Tweetdeck recently published an interesting summary of their testing efforts for a Twitter client for Android. The short post enumerates the set of hardware and operating system versions they had to contend with in testing their software. I counted about 250 different devices and over 100 versions of the OS in the Tweetdeck charts in a population of 36K beta-testers, many of whom, admittedly, are early adopters who are more likely to use wacky devices and odd versions of the OS. But still, that’s a lot of potential wierdness. This proliferation of versions and configurations can be seen as a sign of the vitality of the platform, but it is also suggestive of some potential problems.
While having an operating system capable of running on a range of hardware can increase the market size for applications written for the operating system, too much variability in the infrastructure can make applications (and thus the user experience) somewhat less than predictable. The large number of Android versions and devices make it much less likely that any particular vendor will have tested against (or even heard of) all of the combinations.
This in turn can lead to unstable applications which will undermine the user experience. An open and flexible platform can encourage many developers to contribute apps, but without strong design guidelines and with spotty testing practices, these applications are likely to wreak havoc with users’ expectations.
There are parallels between Android and the PC world of 10-15 years go compared to the software and hardware produced by Apple. Apple’s coherent designs won praise and a small market share, but the PC dominated, driven by enterprise use. The current squaring off between controlled design and unchecked openness may play out differently in the consumer smartphone/PDA market. The current popularity of Android may well be driven by price and a reaction to Apple’s strict policies with respect to everything from software development to the (lack of) choice of carrier.
Whereas large PC purchasing decisions were driven by IT and accounting decisions rather than by end-user preference or user experience, the smartphone/PDA market is considerably different (with the possible exception of RIM devices). Image, viral marketing, and user experience should play a more significant role in this market, and it’s clear that Apple will not let its share erode without a fight. It will be interesting to see over the course of the next couple of years if the Android ecosystem can sustain a good enough user experience without any top-down control, or if the inconsistent interface design and flaky hardware will impact its ability to compete.