Desktop Death Watch: Don't Hold Your Breath!Desktop Death Watch: Don't Hold Your Breath!

Yet another pundit has proclaimed the death of the desktop. I'm not planning to send any flowers just yet.Linux blogger Ken Hess says that he knows when the "year of the Linux desktop" will finally arrive: never. I have some problems with that position, but his underlying reasoning is what really caught my eye: By the time there is widespread acceptance for Linux desktops, it will be too late. Web-based desktops will replace the old school heavy desktop operating systems of today.

Netbooks will be the desktop of tomorrow with no local operating system. Oh there might be a JEOS (Just Enough Operating System)--a minimal system whose only job it is to get you to a network (LAN or Internet) but not a traditional OS with applications, office suites, games or the like. For service subscribers and everyday users, your operating system will consist of a web-based desktop lookalike that is in reality a fancy web page. Renegades, like you and me, will have the choice of web-based desktops or virtual machines that run what looks like a traditional desktop stored on some remote server."

Hess advises us all to "forget the Desktop Wars, the Browser Wars, The Office Suite Wars and any other petty religious nonsensical wars relating to software. Your service provider will determine which desktop, office suite and browser you use when you connect to and use their services."

Oh, and expect to see ads on your new cloud-based desktop -- naturally "subtly placed to entice but without too much 'in your face' interference."

Right. Let me know how that plan works out.

Cloud-based applications and storage infrastructure are already commonplace today. And clearly, we are moving into an era when Web-based desktops will replace traditional desktop environments in some situations.

That doesn't doom the traditional desktop, and it certainly doesn't doom desktop Linux.

This isn't about replacing one dominant desktop computing paradigm with another one. It's about dumping the very concept of a "dominant" desktop paradigm in favor of an IT environment where multiple desktop OS models fill multiple niches.

Mobile workers will move either to fully Web-based desktops or to hybrid models where a lightweight desktop augments a core set of cloud-based applications. Some enterprise employees will rely on Web-based desktops, others will take advantage of a server-based virtual desktop infrastructure, and still others will stick with a more familiar desktop OS.

Creative workers and power users will continue to work mostly with traditional desktop systems combined with both cloud-based and locally-installed application software. IT staff will bare their teeth and rise to defend their custom-built, tricked-out desktop workstations against the JEOS infidel hordes.

And anyone with secrets to keep -- think executives, R&D staff, and stuffed-shirt marketing types -- will be ice skating in hell before they entrust their desktop data to the cloud.

So, what do most of these desktop paradigms have in common? They will work extremely well with a lightweight, flexible, highly robust desktop operating system.

That operating system will make the most of existing hardware resources, allowing businesses to extend their desktop replacement cycles. It will run cloud-based apps, desktop software, and virtualized guest operating systems, all on an as-needed basis. And it will minimize the security risks associated with any networked IT environment.

And when a remote server -- or the network -- flakes out, that "JEOS" had better offer enough functionality to keep us reasonably happy.

Windows 7 and Mac OS X both have important roles to play in this new hybrid desktop environment. But when I think of a desktop operating system that can meet all of the requirements I listed above, I mostly think about desktop Linux.

The desktop is not dead. But the notion that companies can rely upon a single, dominant model for building a desktop infrastructure certainly has seen better days.