I am continually frustrated by the smartphone industry. Smartphones are, essentially, the convergence of what used to be called Personal Digital Assistants (or PDAs) with wireless phone and Internet capability. They can make phone calls, check email, surf the web, and manage your calendars, tasks, and contacts. They can also usually play video and music, take pictures, and synchronize data with your computer. They are extremely useful, if you get yourself into the habit of using them to their fullest capability, but they are simultaneously frustrating.

There are now four major smartphone operating systems that collectively rule the U.S. smartphone market: Apple’s iPhone OS, Microsoft’s Windows Mobile, Palm’s Palm OS Garnet, and RIM’s Blackberry OS. Another system, Symbian Foundation’s Symbian, is very popular overseas and holds a majority of the non-U.S. smartphone market (although it is argubly one notch below being a ‘smartphone’ OS, being instead a ‘feature-phone OS’). There are also various small-market Linux-based mobile operating systems, and at least two major Linux-based smartphone operating systems in the works but yet unreleased: The Open Handset Alliance’s Android OS (supported by Google), and Palm’s Palm OS Nova (which might not be its official name when it is released).

I have now had two smartphones—a Palm Treo 650 running Palm OS Garnet and an AT&T 8525 running Windows Mobile Professional 6—and each has been disappointing. I knew their drawbacks when I bought them, but bought them because—when campared to their available competition at the time—they were the lesser of many evils. I am eligible for a discounted phone upgrade on November 24 of this year but, like my two previous smartphones, I find that I will likely have to settle for the lesser of many evils once again.

My requirements are fairly simple:

  • Reliable: A smartphone should virtually never need to be rebooted. If my computer can run 24/7 for weeks without a system-wide crash, a smartphone should too.
  • Extensible (and Open): The real power of a smartphone is that it’s a mini-computer. Like you can add software to a computer to make it do essentially anything you want, you should be able to do the same with a smartphone. There should be no gatekeeper telling me what I can or can’t install on my phone.
  • Usable: Even a complex phone, like a complex computer, can be made easy and simple to use if its industrial design and its user interface are each well considered and well implemented. Software and hardware must each work together to create a device that just works, and works well, without a lot of hacks and adjustments.
  • Push Email (with no middle-man): IMAP-IDLE is a standard ‘push’ email protocol supported by most email services and most desktop email clients, and should be supported by smartphones. It is a server-based protocol where changes you make are reflected on the server (and, thus, in any other client you’re using). My phone should be able to talk directly to my IMAP-IDLE server seamlessly and without a middle-man.
  • Real Keyboard: There’s a lot of hoopla about ‘touch’ devices, but the reality is that a hardware keyboard—though admittedly not glamorous—is the best way to input information into a smartphone. The ‘Treo’ form-factor, with a physical keyboard of sufficient size immediately below a screen on the face of the device, is the best form factor for my usage patterns.

I don’t think this is too much to ask, yet nobody has successfully brought all of these features and functions together into a single device. Oh, yes, some have come close, but the perfect smartphone remains elusive. Companies that could easily have created this fictional phone like RIM, Palm, and Apple each produce good phones that fall woefully short in one or more categories.

