Why Windows fragmentation will blow Windows to smithereens (cough)

Android fragmentation will wreck Android for smartphones and tablets just as much as Windows fragmentation ruined Windows for the PC.
Written by Steven Vaughan-Nichols, Senior Contributing Editor

The Windows operating system has been plagued with fragmentation ever since the 3.x branch of the OS was released on multiple PCs. (Ahem)

Android fragmentation will blow up Android just as much as Windows fragmentation has blown Windows to pieces.
Image: Google

The transition from XP to Vista, to Windows 7, and, most recently, both iterations of the newest version of Windows, 8 and RT, as well as all patch iterations and dot versions in between, has left a scattered landscape of PCs in various states of OS upgrade version malaise. (Cough)

This has created problems for Windows developers when coding applications, and when they test against different versions of the OS and different target devices. (Oh my!)

The introduction of multiple versions of PCs, as well as Windows virtual machines and emulators running in Mac OS X and Linux, has further complicated this situation by creating additional "forks" of Windows, which have their own unique application issues that developers need to address. (The horror! The horror!)

In case you haven't figured it out by now, I'm not being serious. While I have no fondness for Windows, I know quite well that — even though there are, for example, six major versions of Windows 7 — Windows is now and has been for years the number one desktop operating system.

Why, then, must the fact that there are multiple versions of Android lead to its being blown to smithereens? It seems like hardly a week can go by without someone proclaiming how terrible Android fragmentation is. Somehow, just like Windows back in its early days, Android keeps getting more and more popular.

Why? Because while operating system fragmentation is indeed a pain for developers, it's not nearly as much of an annoyance as some people would have you believe. Just like you can run pretty much any 32-bit Windows application — save Metro and RT-specific apps — on any version of Windows, you can run pretty much an Android app on any Android device.

Mind you, the app won't always work well. Anyone who runs smartphone-specific apps on tablets knows that, but they will run. Smartphone and tablet vendors know that and they like it. So long as they stick with Android as a foundation, they're guaranteed to have tens of thousands of applications from day one with every new device. Even companies that don't use Android, like BlackBerry, know that this is a smart move, which is why BlackBerry 10 comes with support for Android apps with its Dalvik implementation.

I can't see Samsung, HTC, or anyone else that matters in the hardware business trying to fork Android. For customers, there's just too much value in the common operating system foundation and the Google Play apps and store for vendors to ignore.

Software developers see it the same way. Sure, they're willing to support Apple iOS and Android. That's where the customers are. But support Tizen and Firefox OS and Windows Phone 8 and BlackBerry 10 and Ubuntu? That's asking a lot more. Do you really expect them to rush to support Amazon Android, HTC Android, Facebook Android, and Samsung Android? I don't think so!

Look at the numbers today. According to ABI Research, 58 percent of 2013's 56 billion smartphone app downloads will be Android apps. Apple iOS will have 33 percent, and far, far behind, Microsoft's Windows Phone will have not quite 4 percent and BlackBerry will get 3 percent. So, who's going to write apps for less than 3 percent of the market? Nobody, that's who.

Let's be practical. The market can support two, maybe three major smartphone and tablet operating systems; after that, you're looking at marginal players.

Maybe there's room there for small players in some niches. Firefox and Ubuntu, for example, are aiming at the low-end range between feature phones and smartphones. I can see that. What I can't see is the successful Android ecosystem being carved into progressively smaller, more labor intensive, and less profitable portions.

The one company that might fork Android, but I doubt it would bother, is Amazon. It has a unique business model. For it, Android is not so much an operating system as it is a sales channel.

But, again, why would it bother? It has already tuned it to use Amazon Kindle Store instead of Google Play, so it has already accomplished its mission-critical goal. What on Earth could it have to gain by forking the operating system? Nothing, absolutely nothing.

Look at the PC market. For years, Windows has dominated, Mac OS X has had a significant single-digit stake, and Linux hangs on at about 1 percent of the marketplace. There are a lot of reasons that it's worked out this way, but Windows "fragmentation" was never a problem.

For both hardware OEMs and software programmers, the minor nuisance of supporting multiple versions of Windows was never a real issue. Supporting multiple, incompatible operating systems, however, is a problem. Which is why Mac OS X and Linux have remained niche desktop operating systems and why I suspect Windows RT, which is incompatible with other Windows versions, will never be successful.

So, with all that, do you really think Android hardware and software companies will break compatibility when there are many solid business reasons to keep using Android as their base? The Windows ecosystem companies weren't that foolish, and I don't think the Android ecosystem businesses are any dumber than they are.

Related stories

Editorial standards