Last week I got my first close-up look at Live Mesh, Microsoft's new cloud computing-based data synchronization and device management platform. Unlike past sync solutions from Microsoft, Live Mesh maintains a master copy of your data on the software giant's own servers, enabling instant access to the latest version your files from any Internet-connected device. It's clearly a shot over Google's bow.
What's interesting, however, is that while Microsoft seems to be following Google's lead in advocating cloud computing, its actual implementation couldn't be more different, both technically and philosophically. Customers will have to decide which approach to the cloud works best for them -- and, equally important, so will independent developers.
In the traditional computing model, you use an application to create a document (be it a manuscript, a spreadsheet, a database, or what-have-you). Then, when you want to save the document, the application hands it off to the operating system, which maintains a copy of it in local storage as a file.
Google's model represents a radical departure. In it, the cloud is the computer, from alpha to omega. Because there are no disks or volumes for the user to maintain, there is no need for the artificial concept of "files" or a file system to store them in. Persistent storage is reduced to an abstract concept: All that exist are applications and their associated documents.
Google's brand of cloud computing has other advantages, too. Because the applications exist in the cloud, there is never anything to install and no upgrades or security fixes to manage. In fact, the user is freed from all of the day-to-day interactions with the OS that characterize the traditional desktop computing experience. Certainly there is some kind of OS running beneath the servers that power Google's applications, coupled with some form of organized storage; but these are mere technical details, of no concern to the user.
Sense a pattern? No wonder Google makes Microsoft antsy. You don't need an OS to run Google's applications and you don't need to buy them. You don't even need to install them. The Google model invites users to ignore everything that has been Microsoft's bread and butter for the past 25 years.
"OK," says Microsoft, "two can play at this cloud-computing game." But Microsoft isn't about to follow Google's lead in minimizing the importance of desktop software; to do so would be to admit defeat. Instead, Live Mesh aims to deliver the advantages of cloud computing the Microsoft way.
With Live Mesh, your documents persist in the cloud, but they do so in a familiar form: files and folders. What's more, each object in the cloud has an "end point" on one or more devices in the Mesh, where a copy of its contents is mirrored to local storage.
The psychological distinction for the user is subtle but significant. In contrast to Google's ethereal documents, these synchronized files and folders are objects with substance. Although they are instantly available to any device that can access the Mesh, it is the device at the edge of the cloud that takes ultimate charge of them, rather than the server in the center. Furthermore, users create, modify, manage, and organize them using traditional desktop software.
Obviously this model fits much better with Microsoft's view of the software universe. Users gain the persistence and omnipresence of cloud computing, yet they keep buying traditional operating systems and applications.
One could argue that this is also the greatest weakness of the Live Mesh model, since Google offers its applications for free. But Microsoft is rapidly growing accustomed to competing with free software. Consider its recent Microsoft Equipt offering, which bundles Office 2007 with supplemental software on a subscription basis, priced for cost-conscious customers. For Microsoft, overcoming customer thriftiness may simply be a matter of adjusting its own licensing terms.
Google, by comparison, has a tougher row to hoe. Its Web-based applications are impressive achievements, but they lack the polish and functionality of desktop software such as Microsoft Office. Because of this, they are unlikely to win many converts among the "power users" who form Office's staunchest constituency. Plus, enterprise customers may be leery of the idea of storing and managing their documents primarily in the cloud, rather than on local drives.
In short, Google's approach to cloud computing is revolutionary, while Microsoft's is evolutionary. Live Mesh appeals to the status quo, which should make it the easier sell. And yet Google's momentum is undeniable.
For now, customers are free to decide which model better suits their individual workflows and requirements. Independent developers, on the other hand, face a more difficult choice. Do they follow Google's lead down the road of "pure" cloud computing, where both documents and applications exist solely in the cloud? Or do they take a cue from Microsoft and push data into the cloud, while still relying on traditional desktop applications to create and manage the data?
The SDK for Google App Engine is already available, and Microsoft promises to release an SDK for Live Mesh soon. For developers looking to push computing still further into the cloud, the question now is: Which cloud?