the hit of the show. He told Kapor, “I wanna do this thing, I know you wanna do spreadsheets, but I can’t come to work here because I really wanna do this other thing.” He told Kapor he and Halvorsen had tried doing Microcosm but had gotten nowhere with venture capitalists, which was frustrating, and now he wasn’t sure how he was going to get VC money for his new idea. Kapor offered him a scratch-my-back, I’ll-scratch-yours deal. Ozzie would help Lotus develop its follow-on product, called Symphony, an integrated suite of applications, one of the first on an IBM PC, and when that was delivered, Kapor would find a way to fund Ozzie’s project. “He didn’t know what the hell it was,” says Ozzie, “he didn’t understand what I wanted to do, but he knew what he needed. So we did a handshake deal, I went to work at Lotus.” He set aside his “Notes” plans and worked for nine months on Symphony, delivering it in July 1984—ahead of schedule. “He was so happy because it only took nine months, he came downstairs, I remember it vividly the day it shipped, and he said, ‘Okay, you did your part, you no longer have to do anything on this, start working on your specs for what you want to work on’ and we basically discussed that Lotus was so flush with cash at the time—the cash was rolling in so fast we didn’t know what to do with it—so Mitch figured that we would be able to do something Lotus funded. The moment that he rais
ed hopes, like, hey, this could be a reality, I spent like a month putting up the specs and doing up some screen snapshots on an early Macintosh of what the thing would look like.” Filled with excitement, he reached out to Halvorsen and Kawell, inviting them to drop what they were doing and join him in building this new product. Halvorsen and Kawell enthusiastically agreed to join if Ozzie could actually land the funding.
Ozzie did not want his new project to be situated in the Lotus office. “He wanted the thing to have its own culture,” says Kapor. “He put it in a geographic remove, that all made sense to me.” Rather than a conventional venture capital kind of funding, Kapor struck a creative deal with Ozzie that was more akin to a publishing deal: Lotus would advance Ozzie’s company about $1 million from future royalty earnings, which were effectively locked up with Lotus. “It was a unique arrangement, as far as I know, and it was structured that there were incentives to work with each other and to work out issues rather than not to. And there’s a lot of mutual trust….There was kind of an understanding between Ray and me at a personal level that made the thing go.”
When Ozzie finally presented his ideas for “Notes” to Kapor and Lotus’s internal product committee, “they basically thought it was wacky, they didn’t understand how I would do what I was saying, but they all nodded their heads, yes, yeah, sure, what the heck, fund the guy. Except for one guy, the VP of marketing at the time, Jim Manzi.” Manzi told Ozzie he wasn’t sure how he was going to sell this thing. “And so I went back,” says Ozzie, “and Mitch said, ‘Look’—he wasn’t in the habit of overriding his people—he just said, ‘Go change the stuff to make it a single-user product.’ So I said, ‘How the hell do I go and take what I’ve done and make it a single-user product? I mean…what it is, is a group product.’ And so what I did was turn it into a ‘personal’ document management system as the spec, and I got everybody to nod their head yes, we got funded, and the first month that we were out, Mitch approved a change in direction back to the group stuff and that’s when it went back to ‘Notes.’ ”
