Recently Google announced that it will upgrade all US users to Android 2.1. Devices are currently shipping with four versions, so moving to one will help, but that may not be enough to save the platform from fragmentation.The other problem is that there are too many different devices. InfoWorld points out that the Motorola Droid ships with 2.0, but doesn't support multitouch at the OS level while the HTC Droid Eris does through HTC's proprietary Sense UI. However, the Eris ships with Android version 1.5. Even if both get upgraded to 2.1, you'll still have quite a disparity between these two devices. There are dozens of other devices and manufacturers that compound this issue even further.

Buyers that want the latest greatest Android device realistically can't obtain it because there is no greatest. The ultimate device doesn't exist since each manufacturer so heavily customizes it it is impossible to get all of the best features on one phone. Add to that the release dates are so fast that even if you do get a phone you like it will be surpassed in just a few weeks by a different manufacturer with their own features.

That may sound healthy, but it isn't. No one buys a phone and expects it to be cutting edge forever, but no one wants to buy "the latest" at a premium price and expects it to become a distant second within a month or two either. People want some sense of stability. Look at the Apple iPhone. There have basically been three versions since it came out, slightly more if you count differences in storage memory. People gleefully put down cash for that knowing several things:

• Anything they buy now will be upgradable for at least one version and in the case of Apple, likely two and maybe three versions. The only things that won't upgrade obviously are new hardware features.
• There is a tight set of features so it is easy to get the ultimate iPhone - just pick the one with the most storage space.
• Your purchase of the latest is safe for months, maybe even a year, so you will get to use your device a long time without feeling like it is dated, and in that timeframe, you should never feel like you paid too much for an obsolete device.

You get no such satisfaction with Android. Rapid platform releases, disparate hardware platforms and custom manufacturer user interfaces have so muddied the waters it may well kill the platform. Michael Gartenberg wrote of this late last week.

I would argue that to a large degree, this is the same fate that Windows Mobile suffered. Yes, the platform itself stagnated, but that may have been a side effect of device fragmentation. Starting with Windows Mobile 5, Microsoft not only allowed custom features from manufacturers, starting with the Palm Treo 700w, they seemed to encourage it. HTC and Samsung wrote their own user interfaces that so masked the platform, it was hard to even realize you were using Windows Mobile. Microsoft may have rested on their laurels leaving real innovation up to the manufacturers. This fragmentation though hastened the platforms downward spiral. Developers were frustrated that apps written for Windows Mobile might not work on some devices. Custom code had to be written for specific devices, which is a waste of time and resources. Consumers too were confused at what Windows Mobile actually was and seemed to have lost interest.

Microsoft learned its lesson. Windows Phone 7 Series is rigidly controlled by Microsoft, just like the iPhone is by Apple. Unlike Apple though, Microsoft will rely on outside manufacturers to bring devices to market, but gone are the Laissez-faire days where MS cranks out the OS and manufactures do whatever they want.

Google needs to learn this lesson quickly. One possibility is Google can fork the platform and market it two different ways. Android for smartphones should have a very rigid set of requirements to allow device makers to sell it as "Android." Developers could have a stable platform to program to, consumers can buy devices without being frustrated at the vast differences in availability of features on devices currently for sale and upgrades would be more stable. If Google wants to cater to the anything goes market, then they can license it another way that allows the manufacturer to do whatever they want, but the device cannot be called an Android phone. Carriers can make ultra-powerful feature phones with this and manufacturers can cater to specific markets, like industry.

Whatever Google decides to do, it cannot continue down the path it is currently on. Android cannot survive the turmoil.