With all the talk about the deaths of various industries -- from the car companies to newspapers to mortgage brokers -- I'm reminded of the time that my industry abruptly died on me.

It's been more than a decade, so it's easy to forget just how exciting the early days of the Web were. I first caught the bug in the mid-1990s, when I came across a programmer named Philip Greenspun -- a man who was truly ahead of his time. Along with five of his friends, Greenspun had started a Web development company, ArsDigita, at the beginning of the dot-com boom.

And long before blogs, Facebook, and oversharing online became endemic, Greenspun maintained a personal website on which he did something astonishing: Greenspun shared the whole story of the company he was building, live, as it happened. Everything from the big vision to the technical details.

Eventually, these essays formed the basis of a book, Philip and Alex's Guide to Web Publishing. (The "Alex" in the title was Greenspun's dog, a huge white Samoyed.) The final chapter of the book was titled "A Future So Bright You'll Need to Wear Sunglasses." (You can read it, for free, at philip.greenspun.com.) It couldn't help leaving its readers excited about the fantastic new opportunities available on the Internet. Every single industry was going to be turned upside down! New industries would be created! Start-ups would make people rich! Which is really nice, because it's awesome to be rich! And, bonus: It'll never be winter again!

For programmers like me, this was extremely heady stuff, because it meant that, for the first time, we might not be relegated to tasks such as fixing the code on mainframe accounting software. We would have a new job -- designing and building the future itself!

By 1999, everyone I knew was starting an Internet company. The industry magazines (Upside, Red Herring, and The Industry Standard) had so many dot-com advertisers that each issue weighed a ton. And these same magazines profiled some of the stupidest companies I had ever seen, led by kids who were, often as not, entirely incompetent. (All-time stupidest company: CueCat. Google it.) My theory was that if I could start a company and be only partially incompetent instead of entirely incompetent, I would be ahead of the game.

By this time, Greenspun's company had more clients than it could handle and was hiring like mad, and ArsDigita wasn't even the most successful company following this template. There were hundreds of similar Web consulting companies, including a few giant ones. You may remember the names: Razorfish, Scient, Viant, iXL, USWeb, and on and on. Many had wildly successful IPOs, which was perhaps the first indication of trouble.

The immediate problem all those consultancies faced in 1999 and 2000 was that, like ArsDigita, they couldn't find enough great software developers to handle all the new business they were bringing in, so they had to turn away clients. A company I knew of actually auditioned for a chance to hire Scient and was rejected for, well, just not being cool enough.

Yet, even though the consulting companies' revenue was directly linked to the number of developers they were able to recruit, most of them still treated their staff like garbage. They crammed developers shoulder to shoulder at crappy desks, provided them with crappy equipment, and expected them to work crappy hours. Sure, there might be a foosball table in the corner, and the walls were painted romper-room colors to make the place seem full of energy. But really, these companies were terrible places to work.

This, I thought, was my opportunity. If I started a company that treated programmers like talent instead of typists, I would be able to hire the best and the brightest. I had some theories on how to do this and, following Greenspun's example, started writing them up and posting them on the Internet for anyone to see.

And then, I came up with what I thought was another twist. The Web consulting business was great, but it had one problem: limited margins. You could charge only so much for an hour of some consultant's time -- in those days, maybe $200. Some of that went to overhead (say $20) and some to pay the consultant's salary (maybe $70). That leaves you with a mere $110 per hour in gross profit. That's a lot of money, but it paled in comparison with margins in the software industry, in which you can produce additional copies of an application at virtually no cost.

So this was my business plan: We'd start out as a plain-vanilla Web consulting business. We'd look for situations in which we had several clients asking for the same basic thing. Then, using consultants who weren't currently working on gigs, we'd build an application to suit the group's needs. Over time, this product could be licensed far and wide. Eventually, the software side of the business would eclipse the consulting side of the business. That was the theory. Sounds good, right?

Around March 2000, which was, not coincidentally, the very peak of the dot-com bubble, I quit my job as a programmer at an Internet service provider to start Fog Creek. The plan was to spend a few months decompressing and writing the business plan, then take the summer off to spend at the beach. After Labor Day, I turned in the keys to the rented beach house on Long Island, came back to the city, and started working on Fog Creek with my co-founder, Michael Pryor. We signed a big client on our very first day in business and quickly hired a pair of programmers.

Almost immediately, clouds began to appear on the horizon -- mushroom clouds, in fact. Our one big client made it clear it did not plan to renew, and we weren't able to land another lucrative account. The larger Web consulting companies were struggling, too. Razorfish's revenue plummeted from $50.1 million in the fourth quarter of 2000 to $11.1 million in the fourth quarter of 2001. Meanwhile, at ArsDigita, the VCs who controlled the business pushed out Greenspun and brought in new management.

At first, everybody said that the sales cycle was taking a little longer than usual. What was really happening was that all the dot-coms that had hired droves of Web consultants were suddenly running out of money. At the same time, all the Fortune 500 corporations that had also hired droves of Web consultants were deciding they no longer needed to pay snotty 23-year-olds to tell them that They Just Didn't Get It.

Of course, the Web consultancies had by this time invested heavily in recruiting. ArsDigita, which peaked at more than 200 employees, went so far as to promise a Ferrari F355 to anyone who could recruit 10 employees. The prize was never awarded. So when business started to slip, the only thing these companies didn't do was lay off people. Which meant that they burned through their reserves like crazy, paying salaries to programmers who sat idle all day. Within a matter of months, most of these companies went out of business.

One day in May 2001, Michael and I followed suit. We laid off our two employees and then caught an afternoon showing of Startup.com, a documentary about a Web start-up that failed after burning through $60 million in venture capital. I didn't feel depressed until the next day, when I went back to work and nobody was there. Like so many others, our consulting business had gone bust.

Fortunately, we had by that time completed a primitive version of our first software product, FogBugz. The revenue from software licenses that month was modest: It amounted to just $6,463. Our rent was $2,697, and the company's other expenses totaled $2,954. That left $812 for Michael and me to live on. But the important thing was that FogBugz sales were increasing, and we had some personal savings, which gave us the confidence to keep at it.

We were lucky. We started late, and we hadn't had a chance to hire very many people yet, so we didn't burn through cash as quickly as others. And we were fortunate enough to have a software product under development, so that when the Web consulting industry disappeared, we still had money coming in. Because if you can survive the death of your industry, well, you can survive just about anything. In the next issue, I'll tell you how Fog Creek pulled it off.

Joel Spolsky is the co-founder and CEO of Fog Creek Software and the host of the popular blog Joel on Software. For an archive of his columns, go to www.inc.com/keyword/spolsky.