The need for text input is paramount on enterprise-class mobile devices. Sure, we see people text messaging via telephone keypads all the time, but let's face it, this is impractical for most corporate users. The biggest problem is that even with predictive typing technologies like AOL's T9 (which I've always found rather frustrating), the keypad is just too limited as an input metaphor all by itself.
Or is it? Two really interesting technologies extend the keypad's functionality to the point where the micro-keyboard may become a thing of the past. The first of these is MessagEase from Exideas, which is a software product that enables rapid and accurate text input with just a keypad. The way this is done is to assign a particular letter to one or two key presses (or to tap and drag with a pen on a device's screen). Have a look at their site; they have some great demos, and I've found the technique very easy to learn and use. And you can download and try the software, which has been ported to a large number of mobile devices.
The second technology, called Fastap, is from Digit Wireless. This involves the use of a physical keypad, but with more keys, each assigned to a particular letter of the alphabet. The keys are raised a bit and interspaced with the keypad keys (again, check out their site to see how this looks). The advantage of this approach is that the learning time is reduced essentially to zero, but I find that the keys would necessarily need to be quite small on most phones. Still, no QWERTY micro-keyboard required here, either. You can find Fastap on the LG AX490, available from Alltel.
But maybe the best keyboard is no keyboard at all, or so the proponents of speech-recognition technology claim. I hate to say this, but I have to -- speech isn't going to be practical for mobile devices anytime soon, and perhaps never. And there are two reasons for this. The first is accuracy -- speech recognition remains very difficult. Even if we get to 95% accuracy -- some would claim we're at least that far now, with computer-based systems and noise-canceling microphones, anyway -- that's still a lot of errors that need to be corrected. But mobile speech recognition needs to work in noisy environments, and that's very hard, especially with the built-in microphones in cellular handsets and similar devices. The second problem is less obvious -- security. You don't want to be blabbing corporate secrets into your handset for all to hear. And holding down the level of acoustic noise is today a sign of good manners at the very least. I am, by the way, much more enthusiastic about speech synthesis and in having one's email read over a wireless connection by a synthetic voice.
But, all things considered, BlackBerry Thumb may be on the rise for the foreseeable future, although not necessarily as a result of using a micro-keyboard. Unless you wind up in a doctor's office, that's a small price to pay for the convenience of really mobile computing.
This was first published in November 2006