RSS Feed

26 November 2008

Squandering a touch screen

It is sad when such a good idea goes to waste. According to David Pogue's scathing review of the Blackberry Storm and my experiences with the Android G1, both phones fail for different reasons. The Storm is completely right to include expandable memory and an 8GB memory card in the box. Yet, it does not include a fast enough processor, making it seem slow. Insufficient hardware support and poor interface design, according to Pogue, are the Storm's main failings. In a touchscreen device, it is very important to have large targets, and to have the interface designed to act like everyday objects, not like a computer screen. However, the insufficient hardware of the Storm makes a touchscreen not feasable, due to the fact that it does not have the animation capability to make the flick-scrolling lists.


The G1 suffers from the touchscreen not being the main feature, contrary to the Storm which eschews the iconic Blackberry keyboard. The plastic touchscreen on the G1 is not as smooth as the glass on my iPhone; I found myself reverting to using the scroll wheel at the bottom. Including a scroll wheel and keyboard show that HTC, T-Mobile, and Google are not willing to commit to a complete touchscreen experience, and the touch screen probably was not the main feature of testing. Both phones also suffer from poor hardware support; I don't think that the G1 can top the iPhone's 11MB of video RAM, and with five second delays, the Storm sure can't.


All touchscreen phones need to have some perspective on what makes the iPhone great: It is not only the touchscreen, but the touchscreen as the only input source, and the brilliant animation support, and the intuitive interface, and the responsiveness. If I was to predict the iPhone killer, I would predict a touchscreen-only phone with excellent hardware running a stripped-down Linux OS with compositing for animations. And it definitely would not be written in Java, even in compiled Java (way too slow for mobile devices; try a more static language.) Oh, and it will probably cost somewhere around $500.

No comments: