Why open source could kill Android's chances

The fractured fate of desktop Linux should be a cautionary tale to Google and spur it to treat Android like a strong platform, not a starting point

In very Orwellian logic, a couple readers have told me that the lack of gesture support in the Android OS's UI and stock apps was a good thing, because it would let the open source community innovate in individual apps for the Google mobile platform. By leaving such basic functionality as gestures nonstandardized, they argued, the platform would be fertile ground for the open source community to develop untold wonders that would render the controlled platforms such as the iPhone as dinosaurs.

This is precisely why I fear for Android's future. The open source community is much more likely, based on its history, to screw around with umpteen hundred variations that are piled willy-nilly on top of umpteen OS variants, creating a mess that only a few nerds will want to play with.  In other words, we'll end up with a replay of desktop Linux, which in 10 years is still a mess of precious versions and precious innovations, yet lacks basics such as common drivers and interface standards that would let regular people use it and let developers who need to earn a living get enough of a market to bother.

[ Stay up on tech news and reviews from your smartphone at infoworldmobile.com. | Get the best iPhone apps for pros with our  business iPhone apps finder. | See which smartphone is right for you with our mobile "deathmatch" calculator. ]

Google would be stupid to let the open source community turn Android into yet another series of technologist-oriented fantasy fiefdoms sparring like medieval kingdoms. The open source community should either leave Android alone or coalesce into a de facto manager, as has happened with successful open source efforts such as server Linux and MySQL. Of course, that means not being a free-for-all, but a virtual or real "company," as those successful open source efforts actually are.

If the rumors are to be believed, Google gets the problem, which is why it supposedly is planning its own gPhone for early 2010 to set a standard, deeply functional version of the platform that might actually take root.

If it doesn't, we'll see Android fracture. In fact, that's already happening. Motorola has its own developer environment to optimize Google apps to Motorola's devices. And HTC has extended the Android OS with its very nice Sense UI, in essence creating a fork of Android for use in its Droid Eris device. (Yes, I know it's technically not a fork, and that HTC's Sense UI is an overlay, as the first Windows versions were an overlay on DOS. But from a user perspective, it's a different Android.) These emerging variants will forces developers to decide whether to create plain-vanilla Android apps that work at just a basic level across all the variations, or to choose one or more variants and hope they gain enough traction to make the extra work worthwhile. The fact that Google chose to not use gestures in its stock apps is worrying sign as to what will happen.

If that fracturing is not nipped in the bud, there'll be as many Android variants as there are desktop Linux variants, and the platform will have just as much success -- read: basically none -- as desktop Linux.

Of course, if the misguided "everyone do their own thing" strain of open source prevails, that will help the iPhone OS cement its dominance, and perhaps give Palm's WebOS a second chance at gaining market share. The mobile world is lucky that there are some good alternatives out there, and that there's no Windows equivalent to slay before it eats us all.

Don't forget to be part of the InfoWorld Mobile Patrol: Send in your tips, complaints, news, and ideas to comments@infoworldmobile.com. Thanks!

This story, "Why open source could kill Android's chances," was originally published at InfoWorld.com. Follow the latest developments in mobile computing at InfoWorld.com.


Copyright © 2009 IDG Communications, Inc.

How to choose a low-code development platform