It's been a few years since Hawkins and his team worked out the details of what would eventually become the Palm Pilot. It was an ultra-simple device that flew in the face of previous attempts to create a PDA. Most were bristling with ports, big screens, keyboard adapters, expansion slots, trying to replace both desktop PC and clipboards at the same time. But Hawkins, who had already tried those approaches, emphasized simplicity, to just beat paper and be an accessory to the PC, and his strategy paid off. The little Pilot did only a few things, but it did them well, and customers embraced this new tool.

Over these last few years, many good ideas went into the wastebasket along with most of the bad ones. Ideas for clever expansion ports, more software functionality; but the visionary held the line and kept the products simple. A few critics were disappointed that the Palm V was so little different from the Palm III, not vying enough with Windows CE, but most journalists loved its size and look and customers went nuts. The Zen of Palm was established and would be rigorously maintained. It has remained at Palm like a stern father, influencing design decisions long after Hawkins' departure from the company.

Changes did finally come. First with Hawkins' new company, Handspring, which introduced the Springboard slot. Palm eventually followed with its SD expansion strategy, one which Handspring has now joined. Expansion, eschewed in the beginning for austerity's sake, had broken the stodgy old father Zen down. Palm also introduced a few careful and conservative enhancements to the basic applications, but the "leave it out" principle still governs the company; as such, many of Handspring's enhanced applications still offer greater utility and customizability than Palm's core applications. Palm has been more conservative than its creator.

As of this writing, Palm OS 5 has been shipped to hardware and software developers. There will apparently be some strong new features, but most of the core apps, like the Date Book and Address Book, will remain the same. Rather than running in ARM native mode (the new, far more powerful processor that will run the new OS), these apps will run in emulation. That means the powerful processor will be running a program that will make the old apps think they're running on a Motorola 68000 Dragonball processor. We've seen this before, Palm points out, with Apple's transition from the 68000 processor line to the Power PC, which began in 1994. But from what I recall of that messy and frustrating transition, it's not an example they should have used. It eventually worked, but it was a buggy nightmare for many.

Palm's evolution is following a similar pattern to the PC, both Mac and IBM clones running Microsoft DOS, but mostly the latter. Both platforms started out simple, and over the first few years new features were added very conservatively. But nothing significant happened until Microsoft broke away from IBM and the new operating system the two had been working on, OS/2. Microsoft declared itself free from the conservative company, and immediately subsequent releases of DOS began to incorporate new features; that's also when Windows finally started to become a viable OS shell to DOS.

Much like the enhanced Handspring versions of the Palm OS, Microsoft's new applications were most often developed by someone else, then pared down, cleaned up, and rebadged by the company. Programs like DOS's Defrag were actually made by Symantec. Just like Handspring's City Time is really made by CodeCity in Australia. There's nothing wrong with this strategy, it's good business sense for the visionary to recognize a good wheel and buy it rather than inventing his own. In the end, what's important is that the product he ships is enhanced for the customer.

Palm has begun to move in this direction, bundling productivity software like Documents to Go. They've even bought a few great companies and made them their own, like Actual Software, makers of MultiMail.

Both bundling and buying are good practices. Microsoft, while drawing criticism, made DOS more stable through their acquisitions and consolidations, and I as a consumer was grateful. I remember what a mess it was trying to keep my computer running with so many different OS enhancements running--enhancements that were essential to even run some applications, but would crash if used in conjunction with other enhancements.

I want to be clear that I'm not advocating hyper-consolidation and featuritis that takes business opportunities away from developers. The Palm OS has been fertile ground for the many fine minds who were discouraged from developing for DOS and Windows because of Microsoft's overzealous acquisitions and outright attacks, and the Palm OS should remain a field of opportunity. I've long theorized that Bill Gates was lured into the current Anti-Trust case by none other than Larry Ellison of Oracle with his and other disgruntled big name developers' talk of the Network Computer that would one day obviate the need for Windows and PCs; users would instead download their applications from the Web and store their data on servers. What a ruse. The PC had long since destroyed distributed computing, so I remember chuckling with incredulity. But Microsoft, who'd shown disinterest in the Internet up to that time, was sucked right in, diving into the Internet in predictable brazen fashion. Whether the Network Computer really was a trap cannot be proved, but alienating developers is a mistake Palm will do well to avoid, if only to keep them in its camp.

I've said a few of the purchases were good, but what I'm more concerned about is Palm's penchant for leaving well-enough alone; I doubt they'll become overzealous anytime soon. I've mentioned it before, but a good example is the Address Book whose limited number of fields are woefully inadequate for syncing with most desktop PIMs. I'm told by Palm that it will not be updated for the first release of OS 5. There will be no new fields to work with PIMs like Outlook. There is no excuse for this.

Backward compatibility is great, and that's what they're shooting for, but the Address Book's limited number of fields is six years old. Running on an advanced PDA it's a wagon wheel bolted to the hub of an SUV. We've gotten this far with it, but it's an uncomfortable and embarrassing ride about town. It's time to move beyond that original vision, driven at the time by expedience. The circumstances have changed. It's time for Palm to change. Gone are the days when RAM in the little Pilot was 128K. And now with ARM, also gone are any horsepower limitations. All we need now is a new visionary. Someone to take Palm to the next level. Is he or she already at Palm or PalmSource? We'll have to see. Upcoming products will tell the tale. I want to encourage those at Palm to lead, looking forward and not back. Well enough didn't move us from the sea to the stars; let's hope a restless soul takes the helm soon.

-Shawn Barnett, sab@pencomputing.com