The long battle of "thin versus fat" has commenced. From all appearances, Google is angling to end Microsoft's hegemony by disrupting fat client computing on the desktop. The target: none other than Microsoft Office. The weapon of choice: browser-based, thin client applications.

In today's Internet-centered world, the desktop is already becoming little more than a terminal for people on the go. Why should documents and key applications be locked into just one box? Why not move them to the Internet "cloud" so you can work from anywhere you happen to be, using whatever device you happen to have? And once you've sprung your documents from the PC prison, collaboration is just a click away. That's the implicit pitch behind Google Apps.

Of course, we heard similar prognostications when the Web first appeared. Netscape and its ilk were going to change the desktop-bound world -- but they didn't. However, this time around the driving force is not a bunch of startups with big dreams (even though some of them, such as Zoho, are also in the game) but a well-heeled corporate giant with cash to burn and talent to spare.

So is Office as we have known it for the last 20-odd years doomed? Those who've seen an endless parade of inflated possibilities can be forgiven for eyeing this latest paradigm shift from fat to thin clients as just another technology fad that, when it has passed, will leave Microsoft's historic application pre-eminence unchallenged for many years to come. Depending on the time frame, there are strong arguments for and against that position.

Google's plan starts with a stumble

Google is a tidal wave. The talent. The cash. The swagger. If anyone can make cloud computing a reality, and thus take a serious run at Microsoft, Google can.

At the heart of Google's strategy is "the cloud": It hosts your applications and data, providing the ubiquitous reach of cyberspace. No need for a complex desktop OS and application suite, nor the latest, power-sucking hardware to run it. It's a cleaner, more responsible application-delivery model. You might even call it enlightened.

Unfortunately, the cloud is impractical today. Grid computing, software as a service, and the ability to manage application delivery and access at a global scale are all in their infancies. Like the Web in its early days, the thin client apps we see today are merely crude initial constructs trying to point the way forward.

Even with the cloud's limitations, Google needs to do a better job playing up the inherent strengths of the hosted model -- for example, how users can easily share data without having to deploy any servers or infrastructure. Microsoft is already attacking this advantage via its Office Live Workspaces offering, which merely provides a password-protected SharePoint workspace for sharing documents via the Web. When it launches Workspaces as a beta release on December 10, Microsoft could even usurp its competitor's position as the cloud-computing trendsetter. (Of course, despite the name, Office Live doesn't offer Office functionality, so there are only so many stops Microsoft can pull right now.)

Google may ultimately find itself competing against a more functional hybrid product model that lets customers leverage the advantages of the cloud from the comfort of their familiar Windows-plus-Office environment. (That's where Office Live seems to be heading: an adjunct to the traditional desktop-bound Office.) Ironically, the major interface redo of Windows Vista and Office 2007, plus IT grumbles about Office 2007's instability, give Google an opening to redirect those IT investments away from the now-less-familiar Microsoft offering.

But Google faces a more pressing concern: Despite all the hype surrounding its Gears technology, which lets you save files locally, and Google Apps initiatives, the fact remains that none of this stuff works very well. Yes, there are exceptions -- Gmail being the oft-cited example -- but taken as a whole these offerings pale in comparison to the current state of desktop applications. When something as innocuous as accidentally hitting the F5 key can destroy your magnum opus, you know a technology is simply not ready for prime time. (Yes, this has been fixed, but it shows the immaturity in much of Google Apps.)