[Scott R]

Well, this may be the first Palm-bashing article from Ed that I can't disagree much with. Things certainly look grim for PalmSource.

Ed, I'm not completely sure that you got your dates right, though. I was under the impression that PalmSource bought a usable OS from CMS and just needed to slap a skin and some Palm-compatible PIM-beaming capabilities to it. So, that feature-phone platform could be much closer to being ready. Of course, they didn't talk much about it at the developers conference because there would be no point. It's not based on the Palm OS and it's not even an open development platform. It's, as I understand it, just as I described above: the Palm GUI "look and feel" along with compatibility for sharing PIM data with real Palm OS devices.

Cobalt seems all but dead and I think the simple reason there was that it didn't bring enough new to the table from a tangible feature perspective (for end-users) while introducing enough compatibility problems to cause a lot of work for developers to get their apps working on it. As a result, PalmSource's biggest licensee (palmOne) stayed put. This makes perfect sense given the newer management at palmOne which is more Handspring influenced. Handspring stuck with v3.5 of the OS on their PDAs while other licensees had long moved on to v4.x.

The OS formerly known as Palm on Linux seems to be a long ways off and is the mid-2006 deliverable that I think you confused with their feature-phone platform. Of course, while I believe that the Palm-skinned feature-phone platform may be close to being ready, I'm skeptical that PalmSource will meet the mid-2006 date for their "this is really the one guys" Linux-based product.

One problem that PalmSource faces is that they think too highly of their "world class" GUI. I've always been a big fan of their GUI but their GUI doesn't translate so well to smartphones or landscape-native PDAs. I think they recently realized that and that's where their "Rome" project comes in, which aims to take a fresh look at how the GUI should work non-traditional-PDA designs (with most of the focus likely being on smartphones).

As you said, the current state of Garnet on feature-rich devices is quite the mess. Where the Palm OS was once considered rock-solid stable compared to the PPC, the roles have now reversed. I recently started coding an app for the Palm OS and have had to deal with all sorts of compatibility issues between various 480x320-capable Palm OS devices. The app will work fine on my Zodiac, run upside down on the LifeDrive, and not run at all on the Tungsten T5. Oh, but it runs just fine on the desktop simulators for all three. I think the problem here is that in allowing licensees to tweak the OS to their heart's content, the situation for developers has grown uglier and uglier. It might not have gone over well, but I really think they should have required all licensees to use flash ROM and for PalmSource to OK all code changes and release updates that end-users could install, keeping everyone "up-to-date."

While I can't argue with your doom and gloom predictions for the Palm OS, I still see PPC/WM as severely lacking on the usability and GUI front. After numerous OS revisions, there's still too much screen real estate wasted on window elements (and now they've taken away more space by adding those two soft buttons), MS's official design guidelines still say it's wrong to use a real decent-sized OK button on a form and instead tell developers to include that horrid little round OK button in the title bar, there's no task switching "in the box", and the PIM apps still require far too many taps (and have other usability issues) as compared to even my ancient USR Pilot 5000's PIM apps.

As I've said before, when it comes to modern PDAs and smartphones, it's not a matter of whether the Palm OS GUI is better or worse than WM's. The truth is that they both suck. I'm always on the lookout for someone new to jump into the arena with an all-new high-end-PDA and/or smartphone platform with a designed-from-the-ground-up GUI and easy and open developer tools, but it still hasn't happened. I think that MS may get there, though. They've got a large enough userbase and group of licensees such that feedback concerning the GUI issues will get louder and louder and they'll hopefully put some real money and support behind fixing/revamping things.