The real beauty of open source software isn’t that it’s free; it’s that it’s free to change. Developers can tinker with it, strip it down or build it out, depending on their wants and needs.
In the case of Google’s Android, this increasingly means that we don’t have one Android operating system. Instead, we have a family of different Android forks and flavors.
Even to call Android “fragmented” assumes that it was or ought to be unified and singular from the beginning. It makes more sense to start talking about “Android-compatible” devices, rather than Android.
Let’s start with a basic contradiction. Even though it’s open source and free to modify, Google keeps Android’s development tightly under wraps until they’re ready to show it to the world. In the case of Honeycomb, Google even held on to the source code powering devices already in the market.
Because Google controls Android’s development and contributes its name, its software, its services and its prestige to Android, we treat Android like it’s Google’s operating system for mobile devices, in the same way that Windows is Microsoft’s operating system for the desktop. More to the point, we treat the OS as if it’s one thing. It isn’t.
Right at this moment, we have brand-new, state-of-the-art, highly evolved mobile and portable devices in a range of form factors that are based on at least three different major releases of Android. Support for updating existing devices to the newest version of Android is all over the map, even for devices made by the same manufacturer, like Samsung’s Galaxy series of smartphones and tablets.
Tablets with fully skinned, proprietary operating systems like Amazon’s Kindle Fire or Barnes & Noble’s Nook Color/Nook Tablet only amplify the diversity that was already present in the smartphone market.
ZDNet’s Ed Bott argues that for manufacturers and carriers, supporting indefinite at-will updates to the latest version of Android just doesn’t make economic sense:
The problem with Android is all that freedom, which allows hardware makers to take the OS and do whatever they want with it. It is inevitable that that freedom will produce a plethora of devices. Some of them will be incapable of running a new Android update. In other cases that upgrade will require significant engineering investments—time and money—on the part of the handset maker and the carrier. They might decide to spend the money and deliver the update, six months later. Or they might decide that the investment isn’t worth it.
Unless Google lets everyone in on Android’s development process, it most likely never will be. And for many manufacturers — Amazon comes to mind — there’s no real reason either company would even want to be that close to each other.
Still, every single one of these companies gains a tremendous amount through their use of Android. Not only do they get a first-class mobile OS for free (or a licensing fee) without having to develop the whole thing from scratch, they get access to a gigantic user base that guarantees an active and engaged developer community.
In short, they get the apps. Not every Android app will work or work perfectly on every Android device or OS flavor, but the vast majority will; others still can be easily tailored to work well or even better.
This is why I say that we have a huge number of Android-compatible devices. We’ve never really had anything quite like this before in mobile. That compatibility is incredibly powerful.
This shouldn’t be a surprise; just look at the history of personal computing. It’s too easy to say that Android is like Windows and Apple is like the Mac — partly because it’s A vs B, and partly because of the relative relationship at both companies between hardware and software.
But that analogy doesn’t actually work so well. In the Wintel era in the 90s and 00s, Microsoft (and Intel) had much more control over how its machines worked and how its software would run on them. Windows is a “strong OS”: you think of Windows first, and then the hardware. Android is a “weak OS” that recedes into the background. It’s not even comparable.
Android today is more like the PC-compatible computing of the 1980s — only with Google playing the role of both IBM and Microsoft. (Apple is, well, Apple.)
With the Nexus line, Google has its flagships, its models of how its operating system should work. The broader world of Android is more like MS-DOS — the original “weak OS.” Every version of the software is retailored and rebranded by its manufacturer. Even some software has to be rewritten to conform to different machines and their varied specs. But what you get is accelerated development of hardware and broad interoperability between machines, which benefits both users and developers.
Why does this distinction matter?
- If you’re an Android user, you may simply have to accept that you didn’t buy an Android device running Google software; you bought a device from Samsung, Motorola or Amazon running Samsung, Motorola or Amazon software. Unless you have a Nexus, Google’s more or less invisible on your device. Finally, you’re never going to be able to just pull a brand-new software update out of the cloud and stick it on your machine. Not unless you’re up for a little hacking.
- If you’re an Android manufacturer, guess what? You’re in the software and support business! That’s right, the exact business most of you were trying to avoid getting entangled with by opting to run with Android in the first place. It might have seemed like a good idea to throw together a custom UI or some specialty software on top of Android, but now you’re on the line to handle future updates. Before, who cared? It seemed like customers just wandered into carriers’ stores and ran the phone software the two of you gave them. But now, Google and Apple and even you, the manufacturers, have made customers much more shrewd and certainly more aware of their choices. And increasingly, they’re expecting continued software updates to be part of the package. Following Bott’s analysis of the economic logic manufacturers follow, it seems likely we’ll see manufacturers either giving users the shaft or doing more to future-proof their devices by building both hardware and software for the long haul.
- We don’t actually have a software company in the mobile space that’s analogous to the “strong OS” of Windows for PCs that can run unaltered on a wide range of machines. I think this creates a gap to be filled. Maybe Windows Phone will become something like that over time. Maybe Google will morph into something like that in future versions of Android. Maybe a frenemy like Amazon will come along and snatch the Android-compatible market out of Google’s hands, the way that Microsoft did to IBM. Or maybe this is a model that doesn’t make much sense in the post-PC era, where devices have task-specific hardware and are overwhelmingly purchased by consumers, subsidized by carriers, and replaced every two years or so. Apple isn’t going to promise backwards compatibility forever, either, and it’s still selling 3GSes.
- Breaking down hardware devices by OS alone makes even less sense than it used to. Even if you’re a developer who wants to reach the widest number of customers possible, you want to know what variety of the OS a device is running, its screen size, its primary uses, and other information about the hardware.
Ultimately, though, I can’t decide if this is a real problem for Google and Android or potentially a huge advantage. In the short term, it’s been an advantage; It’s let the operating system, user base and developer community grow in a hurry. In the long term, though, it doesn’t seem like Google can continue to maintain tight control of the source code during development and promoting its latest and greatest developments, and then let just about anything go once it’s released while letting less-favored products drift away.
Soon, we’ll have to sever those two questions — what’s good for Android, the family of broadly compatible devices, as well their users and developers, is bound to come into conflict with what’s good for Google, the search and software company who continue to develop Android and put it into the world.