What's the future of Windows in the enterprise? Middleware.

I believe that VDI is simply a "form factor" change for desktops that doesn't fundamentally transform the enterprise's end user computing environment.

A big part of my desktop virtualization road show talks is "The Future of Windows in the Enterprise." As you're well aware, while I love VDI technology, I believe that VDI is simply a "form factor" change for desktops that doesn't fundamentally transform the enterprise's end user computing environment. I also don't believe that the future desktop is as simple as just converting all of our existing desktops to VDI.

So if the future of Windows in the enterprise is not VDI, then what is it?

First, remember that I make a distinction between "The Desktop" (Big D) and a "desktop" (little d). The Big D Desktop is the complete monolithic "brick" of an experience that we've been delivering to users over the past twenty years. It includes the Windows OS, applications, settings, data, shortcuts, configuration options, wallpapers, kittens, the registry, DLLs, and every other aspect of a user's end user computing experience.

But as I wrote in my 2012 article "The 8 components of the future desktop," we can break that old monolithic desktop brick into a bunch of little pieces. Some of those pieces are our responsibilities as IT professionals, while others each user will "bring" with him- or herself on whatever device the user happens to be using at that moment.

Breaking apart the desktop and delivering applications, data, configurations, etc., separately means that we can easily integrate applications based on different runtimes into a suite of services for our users. We're already seeing this today with VMware's Horizon Suite and Citrix's Cloud Gateway. While both companies used to only deliver Microsoft Windows desktop applications, today they can integrate traditional desktop apps with web apps, HTML5, local iOS and Android apps, etc.

Once we're in that world—where all our apps are not just windows apps then there's less of a need to run and manage Windows on the endpoint (as I wrote a few months ago in "Who wants to manage all the "gunk" of Windows just to deliver one app? When will you ditch Windows on the endpoint?" In that world, Windows will be nothing more than a piece of middleware. It will be the platform we use to deliver old school traditional Windows desktop applications.

By the way, I touched on this briefly back in 2010 in my article, "Is Microsoft Windows in danger of becoming the 'XP Mode' of the future app world?." When it comes to traditional Windows desktop applications, the answer is "Yes." But that's okay.

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

The problem is timing and a key assumption.

WRT to timing, it's clear that the big D desktop is not going anyway time soon. The reason for that is simple. Enterprise apps, of which the majority whether they are written in .Net/C#/C++/Java are designed to run on a keyboard/mouse format which is Windows in the enterprise. They are not changing anytime soon. In other words hard to justify cost of development for the majority of apps. So this is why I am confident that VDI/RDS will continue to become a more important part of the so called middle layer to deliver the Big D desktop.

Now onto the small d desktop. I think there is a dangerous assumption here not to think that modern Windows apps won't be written for this world. Windows 8 blue is still a baby and you simple can't write off Microsoft tablets if Windows blue apps start to also make sense on Tablets. There are many more Windows enterprise developers than iOS/Android combined.

So the small d desktop of the future should not assume Windows won't play a big role. Thinking of Windows as Middleware IMO is therefore only warranted in the context of the Big D desktop.


I think that in the fullness of time, you're probably right, but that we're looking at a lot more time than you may think.


1) Applications were written to work within a unified Desktop environment. Applications share data, functions and resources. When you split all of the various pieces and apps into disparate parts of the data center or cloud, it breaks a lot of that functionality. Things like copy and paste become hurdles.

2) Users have been trained to work in that environment, and the vast majority of computer users are no where near as savvy as many industry folks think they are. We had to completely scale back our XenApp deployment simply because our users could not wrap their heads around where their data was or how to accomplish very mundane tasks (real users print ALL THE TIME for example).

Yes, as more services get pushed into apps and cloud based services, the need for Windows diminishes, but but for those of us in corporate environments, that reality is a very long way off.


Windows == Backward Compatibility Layer

EVERYTHING new and innovative is done Mobile First. Even MSFT is telling developers they have to re-write their application to a new stack going forward.

Of course windows will not go away anytime soon but going forward even desktop operating systems will look more and  more like a mobile one from an application lifecycle and management perspective.

- Vittorio


At this point in time, enterprises have a heavy dependency upon the richness and complexity of the traditional windows applications.  Assuming that newer app models and mobility could meet those needs today, it would still take about 8 to 10 years to flush those "legacy" apps out.

And while there are some great capabilities  in the newer apps, there is not a complete portfolio of line of business apps that most of those enterprises need available today.  Mobile apps, while great when you are not chained to a desk, do not allow for the productivity needed for those that are, nor do they allow the richness of interacting with other apps.

So the flush clock hasn't started yet

PS:  It's all about the data and not the apps anyway. And we have a long way to go before anyone is ready to organize computing around data and not apps.


Brian, the point is that once “we’re in that world” we don’t need any kind of desktop. Thus no one will need desktop virtualization. All we need are apps as frontends for the cloud. Look at Windows Blu and you know in which direction Microsoft is heading. And look at Android and iOS. No virtualization needed. It just doesn’t make sense to waste resources in the datacenter to virtualize endpoint stuff considering a little smartphone is now more powerful than a PC a few years ago. And wait a few more years and you can run a datacenter on a smartphone.

Desktop virtualization will just help a few organizations to master the transition a little faster. I believe we will live in the new world sooner rather than later. Many seasoned IT people underestimate the speed of changes because they think linearly. They look at the changes in IT of the last 10 years and then come to the conclusion that such profound transformations “take 8 to 10 years”. However developments in IT are not linear; they are accelerated. Just think how long it took the desktop to enter the enterprise and compare this to smartphones, and then to tablets.

It is better to accept these changes now or you will share the fate of many mainframe experts who seriously believed that their jobs won’t “go away anytime soon” because of these tiny PCs. Desktop virtualization never played an important role and soon it will completely disappear. Don’t waste your talent for an obsolete technology.