Computerworld — Palm is at a critical crossroads. It is not clear whether it can survive, let alone regain its former luster. The company that previously defined the smart-phone market has lost its way, as clearly indicated by its declining market share and lackluster financial performance. Why did this happen, and what can Palm do about it?

Palm has fallen well behind the market with its geriatric Palm OS devices and needs a good "kick in the phone" to get market share back. Even die-hard Palm fans are choosing more modern platforms. To supplement its market presence, Palm has relied on Windows Mobile devices to fill the gap, with some success. But that tactic only takes Palm so far, and it has not produced a major differentiating product for the company, especially in the consumer market. And although the Centro has been popular at the lower end of the market, it can't compete on features and functions with newer smart phones. Palm needs a competitive high-end product to compete with next-generation platforms from Symbian , Android , iPhone and so on, but right now Palm OS does not deliver a user experience or feature set close to what those platforms have.

Palm has been working on a new Palm OS based on a Linux kernel for a long time. It?s at least three years late after multiple slips. Now rumors say that it will be announced in January. This really is Palm?s last chance to come out with a compelling, competitive platform. If it doesn't, the market will continue to pass Palm by, and it will be even more marginalized. That's sad for a company that virtually invented the smart phone, but the reality of the current market is that some highly compelling platforms are available and winning over users.

Complicating matters for Palm is the fact that it may not be able to rely on its shrinking base of loyal Treo fans to quickly adopt any new operating system. Palm faces the difficult task of providing an upgrade path for existing Palm OS applications to a new operating system with a Linux core that is completely different from the previous operating system. Interface enhancements will be essential for this new operating system to be competitive, but it's unclear how backward-compatible it will be. So, Palm may have to entice application developers to its platform, a difficult task with so many other platform choices.

It is still unclear whether Palm will finally announce its new operating system in January as predicted. But it is very clear that it has little time to waste. And if this operating system is not a compelling alternative to the myriad new-generation platforms already in the market, it will become Palm's swan song. That would be a sad ending to a pioneering company.

Jack Gold is the founder and principal analyst atJ.Gold Associates LLC , an IT analyst firm in Northboro, Mass., covering the many aspects of business and consumer computing and emerging technologies.