I had a whole post planned this topic but Ars Technica beat me to the punch. The gist of their article: Google actually pushed out a significant upgrade to Android without actually upgrading Android. They pushed out major changes to location services, game services, music, search, notifications, messaging, and social networking through Google Play and Google Play Services. By doing so, they were able to update these services for about 98% of their audience, without requiring a version update!
According to the Google developer team, they went this route to break their dependency on version updates for their innovations. Now, they can deliver API updates to phones without intervention from carriers or OEMs. Ars asked “is there a light at the end of the [version] fragmentation tunnel?”
What Ars missed, though, was that this approach solves OS fragmentation problems as well. Access to Google Play, Maps, Gmail and other services requires a licensing agreement; that’s why you don’t see these services on other forks of Android. Going forward, these Android forks will need to support Google Play to get these major feature updates. And I’ll bet major new APIs built on top of the low level stack (for example, Google Now cards) will require Google Play Services to work as well.
So, if Samsung wants to fork Android, they can go ahead and do so – but if they don’t include Google Play they’ll miss out on a growing chunk of the Android API that developers will probably be implementing. Same goes for anyone else in China or the Far East that wants to fork Android. And Amazon, you’re already screwed.
Nice job, Ars, in catching an important nuance to Googlel I/O that every other outlet missed.
EDIT: another great post on the subject by Android Central.