In early November 1987, we drove across America in tandem—Kelly in her Ford EXP and me in my Yugo—but it was not the road trip we’d envisioned. In Nevada, the heater in Kelly’s car died, and we wound up switching back and forth between cars so we could both be warm some of the time. In Albany, the Ford’s transmission died. We left the car there to get repaired and drove together, in a race against the clock. I was due in Manchester, New Hampshire, to start work the next day, and the truck delivering our stuff was arriving, too.
Nothing was easy back then. We didn’t have a lot of support, fiscal or moral. One month earlier, when we’d decided to get married, Kelly’s parents were not at all pleased. Mormons, generally speaking, don’t believe in unwed couples cohabitating or having sex before marriage, and they were understandably upset about their daughter. On the other hand, we didn’t think we were doing anything wrong. My parents liked Kelly, but they were quietly wary about two young, inexperienced, and underfunded kids building a life together. My mom never said as much, but Kelly was following in her footsteps by marrying a Romero straight out of high school, and Kelly, like my mom, got pregnant early on. I’m not sure if these similarities made my mom want to urge caution, or whether the similarities prevented her from saying anything because it might have sounded hypocritical. It wouldn’t have mattered much. Once Kelly was pregnant, we wanted to get married. We went to city hall and took our vows and that was that.
I started at Origin in mid-November, the same week Nibble magazine splashed my game on the cover of its new issue. “MAJOR MAYHEM,” read the featured headline, “Complete Type-in Arcade Action Space Adventure.” The accompanying artwork showed our hero inside his rocket blasting out from the screen of an Apple II over the legs of a nasty tarantula. Puffy clouds of smoke—presumably from the rocket lifting off—poured out of the monitor and swirled around the spider and the rocket.
I had mentioned the game in my interview with Origin. Now it was out, and I was here. It was perfect timing, helping me cement a good reputation. My manager, John Fachini, was the lead coder on Ultima V for the PC, and he had programmed for PCs and the Apple II, so I was impressed with him. I had no experience coding PCs, but they were becoming a larger segment of the home computer market. Despite my love of Apple IIs, I recognized that Fachini was a programming badass, and I could learn from him. Fachini was happy to share; he was a great boss who explained the industry to me in terms of salaries and expectations, stuff that I had no way of knowing. For instance, he told me that $30,000 was the top pay for someone who did straight coding. He told me he made $45,000 because he was the lead coder and managing a couple projects. “The more responsibility you take on, the more money you’ll be paid.” He was only six years older than me, and we really clicked. We ate lunch together, discussed games, projects, and programming.
One night in December I was working late at the office and Robert Garriott, brother of Richard Garriott and cofounder of Origin, asked me to help him for a minute. It just so happened that Robert was having a meeting with Sir-Tech Software, publishers of Wizardry. The first Wizardry game came out the same year as Ultima and was just as influential in the role-playing game (RPG) universe. In my personal pantheon, it ranked as one of the most important computer games in history, and now somehow at 9 p.m. I was randomly meeting the company’s cofounder, Robert Sirotek. With Robert was a game writer named Brenda Garno. She was just a year old than me, and she’d started at Sir-Tech as a teenager, running the Wizardry phone hotline. Like me, she was working her way up. It was nice to meet someone else who’d started in the industry as a teenager and who was into RPGs, too. When the meeting was over, she went back to northern New York, where Sir-Tech was based, but she’d made an impression.
I went back to my work for John, being a department of one, responsible for porting—re-coding a game built for one computer system to work on another. I started to examine 2400 A.D.’s Assembly code for the Apple II, and my job was to convert it to run on the Commodore 64’s hardware. It didn’t take long for me to realize there must be a way to move the files from one machine to the other. In 1988, there were no standard interfaces between different computers, no compatible floppy disks, and no cross-platform APIs (application programming interfaces)—the intermediary programs that now allow different systems to share information seamlessly. If I couldn’t move files between the two machines, I had to retype the Assembly language for the gameplay code and then write new code for the input/output (I/O) to make it work on the Commodore 64. Retyping all that Assembly wasn’t an exciting option because the Commodore 64’s keyboard was horrible compared to the Apple IIgs’s keyboard.
I asked Fachini if there was a way to connect machines to move a lot of the code.
“We don’t have anything like that here,” he said. “Why don’t you call the tech team in Austin and ask them.”
I contacted the person who wrote software development tools in Austin, Alan Gardner, and got the same response: “We don’t have anything like that.”
“Well, what do you normally do for porting between systems?”
“We just re-create everything from scratch on the new computer.”
That makes no sense, I thought. I can figure this out. I went to RadioShack and bought what I thought the job would require: a soldering iron, solder, some four-wire telephone cable, and a nine-pin D connector, which plugged into the Commodore 64’s joystick port. For the Apple II, I got a sixteen-pin IC chip with two rows of eight pins that fit inside the computer. Back at the office, I soldered four wires to the IC chip. Then, I soldered the remaining four wires on the other end of the cable into the D pin connector. Once finished, I plugged it into the joystick port and plugged the sixteen-pin IC chip into the Apple IIgs motherboard. Now that the machines were connected—the Apple II I/O port connected to the Commodore joystick port—the Apple could send data to the Commodore. I had to write the code so the Commodore 64 would accept data, but it was a net time win for me. The Apple II was much faster than the Commodore, so I had to figure out what speed to send the data without causing the Commodore to lose any of the transfer.
