Operating System Grist for the Google Rumor Mill

Last week, Net Applications reported Microsoft's operating system share had fallen below the 90 percent mark on the 40,000 or so websites where it gathers its traffic statistics. InternetNews is reporting that Net Applications made another interesting, if puzzling, discovery.

It seems the statistics gathered from Google.com (this only includes Google employees -- not the public using the search engine) were showing that a third of these users were accessing sites with an unknown operating system. It's more interesting when you consider that proxy servers block all identifying information, but the Google.com "unknown" systems only had the operating system information obscured.

Theories abound, of course, as to what Google might have up its sleeve. There's the "Google is bringing Android to the PC" school of thought, and the "networking application infrastructure development" theory.

The InternetNews piece gives some insight into both schools of thought. Google has indicated it would like to expand the Android market, and this would be a way of doing so that would allow Google to compete with existing operating systems on its own terms, building on its own strengths.

The Android idea has gotten the most buzz, of course. I am not sure I think it's the most likely scenario. There have been rumblings of an open source Google desktop for years (2006 brought reports of Google using Ubuntu for "unknown" reasons and a bit of confusion from open source leaders as to why the company would move that way at that time). Perhaps the time is drawing closer, and Android is a major piece of that master plan finally falling into place. It sounds dramatic. If it should happen, though, it still is a long way off.

Clay Ryder of The Sageza Group explains nicely why an network application infrastructure makes more sense -- a "software-as-a-service" platform, rather than an all-out desktop-oriented operating system. He points out that "...an operating system really connotes the stuff that makes the hardware and software talk to each other, and they are not in that business."

He goes on to say that developing an operating system to use as an infrastructure for network applications that could be deployed virtually anywhere is more in keeping with Google's past service offerings.

Android ported to the PC, or even the 2006-era dream of a "Goobuntu" desktop are of course possible, but if not overly costly for Google to undertake, would at least be major time investments. Those sorts of investments might pay off over time, but a "software-as-a-service" product gives Google an advantage in the operating system market, should have a faster return on investment, and complements the free services Google already offers.