The Super Synergy Effect
Riding back in the car after the meeting with Electronic Arts, Yoo Jae-won recounted every detail to his colleagues. The reaction at ID Software was explosive. John Romero, ever the excitable one, stomped his feet and cheered loudly. Their enthusiasm was entirely understandable. A fledgling studio like ID Software faced countless obstacles. Making a great game was hard enough, but finding a way to sell it was another battle entirely. Most developers ended up leaving copies at local computer shops or uploading demos online. The real dream was landing a major distributor. Large publishers printed enough copies for nationwide release that even a modest seller could turn a profit. Securing a deal with Electronic Arts or Activision was the ultimate goal for any new developer, and ID Software had been no exception. Thanks to Yoo Jae-won, they had just cleared their biggest hurdle. The team felt they had nothing left to wish for.
“Let’s eat first!” Though the youngest person present, Yoo Jae-won exercised his authority as president and led Choi Kang-wook and the ID Software staff to the finest restaurant in Dallas. The place was usually booked solid even on weekdays, but he had made reservations on their first day in America. He treated John and his colleagues to the best steaks until they could eat no more.
“Unbelievable.”
“What did you call this technique again?”
“Image interpolation, obviously.”
“More precisely, it’s anisotropic texture filtering.”
John Carmack and the rest of the ID Software team crowded around the computer Yoo Jae-won was using, chattering excitedly. After a hearty lunch at a downtown restaurant, the developers threw themselves back into their work. Though they were all in their early twenties, they knew their roles better than anyone. Yoo Jae-won joined them. His visit to ID Software was partly to build rapport with John Carmack, but also to help them overcome the technical obstacles they were facing. He sat at John’s machine, reviewed the code, pointed out bugs and suggested improvements, and guided them toward the right direction. None of it was particularly difficult for him. John simply hadn’t fully awakened to his own genius yet.
The most pressing problem at ID Software was handling image scaling. The current stage in Wolfenstein was a 3D dungeon that the player could freely explore while controlling the protagonist, the Super Soldier. The dungeon wasn’t built from polygons, and the objects weren’t 3D models but images. At the time, rendering large numbers of polygons simultaneously was only possible on Silicon Graphics workstations. The issue with images was that pixels became comically large up close, destroying visual quality.
Yoo Jae-won’s solution was a simple image-processing technique. By inserting intermediate values between pixels, he prevented the screen from turning into a mosaic when the player approached walls, floors, or ceilings. The improvement was dramatic. However, the method demanded significant CPU power, so frame rates dropped noticeably on 286 machines.
That was not the only new technology he demonstrated. He also showed an image-compression method that took advantage of the unique memory architecture of video cards, improving both image quality and loading speed. He introduced dynamic background music that intensified when enemies approached and calmed once the danger passed. From hardware scrolling to techniques unimaginable in the 1989 game industry, he unveiled one secret after another. With each revelation, John and his colleagues were left speechless. Only the method for accessing memory beyond DOS’s standard 640KB limit drew puzzled looks. They saw no need for it when 640KB was considered sufficient and machines with more than 1MB of RAM were rare. Yoo Jae-won knew that as graphics quality and scale increased, 640KB would quickly become inadequate. In time, the industry would struggle with EMS, XMS, DOS4GW, and countless other workarounds. Thanks to the techniques he shared, Wolfenstein’s visuals and gameplay would far surpass what they had been in his previous life.
“Computer hardware is already several times more powerful than arcade boards, yet computer games still can’t match arcade visuals because these techniques were never standardized. The few people who knew them kept them secret. We’re going to change that.”
This was the true purpose of his visit.
“You can’t be serious.”
“What if we bundled hardware scrolling, video-memory compression, sound control, and everything else into a single library and released it publicly?”
In 1989, computer hardware manufacturers and software developers had almost no common ground. Each manufacturer used different drivers and supported different features. Developers had to write separate code for every piece of hardware. Unpopular cards received poor support, and some features simply failed due to compatibility issues. Even Keyboard Warrior’s hardware scrolling depended on the specific video card. John was stunned. Normally, anyone with an advantage kept it hidden to make their own games stand out. The smooth scrolling in Keyboard Warrior had been a major reason for its success.
“Wouldn’t that just be helping everyone else?”
Yoo Jae-won smiled at John’s reaction.
“It might look that way at first. But these techniques will naturally become obsolete as technology advances. New hardware and new operating systems will appear, and everything will have to be adapted again anyway.”