As I was doing that, I discovered something about the Commodore 64—coding on it was physically painful. Commodore was a calculator company, and the 64 had calculator keys that were not ideal for sustained typing. The Apple IIgs keyboard, in comparison, was a thing of ergonomic beauty, even in 1988. It was low and sleek with shallow keys that didn’t feel like I was pressing down on large buttons. I decided the most efficient course of action was to program the Commodore version of the game on the Apple IIgs, and then just use my new interface to send the program over to the Commodore. Years later, I found out Will Wright, creator of The Sims, did the same thing for one of his early games.
I finished this project in about four days. I showed it to Fachini and explained how it was programmed on the Apple IIgs and then transferred to the Commodore. The next week, I got an 18 percent raise. My salary went from $22K to $26K. I had been there less than two months, and I was thrilled.
Kelly, thankfully, had a trouble-free pregnancy, but our timing for the actual birth was a little off, thanks in no small part to an arctic-level snowstorm that gripped the northeast US on February 13. As two kids from Northern California, we’d never seen anything like it. Concord, New Hampshire, just 20 minutes north of Londonderry, where we had settled, reported thirteen inches of snow on the ground. Snow and ice were everywhere when Kelly’s contractions started getting faster. I rushed out to the car and found the doors frozen shut and the keyhole iced over. I heated the key to slot it in and turned the lock, but the door itself was still frozen. Finally, I put one foot up against the car and wrenched the door open. I got Kelly inside and poured windshield wiper fluid all over the window to melt the ice. We made it to the hospital, and twenty-four hours later, on Valentine’s Day 1988, Michael Alfonso Romero made his entrance into the world.
I was a twenty-year-old father with a new job, but I made it a point to work 9 to 5 and come home and give Kelly some downtime. Unfortunately, Kelly didn’t have a lot of other support. She’d barely been in town three months. She wasn’t working, hadn’t made any real friends, and our families were thousands of miles away.
In April, the project I was hired for at Origin was canceled. The game I was porting hadn’t sold well in its Apple II incarnation, and the company figured it wouldn’t sell well on other computers either, which seemed logical. I switched projects to a game called Space Rogue, a 3D flight-simulation game with RPG elements being developed by Paul Neurath. While I was working on it, John Fachini approached me. He was unhappy about the internal politics at Origin and asked me if I wanted to form a start-up. Whatever politics there were, they were over my head. I loved everything about Origin. He pitched me on forming a game company that would initially code ports and eventually build original games. He wanted me as his cofounder. My response was immediate: Yeah, that would be awesome. As much as I loved Origin, I was fresh off the 2400 A.D. cancellation. What if they canceled Space Rogue, too? I couldn’t afford to be out of a job now, especially with a new baby boy. Running my own company with another person I trusted would be a great learning experience. As a safeguard, I told John I couldn’t leave unless my salary was guaranteed and equal to what I was making at Origin.
I didn’t know when another opportunity like this might come along. Finding jobs in the 1980s, particularly in a fledgling industry, wasn’t as simple as hopping on the internet. Most jobs were filled via word-of-mouth, advertisements in local papers, or listings in computer magazines. Résumés and cover letters were sent by mail, provided you knew the company and its address. So I took the chance I had, reasoning that it would be in our best long-term interest. I was willing to work hard to make sure we succeeded.
As it turned out, that “next chance” happened just a week later. Paul asked if I was interested in cofounding his new start-up after I shipped Space Rogue. Paul’s studio, Blue Sky Productions, went on to create many seminal titles, most notably Ultima Underworld. Sometimes, I do wonder what might have happened if I had been paired with their amazing programmer Doug Church. Why didn’t I join Paul? Two reasons. First, I gave John my word, and he had already started talking with potential clients for two ports. Second, I hadn’t worked long with Paul and didn’t know him as well as John. There was also a third reason, I suppose. Paul used the ORCA/M assembler, and I used Merlin 16, a much faster 6502 assembler. It took hours to assemble Space Rogue because it used ORCA/M, and I never understood why he wouldn’t switch to the faster assembler. Paul used a lot of macros in his Assembly language, and macros take time for an assembler to interpret. Although its name was macro backward, ORCA/M was far slower at processing macros than Merlin. It’s an odd reason not to join a start-up, I suppose, but these little details matter to programmers.
I left Origin without meeting my idol Richard Garriott, who worked out of the company’s Austin office, just seven months after landing my dream job there.
