A Real American Hero

Reference designery, and the proliferation of Android

Generally-interesting-guy-I-follow, Hank Williams, talked about the Kindle a little bit in his post on Arrington’s great Kindle idea. In it, Hank discusses the merits of creating a reference design based off of Amazon’s Kindle for other companies to emulate and create a class of “readers.” However, he mentions something about Android which I’d like to comment on.

Interestingly, this is really what Google should be doing with Android. Google is indeed licensing the Android OS to third party phone manufacturers, but by not creating and controling an initial reference design they are leaving important pieces of the design to third parties, in a field (mobile phones) where important design elements can be critical.

I definitely agree with the Kindle software being offered up as an “OS,” but I’m not entirely convinced on the whole reference design concept. In the reader market, the paucity of players provides for an opportunity for Amazon to “set the bar,” so to speak, on performance and quality.

When you’re talking about computers and operating systems, its the chip vendors that are setting the reference design.  The i686 architecture, for example, supports a particular instruction set, communications/bus protocols, and etc.  The motherboard vendors have to adopt certain standards, offered by the chip maker in design guides, to make their product.  The OS vendors have to make the OS work with the instruction sets of the CPU and the various bridges/buses and etc.

Now, when you’re talking about Android, how would the reference design benefit the third party manufacturers? The benefit is already in the fact that the Android contributors have already done the legwork in designing the OS, so all the phone vendors have to do is make sure they pick hardware that can talk to it.

If you were to look at it from the reverse, Google has extraordinarily limited experience in hardware design. Is it really in their best interests to design some “uberphone” that can run Android when most of the phone makers have a pretty good pulse on what the market wants?

Sure, one can argue that the phone vendors don’t have a clue, and that’s why the iPhone is “smashing” other handset sales. There are a rush of copycat designs that try to approximate the iPhone in feature set and functionality, and all hit the mark to one degree or another. However when you look at the global scope of phone sales, the overwhelming leader in mobile web browsing (and probably total handsets sold) is actually a non-smart phone — the Motorola Razr.

I think, to a certain degree, that the “best phone” on the market will always serve as a type of reference design, and the Android OS, in and of itself, will do the same. By being open and transparent and accessible, we will see both a large number of products/apps developed for Android, as well as variations on Android’s components that will make each handset unique in its own way, should the manufacturer choose to do so.