The Desktop Isn’t Dead, Yet

I just wrapped up a review of OpenSolaris 2008.11, which, among other things, represents the most recent fruit of Sun Microsystems’ intermittent and arguably quixotic efforts to field a viable desktop alternative to Microsoft’s Windows.

Over the past several years, I’ve reviewed quite a few of these desktop challengers–perhaps too many, considering the slender combined market share of non-Windows desktop options.

What keeps me forging on in this coverage is my belief that despite the heavy buzz that surrounds cloud computing and all things Web, the desktop (in all its incarnations from the netbook to the workstation) is far from dead.

Microsoft has it exactly right when officials there talk about software plus services as the way forward for computing. For all the amazing attributes of the Web as an application platform, it’s no substitute for the PC’s local storage, memory and computing power potential.

No, the desktop hasn’t reached its end of life, but the desktop does appear to have shifted into maintenance mode. These days, the center of application innovation has moved to the Web, where SAAS (software as a service) is all the rage, and Facebook is achieving eye-popping penetration rates among mainstream users.

Cruise over to’s software department and scroll through the top 150 or best-selling titles. Putting games aside, the list is dominated by familiar old faces: office suites, tax software, image editing, anti-virus.

You’ll also find a handful of virtualization applications, an exception that proves the innovation void rule, since the primary purpose of products such as VMware Fusion is to enable users to run, albeit clumsily, those same old Windows applications on OSes too small to get a native port.

Along the same exception-proving lines is the undisputed innovation king of the desktop, the Web browser, the makers of which are chugging steadily along toward turning the browser plus Web into a sort of desktop OS of its own.

Now, it’s not that all the desktop application ideas have been used up, or that the browser is somehow superior as a desktop application platform.

What’s missing from the desktop world, but alive and well on the Web, is the sort of fierce competition that arises from an open platform that is governed by standards, but accessible to a diversity of hardware and software components at every layer of the stack. As a result, Web application builders can mix and match the components of their choice to build their works on the back end, and tap a client-side market that includes every sort of device that can support a Web browser.

In contrast, the desktop is, for the most part, a closed platform, where a single company calls the shots for the majority of the platform stack on about 90 percent of desktops.

I submit that while the Web is indeed an attractive application platform, the level of Web and cloud activity that we now see is artificially inflated, as efforts that would otherwise be trained on the desktop are channeled instead into the Web.

Diversity is the antidote for the innovation funk in which the desktop finds itself, and the desktop stakeholder with the greatest ability to encourage that diversity (as well as the party with the most to gain from a desktop revival) is Microsoft.

I’ve written about an open-source Windows as a means for expanding and strengthening the platform. I believe Microsoft could help achieve some of the same ends by focusing higher up the stack, on the company’s managed code technology.

Microsoft could work with Novell and the open-source Mono project to make Linux and related Mono-supported operating systems into first-class hosts for applications developed with the framework.

The effort would require enough licensing changes and patent assurances to make .NET/Mono palatable for inclusion in any open-source OS distribution, and to allay developer suspicions of a possible intellectual property trap. However, the move would be a boon to application builders and users alike.

Most importantly, the move could give the desktop, as a platform, the shot in the arm it needs to retake its place on the leading edge of computing innovation.