John nodded. He had dealt with compatibility headaches every time the system changed while developing games for the Apple II at Softdisk.
“On the other hand, if there’s a standardized game-development library, hardware makers can design their products around it. Software developers will create better-performing games through the library. The user base will grow much larger. Everyone—component manufacturers and game developers alike—will benefit.”
“I see the necessity,” John admitted. Still, one question remained.
“But what do we actually gain? It feels like we’re just giving away our secret techniques and spending time building a library for everyone else’s benefit.”
Yoo Jae-won wagged his finger.
“Will computer technology stay exactly the same a year or two from now? New technologies keep emerging, so the tricks we have today will lose their value over time. Even if we release them as a library, we’ll have to update it regularly to match progress.”
“That sounds like a lot more work.”
“In exchange, we keep the copyright and control of the standard. We decide which techniques go into the library and which ones get phased out. Hardware manufacturers will have to support our library, and software developers will use it.”
Hardware makers were already begging for the technology in Keyboard Warrior. Yoo Jae-won could now turn the tables and demand concessions from them. Implementing anisotropic texture filtering directly on the VGA card, for example, would reduce CPU load. Image compression and turbo memory structures could also be moved into hardware. His words resonated deeply with John and the ID Software team. He wasn’t finished.
“Once the library is complete, we’ll release it for free. Later, we’ll include it in the next-generation GUI operating system that will replace DOS.”
John and his colleagues stared at him, mouths agape. The full scope of his vision was finally clear.
“As you know, DOS is extremely inconvenient and can’t fully utilize a computer’s capabilities. The 486 is coming soon, yet DOS won’t be able to harness its power. Microsoft shows no sign of replacing DOS. ID Technology intends to fill that gap.”
John finally understood. A library alone would make development easier and improve compatibility, but it wouldn’t directly benefit Yoo Jae-won or himself. A next-generation GUI operating system changed everything. The biggest weakness of any new OS was compatibility—there would be no hardware or software supporting it at launch. It could never break DOS’s dominance. With a library, however, developers could quickly create compatible software and hardware makers could adapt easily.
Yoo Jae-won had prepared even more. Office applications that would serve as killer titles for the business market were already in development, and the Ribbon Interface from Keyboard Warrior would be integrated into the new GUI.
“A GUI operating system? Like Windows or the Amiga OS?”
Yoo Jae-won rejected the comparison outright.
“Of course not. The kernel will be based on Unix, and the GUI will use the Ribbon Interface from Keyboard Warrior. The codename for the operating system is Android, and the library is called Glide X.”
John was stunned. Every piece fit together perfectly. He had already found the Ribbon Interface novel and convenient when he tried it. If it became the foundation of an operating system, even complete beginners would find computers easy to use.
“By the way, how many copies of Keyboard Warrior have sold so far?”
“One and a half million.”
“Good heavens.”
With 1.5 million people already familiar with the Ribbon Interface, Android had that many potential users from day one—and the number would only grow. The library, the interface, the office suite, and the new operating system together created an enormous synergy effect. If everything went according to plan, Microsoft would be driven out of the PC operating-system market.
“You’ll be handling Glide X as well, won’t you?”
John Carmack’s perspective had completely changed. Being entrusted with the project now felt like an honor. Despite his pride, his résumé consisted only of a few games developed at Softdisk. Receiving such a major responsibility was almost unbelievable.
“It would be an honor,” John and his colleagues answered in unison.
Yoo Jae-won stayed in Texas for two more days, passing on additional tips and know-how. He even implemented several of the features he had described for Glide X. Choi Kang-wook and Robert Halley assisted with documentation. They drafted nondisclosure agreements that John and the rest of the team signed, and Yoo Jae-won provided a roadmap for Glide X’s development and release strategy. Afterward, he returned to San Francisco, appeared on PBS’s Computer Chronicles, and delivered another shock to the nation while promoting ID Technology and Keyboard Warrior. As a bonus, he effortlessly cleared the infamous final stage of Keyboard Warrior before heading home to Korea. With the stage perfectly set for 1989, the return flight felt far lighter than the journey to America. All that remained was to create the works that would trigger the synergy effect. Through the airplane window, the golden world outside looked close enough to grasp. I can do this. Without the slightest hesitation, Yoo Jae-won reached out and seized a handful of the golden sky.