  • RIM Blackberry: The Blackberry has become synonymous with mobile email, yet—ironically—email is where Blackberry fails my test. The proprietary Blackberry operating system is reliable. The system is extensible and (with the correct settings) allows the installation of third-party software without any approval from RIM (except when the application deals with certain ‘deep’ system functions). The Blackberry OS is surprisingly easy to use, especially given that no Blackberry (yet) has a touch-screen and system navigation is performed entirely with the trackball and keyboard. And, Blackberrys (so far) all have real, physical keyboards. Email, however, is extremely limited. Email delivery goes through Blackberry servers (not directly from your email server to your phone) and only a limited number of emails can be downloaded. Two-way sync between your Blackberry and your IMAP folders is extremely limiting and unintuitive. This would be a wonderful opportunity for a third-party email program, but the only one I know of (LogicMail) does not yet run silently in the background as a ‘full’ email client.
  • Palm OS Garnet: I admit to being a big fan of Palm. Their Garnet operating system is wonderfully simple and easy to use, Palm-built devices (even those running Windows Mobile) have classy, useful features like a physical ringer switch to put the phone in silent mode. There are thousands upon thousands of third-party applications for Palm OS Garnet. Push email through IMAP-IDLE is available with the addition of third-party programs like ChatterEmail. Palms have real keyboards. Where Palm OS Garnet fails though is its unreliable, antiquated core. Garnet has been around, in one form or another, since the early 1990s and has seen few major architectural improvements. While some multimedia and other modern features have been bolted on, the OS still crashes regularly.
  • Microsoft Windows Mobile: There is no doubt that Windows Mobile, which is available in the ‘Pro’ (touchscreen) or ‘SmartPhone’ (non-touchscreen) edition, is powerful. But, as I’ve discussed before, the system requires an incredible amount of hacking to get it working acceptably. On the positive side, it does offer extensibility and the easy addition of third-party software. It also provides push email with the addition of the third-party FlexMail software. Some WM devices come with real keyboards, such as Palm’s WM devices, the Motorola Q, the Samsung BlackJack, and the AT&T Tilt (slide-out). But Windows Mobile crashes regularly. Worse, it is perhaps the most unintuitive operating system ever created for any purpose. WM can be made usable by applying hacks-upon-hacks to repair its shortcomings, and some implementations have been proactively repaired by phone manufacturers (Palm’s customizations for its Treo line have received wide acclaim, as have HTC’s TouchFlo customizations), but ultimately it is a kludge.
  • Apple iPhone: I expect failure from Microsoft; I did not expect it from Apple. As a Mac user since 2001, I was thrilled to hear that Apple was entering the smartphone universe. “If anybody can do this right,” I remember telling my friends at the time, “it’s Apple.” I was wrong. With a Mac OS X underpinning, the iPhone is reliable (or at least should be; there have been some problems initially). The phone is immensely usable, having presented a ‘game changing’ touchscreen interface that is extremely easy to use. The iPhone is the only smartphone I am aware of that supports IMAP-IDLE push email out-of-the-box without any third-party software. Apple’s on-screen keyboard is, however, terrible. Even with learning self-correction built-in, which makes it much better than other software keyboards, typing speed and accuracy still does not approach that which is possible with a real keyboard. Much, much worse though is Apple’s attitude toward third-party software. Software can only be installed through Apple’s iTunes store, and software can only be in the store if Apple has reviewed and approved it. I’m fine with Apple having a store, and I’m fine with Apple keeping some software out of its store. I am not fine with Apple’s store being the only legitimate way to install software on the phone (without hacking it).

So what’s a prospective smartphone buyer to do? Well, we can look toward the future—though, if the past is any indication, the future is likely to come with even more dashed hopes. Blackberry could start offering ‘direct’ IMAP-IDLE support or there could be further development of third-party software like LogicMail. Microsoft could . . . start over. Apple could open their software ecosystem and release a phone with a keyboard. I’m not holding my breath for any of these.

I am, however, cautiously hopeful. The Linux-based Palm OS Nova operating system is likely to bring the ‘zen of Palm’ to a solid technological base some time next year, though there have not been any public demos to-date so we don’t really know what to expect. Google and the Open Handset Alliance are readying the Linux-based Android operating system for a release this year, which is the most exciting thing to happen in smartphones for some time and could very well be the first phone to-market to meet all of my requirements . . . though we don’t know for sure until it’s shipping, and there has been little word (if any) about email support at all on Android—let alone IMAP-IDLE specifically. Finally, the establishment of the Symbian Foundation by Nokia with the stated intent of unifying the Symbian platform and moving it to an open-source license has the potential to improve that system into a full-fledged smartphone OS that may well meet my needs.

Having said that, I am wary of looking to Palm, Google, or Symbian as a smartphone savior. I expected Apple to fix the industry, and was sorely disappointed. Once bitten, twice shy.