Should mobile operating systems follow the growth strategy of their desktop and notebook counterparts? There are risks inherent in that path.
I got my start in computing working on operating systems. I was fortunate enough to attend a college that had a mainframe (an IBM 360/67) that had an operating system (CP-67/CMS) that implemented virtual machines. It was relatively easy to develop bootable systems, and consequently it was a great learning environment. I later went on to design and develop a real-time operating system (OS) for a process control application (automating a sewage-treatment plant -- the scariest job I've ever had!) and working on the design of a multi-tasking OS for an early personal computer. Operating systems have the job of making the unreasonable environment of "bare metal" something that can support applications, and they include such common functions as task management, file management, exception handling, user interface, and networking support. Like everything else in computing, OSs only grow over time, and that complexity is especially troubling on handheld devices.
Which brings up an important question: Should your handheld device, whether PDA or simply a phone, even have an OS? Operating systems introduce the possibility of configuration errors, bugs (in both the OS itself and in the applications it runs), and viruses and other malware, along with the not-insignificant associated training and support costs. As both William of Occam and Henry David Thoreau noted long ago, simplicity in all things really should be the goal, and this is especially true while mobile. And, as I've written before, mobile Web services are (to me, anyway) clearly the future and demand primarily a desktop-class browser. The Apple iPhone fits this model well, but even the iPhone has a desktop-class OS (the Mac OS) to support that browser (Safari). This illustrates that mobile operating systems are likely to have a bright future, as a bare-metal browser seems unlikely. Significantly, though, note that Apple and AT&T are restricting access to much of what the OS can do again to (hopefully) hold down those potentially troubling support costs.
But the Mac OS clearly isn't the only game in town. There are other popular mobile operating systems, as follows:
- Windows Mobile: It's a little hard to understand Microsoft's Windows Mobile strategy because the company keeps changing the product's name and nomenclature. But as the premier OS company in the universe, Microsoft clearly has a vested interest in increasing the range of functionality in its handheld OS. The latest version (WM 6) is quite robust and very popular, and I expect Windows Mobile to win the mobile OS wars -- sort of. What I really expect to see (pure speculation on my part here) is Microsoft, in response to the iPhone, in fact basing future WM platforms on a scaled-down edition of the embedded version of Windows XP. This will allow Microsoft to run a big browser (Explorer) on future handhelds. Apple's iPhone is going to set the pace with respect to mobile Web access.
- Symbian: A joint venture of Ericsson, Nokia, Panasonic, Samsung, Siemens and Sony Ericsson, Symbian has the largest installed base in the mobile OS space today. Symbian is an elegant, robust and functional mobile OS that is especially visible on phones from Nokia. Being the cynical analyst that I am, I originally suspected that Symbian was formed simply to squeeze price concessions out of Microsoft. But Symbian has seen continual development and enhancements over the years, and it is in general a pleasure to use.
- Palm OS: Many believed that the Palm OS would be the eventual winner of the mobile-OS wars, because it is simple, elegant and functional. But Palm [ (the "hardware" company) spun off Palm OS to Access Corp. of Japan, where it is now known as Garnet. A huge number of applications exist for Garnet, and it remains a favorite of many because of its ease of use. Palm (again, the hardware company) intends to run Garnet on top of a Linux core -- an interesting direction, to be sure.
It's pretty clear, in fact, that Linux is going to play a big role in the future of mobile devices. It's free, it's fairly robust, it's open-source, and it's running on a lot of mobile platforms today. Check out the OpenMoKo effort for an example of this trend.
The big question is how much access the device manufacturers and the mobile carriers will allow to the features of the mobile OS. Microsoft, Symbian, Garnet and Linux all seem committed to providing a traditional programming environment as supporting local execution of code. And, as Apple has illustrated so well, a big OS is required to support a big browser. But the degree to which applications will be allowed to run on the mobile device will be the key subject of debate going forward, and the result of this debate will determine how mobile operating systems evolve. My money is on Web services dominating and the mobile OS fading into the background -- but we'll see.
About the author: Craig Mathias is a principal with Farpoint Group, an advisory firm based in Ashland, Mass., specializing in wireless networking and mobile computing. The firm works with manufacturers, enterprises, carriers, government, and the financial community on all aspects of wireless and mobile. He can be reached at firstname.lastname@example.org.