The new company, Inside Out Software, began with promise. We got a business loan to buy John a 20Mhz 386DX, a state-of-the-art PC, for $5,000, and John signed a porting contract to translate the Apple II version of New World Computing’s Might & Magic II to the PC and Commodore 64. I started working in John’s apartment on his Apple IIgs and, after a couple months, we rented space down the street from Origin’s offices.* Things were going great, and not just for the business, either. My wife was again pregnant, and we looked forward to expanding our little family. She missed her family back in California, though, and the pregnancy only increased that longing. She supported this New Hampshire journey of ours, and our new company, and we hoped for the best.
Before long, Inside Out Software had a relationship with Infocom to finish developing Brian Moriarty’s Timesync (and move the Mac game to the PC), and to create an original game based on the 1979 movie Alien. We also got a conversion contract to move a great Mac game called Dark Castle to the Apple IIgs. With more conversion work coming in, we decided to hire more staff, a couple of coders to finish my Might & Magic II port, and I started on a new port called Tower Toppler (Commodore 64 to Apple II). For Timesync and Alien, we hired two artists. As luck would have it, one of my friends at UpTime, Apple programmer Lane Roathe, told me he was looking for a new gig. Both Lane and my old editor at UpTime, Jay Wilbur, said that UpTime was hitting the skids. Lane immediately accepted our job offer and moved to Londonderry.
Lane and I had a real bond. We both loved programming, games, and Apple IIs. Lane, though, was much more of a purist than I was. He hated PCs with a vengeance, and his contempt rubbed John the wrong way. John was agnostic on the PC versus Apple question. It didn’t help that John was coping with business growth issues at the time and didn’t want Lane’s negativity in the office. In less than a month, the relationship between my partner and my pal became toxic. Eventually, Lane had had enough and quit, which was a fatal blow for the Dark Castle port. We needed an Apple IIgs programmer, and Lane was a total pro. Adding fuel to the fire: John delayed paying Lane’s final check, and Lane filed a complaint with the Better Business Bureau, something no start-up company wants. Then Lane placed a call to Three-Sixty Pacific, the company that hired us to port Dark Castle.
“I’m the guy who was handling your game, and I just quit,” Lane told them. “How about you guys transfer the contract to me?”
This sent John over the edge.
“You brought this fucking guy into the company, and he’s screwed us!” he yelled at me.
“Don’t get mad at me because you drove him away, and we don’t have the person to do the contract.”
“He just stole our client!”
“John, if you think about it, Lane actually just did us a favor. We don’t have anyone to do that work. If he didn’t take it, we would have blown our deadline, for sure.”
Things were tense in the office after that, and in February, they got grimmer when the Tower Toppler client canceled their porting contract. I had been in the industry for fifteen months, and this was now the second canceled project I’d worked on. Was this normal for the industry? I wondered. The answer was no. I realized that the computer industry was undergoing a shift. An entire generation of computers that limited a unit of data to eight bits was about to go the way of the dinosaurs. Computers like the Apple II had lasted longer than they should have—most likely because they were the first generation of a new technology. The PC, introduced in 1981 by IBM, had a 16-bit processor. The additional power and lower cost of these machines translated into growing sales. The market for the computer Lane and I loved was shrinking and destined to become obsolete.
I knew I needed to bite the bullet and become fluent in PC programming as quickly as possible. That old feeling and fear of falling behind returned. John was stressed by financial pressures and his own personal issues. I decided to seize the initiative.
“John, my project is canceled,” I said. “That was money we needed to pay me and run the business. I know we’re partners, but I think this is going to be easier for both of us if I sign over all my shares of the company to you, and I’ll just move on.”
That marked the end of my first “real” start-up. I had tried to make my own luck and failed within nine months, largely due to factors beyond my control: the computer industry’s evolution—which I probably should have seen coming—and the personality clash between John and Lane. Both these things were valuable learning experiences, as all failures are. In the future, I told myself, I need to think about game development from a broader perspective, one that factored in technology trends like increased processing power and the new computers that really drove the industry. I also needed to safeguard any companies I founded by making sure our revenue stream was diverse and not able to be so easily wiped out (this desire to grow the business—to have multiple streams of income—influenced me heavily during the later id Software years). I also needed to remember that not everyone had grown up with an Alfonso Romero or John Schuneman in their life. I survived by trying to avoid creating friction with others and steering clear of disagreements that didn’t involve me. That was one of my survival strategies. I needed to either pick better partners or be a bit more proactive in helping iron out workplace differences.
Tower Toppler’s toppling created a cascade of issues at home, too. Kelly was eight months pregnant now, and more than anything, we needed stability. Life in Londonderry was lonely for her without her mother and father, and she wanted them around when she gave birth. So she moved back to her parents’ house in Folsom, California, to have the baby while I figured out our next steps. Michael, of course, went with her.
Just fifteen months earlier, I’d been on top of the world, living out my dreams and planning for our future. Everything seemed possible. Fortunately, I didn’t bottom out into despair. I just needed to make something happen. My other survival strategy—solving problems to eliminate friction—needed to kick in.
* We heard that Robert Garriott blocked us from renting next to Origin so we wouldn’t hire its employees.