In 1997 the concept of storing thousands of contacts, notes, appointments and "to do" lists on a calculator sized device was absolutely cutting edge. If such a device could also function as a platform for applications from Independent Software Developers, all the better, and add the funky ability to share the programs, contacts, and appointments through infrared 'beaming' and it sounded like the stuff of sci-fi novelsat least in those days. And yet, this marvel is exactly what the PalmPilot brought to the world.

What made many of us enamored with the PalmOS was that, unlike the Windows OS that most of us used on a daily basis, there was no BSOD (Blue Screen Of Death), no crashes, little waiting, and little complexity. The PalmPilot did a few important things, and it did them VERY well.

The PalmPilot also introduced most of us to something new: synchronization of data across multiple devices. With the PalmPilot, users did not need to enter data through the restrictive UI of the handheld device, but could instead use a partner program on the PC to take advantage of the existing data of a user, and the keyboard and screen on the desktop. Information (contacts, appointments, etc.) from the PC could then be easily synched to the PDA. Information updates, new appointments and contacts received from beaming would also be synched from the PDA to the PC. For people used to "triple tapping" contact information into a cell phone, synch was fantastic.

All of these revolutionary advances in handheld computing launched the PalmPilot and its unassuming OS into the minds of tech users everywhere, and well ahead of any competitors. How, then, did the PalmOS stumble so badly by 2005? Why is it destined to fade while other later OSes rise? Why has Palm finally chosen to use a Windows OS on the Treo 700w? The answer is simple: the PalmOS is a buggy whip in an age of automobiles.

How is it a buggy whip? Well, the original value proposition for the PalmOS was: do a few important things well, with few bells and whistles. Contacts, To Do, Calendar, and Text Memos. The screen? Black and white (er'green). Simple tasks for simple processors and memory limits. When Microsoft came out of the gates in 2000 with PDA efforts, they tried to be too much: color, office documents, PDA and mini laptop form factors, media, games, sound, etc. The upshot was that while the PalmOS did a few things well, the Windows CE devices did many things poorly.

But just because the PalmOS was well positioned for 2000, doesn't mean it is well positioned for 2006. Because of Moore's law, as time passed, PDA hardware became more and more capable. The Microsoft OS's abilities seemed less and less like foolhardy ambitions, and more and more par for the course. The PalmOS's abilities seemed passe. By 2003, Palm suddenly found itself playing catch up by adding color, larger memory, faster processors, more sound functionality, support for external memory, support for external devices (ex: Wi-Fi), higher screen resolutions, greater and greater media support, e-mail, IM, web, Office apps'until the PalmOS became a copy of Microsoft's OS, which Redmond had been steadily improving for 5 years.

Over those years, the table stakes for being a compelling PDA changed. Customers wanted the greater features of the advanced PDAs, and were no longer satisfied with the original simplicity of the PalmPilot. The PalmOS had to up its ante.

And a funny thing happened over those years. It was subtle, and thus hard to notice. Like a child growing up, you don't see it grow. And as a Palm user, I don't know when it started'but my Palm devices started to crash. Under the weight of so many layers of improvements and enhancements, the PalmOS lost its simplicity, and with it, lost its stability. My Treo 650 crashes more than Windows95 with an IRQ conflict (If you know, you know). But my PalmOS Treo doesn't even give me the BSOD, it just neglects my incoming phone calls while it runs some endless loop. Somewhere in those years, PalmOS lost its innocence, and started playing feature catch-up with Windows Mobile. But it turns out that Windows is better at being Windows than PalmOS.

Last year, Microsoft finally enabled non-volatile memory for their PDA OSes, which means that at last, when their devices crash, you don't lose all your data and need to synch to the PC to restore a backup. This was the last critical feature that the PalmOS had over Windows Mobile. In fact, the PalmOS is now notably less capable in that it doesn't have a multi-tasking capable run-time environment (it can't do any tasks in the background).

The unfortunate thing for consumers is that instead of converging towards the PalmOS's original stability, the market has converged towards increased functionality at the cost of stability and reliability. Now they all crash and hang.

If you're wondering where Symbian, the Apple Newton, etc. are in this article, I just kept it to PalmOS and Windows to keep it brief. Symbian is also one of the modern winners in the Smartphone OS market, owing to its good functionality and its strong relationship to phone handset vendorsbut it is more widely used outside North America.

To conclude, when looked at in hindsight, the PalmOS was visionary, but it was an OS designed and built for the hardware of the last century. Unlike MSFT, which designed a complex PDA/Smartphone OS and then waited for the hardware to catch up, the PalmOS has been surpassed by the hardware, and shall become a vestigial technology. Thanks PalmOS for the good times, and for putting the pressure on the other handheld OS vendors. You'll be remembered fondly.