I wonder if executive compensation schemes are at least partially to blame for failures like these? After all, next quarter's numbers will look a lot better if you stop spending money on R&D which, if it ever does pay off, won't for a few years down the road. So, if you're running Kodak in 1992 and you don't expect to be the CEO in 2002, why negatively affect your stock options by ensuring the company's future? Investors aren't given a breakdown of R&D spending by expected first payoff year, so no one can tell that your earnings look good because you've sold off the future.