Google Android Is Doomed to Self-Destruct
Simply put, it's too much of a good thing. Every few days, another Google Android device is announced, as hardware makers and wireless carriers rally around the mobile operating system as the de facto smartphone platform alternative to Apple's limited-availability iPhone and RIM's limited-capability BlackBerry.
That flood of options should be a good thing -- but it's not. In fact, it's a self-destruction derby in action, as phones come out with different versions of the Android OS, with no clear upgrade strategy for either the operating system or the applications users have installed, and with inconsistent deployment of core features. In short, the Android platform is turning out not to be a platform at all, but merely a starting point for a universe of incompatible devices.
Who wants to commit to a two-year cell contract for an Android phone when it's not clear if a better version will be out next month or if the operating system and apps you put on your Android device will be supported in the future?
Consider what's happened since the Motorola Droid-fueled "reboot" of the Android platform in November: The Droid used Android OS 2.0, as did the HTC Droid Eris. But the Motorola Droid didn't support multitouch outside of a few third-party apps. The Droid Eris did so using HTC's proprietary Sense UI overlay but is underpowered. Meanwhile, users with first-generation Android devices (running Android 1.6 or earlier) were largely out of luck getting the current Android OS.
In 2010, the situation became even murkier. Some carriers updated earlier-version Android smartphones to the 2.1 OS -- but many did not. The 2.1 OS came with Google's Nexus One, what Google called a "superphone" meant to be a standard bearer for the Android platform. Sorry, Droid buyers -- it too lacked multitouch and didn't have a keyboard, plus Google thoroughly screwed up the product support. Next month, Nexus One's manufacturer, HTC, will ship the HTC Desire, essentially the Nexus One with multitouch added using HTC's Sense UI; anyone who bought a Nexus One must feel like a fool for adopting the alleged flagship Android smartphone.
But before HTC Droid Eris and Desire customers gloat, it's not clear what their upgrade path is for apps and the Android operating system itself. Because of the Sense UI (which really should be a standard part of the Android OS), you can't just upgrade the operating system or be sure that your apps will work -- a new OS may break Sense UI, and a new Sense UI may break your apps.
It gets worse. Later this year, Sony Ericsson plans to release its Xperia Mini series -- inexplicably based on Android OS 1.6 and using its own proprietary UI on top.
This mess leaves developers and users in an unstable position, as each new Android device adds another variation and compatibility question. Compare this self-demolishing behavior with Apple's iPhone strategy and Palm's WebOS strategy: a few models each year, with the same operating system applied to all of them and made available for previous models immediately. Both users and developers know that they can count on high, predictable levels of consistency on these platforms. Even Microsoft has realized this: After years of letting Windows Mobile be customized into a mess of incompatible versions, its forthcoming Windows Phone 7 operating system follows the same strategy as Apple and Palm. Even RIM, who for several years had multiple versions of its BlackBerry operating system floating about, has largely developed one-OS-for-all devices. (Other device capabilities remain less consistent.)
You might think that Apple, Palm, and RIM have it easier than Google in enforcing platform consistency, since they all make their own devices. By contrast, the Android environment is open source, with every device maker free to do what it wants. That's a policy problem, not a technical one. As Microsoft seems to have learned with Windows Phone 7, you can enforce a consistent, predictable platform even if you don't make the hardware. Oh wait, that's how the PC industry works -- so it's not so novel a concept.
A couple months ago, I wrote that Google's naive approach to open sourcing Android would result in a free-for-all that confused users and developers. I urged it to treat Android as a managed platform, where users and developers could buy devices and code for them without fear they were on a dead-end fork. The Nexus One at first seemed to be that attempt -- obviously not.
That's too bad, since it leaves Android users in a precarious position that will likely cause a market rebellion. I can understand Google's open source impulse to adopt an approach more like that of desktop Linux in which the community can experiment freely. Of course, desktop Linux is an asterisk in terms of actual adoption. But even its approach is not as bad as Android's; at least with desktop Linux PCs, your PC is not tied to a specific variation of Linux or to Linux at all. Users can at least retain their hardware investment and install a different operating system if they chose a fork that goes nowhere. But there's no such do-over option on smartphones.