Jack Schofield is computer editor of The Guardian
Five years ago, the papers were full of stories about how wonderful Java was and Corel announced it would ride "the Java tidal wave" with a suite of office programs written in this trendy language. Did anybody bother to ask whether it was possible to rewrite Word Perfect Office in Java and, if it was, whether a significant number of customers would want it? In fact, as soon became obvious, the whole thing was a waste of time.
Then network computers became all the rage. The NC was poised to sweep away the PC industry. Corel had a hardware division developing the Video Network Computer, announced in 1998 as the Netwinder. Needless to say, NCs were a flop.
By 1999, the GNU/Linux clone of Unix was all the rage, and again the press was full of stories about how wonderful it was. Corel was at the forefront, as usual, promising an office suite for Linux, and its own version of the operating system. It was a big story, and Corel's share price jumped 41%. Once again, predictably, the whole thing turned into a financial disaster.
Corel is now flogging off its Linux business for about $5m (£3.4m). Last year, that was barely enough to buy a box of popcorn with a penguin label.
Finally, last week, Corel rolled out a "new corporate strategy". Was it going to enter the mobile phone market, launch a wireless electronic organiser, or bring out an Internet appliance? No, it was going to try to cater for the needs of its customers - the people who buy and use Corel Draw and Word Perfect. Obviously, this is not a story. Corel's shares slumped.
Of course, Corel is not the only company that has exploited successive waves of hype: just do a Web search on Wap, dot.com, ASP, and so on. But who should we blame - the companies who originate it, the journalists who pump it out, or the readers who swallow it?