I've been building software with Delphi for nearly 30 years. It shaped how I think about code, about architecture, about what good software looks like. Delphi, and its predecessor Turbo Pascal, were made by Borland, a company that once rivaled Microsoft in developer tools.
At some point I decided to write down the history. That became a book, Pioneering Simplicity. But early on I realized: most of the people who built these tools are still alive. I could just ask them what really happened.
So I sent Teams invitations. Anders Hejlsberg, who wrote the original compiler. Chuck Jazdzewski, who designed the visual component library. Tim Berry, who was on the board. Zack Urlocker, who managed the Delphi launch. I expected polite rejections. Instead, everyone said yes.
They wanted to tell their story. And their story turned out to be different from what I expected.
The myth of the vision
We tell stories about visionaries in boardrooms. Strategic plans. Five-year roadmaps. The reality is messier. The idea for Delphi came together in a parking lot. Chuck Jazdzewski told me how it happened: "Anders and I, in the parking lot one day, came up with this idea." Two additions to the language, properties and method pointers, would enable event-driven programming. Combined with a visual designer and native code compilation, they could match Visual Basic's ease of use with the performance VB couldn't deliver. "We realized that we could do everything that Visual Basic was doing with just those two additions."
Anders Hejlsberg taught himself to write compilers. No formal training. No computer science degree in compiler design. "I had absolutely no knowledge, never taken a compiler course, nothing," he told me. He learned from one book: Niklaus Wirth's "Algorithms + Data Structures = Programs." He wrote everything in Z80 assembly code. The compiler that would sell millions of copies started with a guy reading a book and figuring it out.
The $49.95 price point is legendary now. A brilliant strategy that disrupted the market. Compilers cost $500 back then. Borland sold Turbo Pascal for fifty bucks. But Anders, who had built the compiler in Denmark and licensed it to Borland, couldn't believe it. "We thought you guys are nuts," he told me. "You're going to ruin the entire market." Tim Berry, who was on Borland's board, was equally skeptical. He had an MBA from Stanford. "Why in the world, when you could sell this for $400?" Philippe Kahn, who ran Borland, explained the math: the product cost $2.70 to manufacture, so $49.95 was still a 90% margin. Tim told me: "I was so wrong. The times when I've been wrong are more interesting to me than when I'm right."
Form inheritance, one of Delphi's defining features, exists because of spite. Chuck told me about an internal document from a team Borland had acquired. It claimed their product would always be better than Delphi because Delphi couldn't do form inheritance. Chuck read it and thought: "No, I think I can do that." He did. It shipped in Delphi 3.
Delphi itself almost didn't happen. Chuck and Anders spent a year and a half trying to convince management. Nothing worked. Then the old management got fired. A temporary manager came in from the Quattro team. Chuck and Anders ambushed him after his first team meeting. They pitched the idea. His response: "Sounds good. You guys should do that." No strategic review. No business case. Just a yes from someone who didn't know he was supposed to say no. That's how Delphi got greenlit.
The myth of the company
Frank Borland was the face of the company. A prospector who lived in the Santa Cruz Mountains. A folksy character who made the company feel personal. Except Frank Borland never existed. Tim Berry told me he made him up. "I just, playing for the fun of it, invented some preposterous story about a prospector named Frank Borland." It was a joke. But Philippe Kahn saw the value. He had no developers in the US; Anders was still working in Denmark. Frank Borland became the explanation for the invisible engineers. The mythology served the company well.
Philippe Kahn built Borland into a major company. He's a legend in the industry. But the people who worked with him told a more nuanced story. Philippe was brilliant at marketing and vision, but the credit didn't always land where the work had been done. Anders grew frustrated. He had built the products that made the company successful, but wasn't allowed to shape its direction. Then Bill Gates called with an offer: come lead Microsoft's language development, with full freedom to shape its direction. Anders left.
Behind every visionary, there's someone who makes things actually work. At Borland, that was Spencer Ozawa. He was Vice President of Operations from the beginning, the one who turned Philippe's ideas into a functioning company. I had never heard his name before these interviews. Tim Berry put it simply: "Philippe was the brilliant idea guy. Spencer made all that come together." While Philippe was dreaming up the next big thing, Spencer was hiring people, shipping products, keeping the lights on. He's a footnote in the histories now. He shouldn't be.
The moment
February 14, 1995. Moscone Center, San Francisco. The Delphi 1 launch. The room was packed. Anders opened the Delphi IDE inside the Delphi IDE and compiled it. The audience went wild.
But Marco Cantu, a Delphi author who was at the launch, told me about a different moment. The demo that got the standing ovation. A one-button application. Click the button, it triggers an access violation. And the application doesn't crash. "At the time, that was hard to believe. Any unhandled exception would crash your application." They showed you could debug through the access violation. The whole room stood up. Not for a feature. For a bug that didn't kill the program.
What I learned
The tools that shaped my career were built by people who were figuring it out as they went. Anders didn't know how to write a compiler until he taught himself. Chuck and Anders sketched Delphi in a parking lot. The pricing that revolutionized the industry was a gamble that almost everyone thought was crazy. The feature that defined the product existed because someone wanted to prove a critic wrong.
The story we tell about technology is about vision and genius. The boardroom presentation. The strategic insight. The brilliant founder who saw what no one else could see. The reality is messier. Parking lots and gambles. A temporary manager who said yes because he didn't know he should say no. A feature built out of spite.
The lesson
Zack Urlocker was the product manager for Delphi 1. He described what it took to ship: "It was like running a marathon, reaching the finish line, and then being told to run another fifty miles. Everything was harder and took longer than we expected. But we never lost faith."
At the end of our conversation, he said something that stuck with me: "In the end, it's one of the things I'm most proud of in my life."
I understand why. Not because they had a vision. Because they kept going when it was hard. Because they tried things that might not work.
The best things aren't planned. They're built.
