An interesting question, no?
Attempting to predict the future of technology can be tricky business for sure. Still, I believe there are some strong signs to support where I believe we are headed, and it appears that at least one major technology vendor agrees with me. Before I get too far ahead of myself, it might make sense to take a quick look back at where we have been and where we are today.
Setting earlier data input techniques such as punch cards and teletype terminals aside, the earliest of the true desktops were your typical CRT based computer terminals. These devices were sometimes referred to as green screens or dumb terminals. In this model, the applications themselves ran on centralized mainframes, and the terminals were used simply as input/output devices.
As the client/server revolution of the 1980s began to take hold, the Personal Computer became the working desktop, with Microsoft Windows and the various Apple Macintosh systems being far and away the two most common. Client/Server computing, still the most pervasive computing paradigm deployed today, involves applications running on the local PC while interacting with network based file, database and business intelligence servers. While pushing the processing out to the PC certainly advanced the industry, it also created an unforeseen problem: Desktop Management.
A couple of approaches have begun to emerge to deal with the so-called "Desktop Management Problem," and The Cloud is at the center of both of them. One approach is to convert client/server applications to web based applications and deliver them via web browser. This approach largely marginalizes the desktop operating system, meaning it no longer matters what version of Windows or Mac OS a given user is running. As long as an application is written to industry standards, and an end user has a web browser and internet access, the application is available to them for use. Commercial software providers saw the opportunity afforded by this deployment model, which later became known as Software as a Service, or SaaS.
While the SaaS model continues to gain momentum, the vast majortity of applications still require a traditional desktop operating system to run on. To cope with this requirement, Server Based Computing (think Citrix or Terminal Server) and Virtual Desktops are leveraged. Both of these solutions involve running legacy applications within a centralized data center, whether private or public, and making them available from most any device that is connected to the Internet.
This is where we are today. We now have SaaS, Server Based Computing, Virtual Desktops, and traditional/legacy desktops dotting the IT landscape. I believe the future, however, will be very different and we have a concept known as "CoIT" to blame for it. CoIT is an acronym for the Consumerization of IT, which is an ideology which suggests that corporate/enterprise IT is now, at least partially, being driven by consumer technologies - or more specially, the users that own them.
With iPhones, Droids, iPads, Xooms and other consumer devices in mind, imagine for a moment that your next desktop is less of a desktop and more of a virtual workspace. This workspace would allow legacy applications, even those requiring completely different desktop operating systems, and SaaS applications from multiple providers to be accessed from a single portal. It would allow those applications to actually interact with one another in ways similar to applications that run on a single desktop. To appease security concerns, the security for this portal would be maintained within corporate firewalls. To address the management aspect, an application catalog similar to the "App Stores" from companies like Apple and Google would be available, only this catalog will contain strictly those applications approved for use by corporate management. In summary, we are talking about a workspace that is less focused on the desktop or network operating systems and, some would say finally, more focused on the applications themselves.
Does the above sound too good to be true? If it does, I would advise you to take a look at VMware's new Horizon project. While the first iteration does not deliver on all of the above promises, it delivers on many of them, and it is clear that the above is where the project is headed. If they are successful, it will be a safe bet that there will be others that follow in their footsteps. So whether the leader in this space turns out to be VMware's Horizon or some other technology, I believe pretty strongly that the desktop of the future will be less of a desktop and more of a virtual workspace.
It is an exciting time to be a part of technology, and Cloud Computing in general!