Building Search Engines is Hard

Having built a search engine from scratch, I can verify that the task is far harder to do right than most people assume, and that's even leaving aside the need to do something cleverer than everyone else with all of the data you're busy gathering and indexing. Rather than go on at length about the precise whys and wherefores, I'll let this ACM Queue article do a little of the work for me; I say "a little" because it addresses only the basic technical aspects of the exercise, while even a stirringly successful development push will still leave the daunting task of commercializing the product. There's nothing to sour one on the competence and vision of the supposed leading lights of the financial world like dealing with venture capitalists who insist on telling you that "search is dead" just 2 years before Google goes public ...