Is Microsoft Windows in danger of becoming the "XP Mode" of the future app world?

As many of you know, Microsoft made a lot of changes in Windows Vista (and by extension Windows 7) that can break older applications.

As many of you know, Microsoft made a lot of changes in Windows Vista (and by extension Windows 7) that can break older applications. Since breaking applications is not a good thing for a new OS, Microsoft is addressing legacy Windows application compatibility via something called "Windows XP Mode" for Windows 7. XP Mode isn't really a "mode" at all—it's actually a fully functioning Windows XP virtual machine that runs in the background where legacy Windows apps can run. Microsoft has gone to great lengths to hide this from the average user, with the experience being that applications running in the XP VM appear to seamlessly integrate with native applications running on Windows 7. While XP Mode isn't necessarily the most elegant solution to the legacy application compatibility problem, it does solve the problem and it does provide a solution where any application can work in a Windows 7 environment.

What's interesting is that while Microsoft is struggling for a solution to make old Windows apps work on new Windows, the whole IT world in general is having conversations about the future of Windows apps. A lot of people are wondering whether Windows apps will even exist in the future, and people are talking about HTML5 and rich internet apps and SaaS and corporate app stores and all sorts of ways to deliver non-Windows applications.

This got me thinking: Is Microsoft in danger of Windows becoming the "XP Mode" of the future cloud-based app world?

Everyone can agree that Windows apps will never 100% disappear. (After all, even today we have 16-bit apps that people still use. Heck, we have mainframe apps that people still use!) But in a world where apps are delivered on demand and via new architectures, what is Windows' role? If Microsoft isn't careful, the whole concept of "Windows" will be relegated to a middleware compatibility layer for legacy applications.

We already talk about how we use 32-bit Terminal Servers based on Windows Server 2003 to deliver 16-bit apps to "new" environments. We all know how we can use XenApp or Dazzle to deliver Windows apps to iPads, Androids, and all the other new platforms out there.

There's a battle going on now for the hearts and minds of the developers. Apple wants things done their way. VMware wants apps to be developed for the cloud via SpringSource. Lots of folks want apps developed for HTML5 delivery. And Microsoft is struggling to convince Windows developers to use their same dev tools to port their apps to Silverlight and Azure. None of these application platforms require Windows. Sure, the notion that Microsoft has to change their business model is not new. But in this new app world, where exactly does Windows fit in? Is there any viable future where Windows is not relegated to the datacenter, demoted to an eternal role of delivering old apps? (Even Microsoft's arcane device-based licensing schemes can only keep Windows around for so long, like labor unions resisting technologies that eliminate jobs.)

How soon will this happen? Three years? Five? Then what?

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

Brian, very good question and nobody knows the answer as there are so many factors beyond XP Mode that may influence the future of Windows. But why don't you just predict that Windows will be a legacy technology in three years from now? If you do so, according to the Madden Paradox, we can all sit back, relax and continue our work around Windows in the same way as we did since the last century. LOL


Symantec SWV and VMware Thinapp are alternatives which cost a lot less to run legacy apps.


In theory, it does appear that Windows is quickly becoming legacy, however in practice, I doubt that will be the case anytime soon. Windows is so entrenched in business that it will takes years/decades for a major shift to happen.  Look at the browser "wars".  Even though Firefox and many other browsers are now out and have been since 2000, IE still commands roughly half of the web traffic market and changing between firefox and IE is not nearly as difficult as going off of an operating system.  

It is true that many organizations are moving into the SaaS field for some of their apps (SalesForce), but we are talking about thousands of applications. And the discussion that HTML5 will reshape applications is again wishful thinking.  It is 2000 happening all over again with people talking about applications moving to the web.  


Anyone who says that all apps will run off the web are dellusional.  HTML5 is not the answer to everything in the world of tech.  While it does some awesome stuff it is nowhere near as robust as apps that run on Windows.  Maybe in 10-15 years when you get enough people to agree on standards then robust 'cloud'-based apps may have their place but it will never match the functionality a full native OS brings.

For some people these web apps may just be enough to get by with. Google docs, facebook, watching video, tv, email.  If this is all you do then hey a thin OS that runs web apps will work perfectly fine for you.  The rest of us won't be satisified with the sub-par performance and limitations.

Part of the problem with Windows is the backwards compatibility it must retain.  If they could just draw a line in the sand and say, "Any 16bit or 32bit apps won't work on Windows 8/9" then they may just be able to take leaps forward.  No way MS will ever do it because people would go crazy.  Look at Apple, when they went to a new version of their OS didn't it simply break a ton of stuff and they said "Oh well, upgrade your code to make it work"?  

In the end I am not worried about Windows apps going away for a looong time. And not in the 'legacy mode' long time perspective.  


@Watson: with SWV and Thinapp applications are still running on the host OS. So Apps not working on Win7 will still not work with ThinApp.

If they do, you can probably make the app to work by just using a Shim and not virtualise it at all (unless it is a dll error).

I think Microsoft first should do the same as Apple (as they are allready doing that). Just quit the legacy support. First step is virtualisation of the old OS to run them (like apple did in the first release of OS X) and than drop support in the next version.

As for web apps. There will still be Windows, running in the background on the servers.

But I think organisations will not quickly run all apps in the cloud. Some applications still need performance and cannot be quickly transformed to web services.


App compatibility is truly a very big problem for organizations migrating to new OSs like Windows 7. The interesting thing is that this is especially true for companies that are using a lot of "web based" applications, as these tend to be not web-based at all and, in fact, rely on a ton of plugins that are physically installed on the OS, thus creating dependencies on both the OS and the browser.

XP Mode could be a good solution but it is lacking any real enterprise appeal -- such as the ability to centrally deploy / update / manage the apps, not to mention the limited integration the apps will have with the host OS. MED-V adds some central management, but it also tries to shoe-horn a whole VM onto your OS for the purpose of making a few apps work.

Also, the vast majority of app virtualization products cannot really solve the app-to-OS compatibility problem because they don't really isolate the app from the underlying OS.

InstallFree works differently from other app virtualization products by including a full compatibility layer for each virtual application. This way we can solve app-to-OS compatibility without requiring a full VM and while providing full management and OS integration.

So until everyone is ready to dump their fat apps and resolve compatibility problems at the source, it's good to have some options... :-)


Hey guys, good comments all + thanks to Benny for delivering todays paradoxal ROFL :)

Windows will be around for years to come due to far too many factors to mention. Beyond that? Who knows...

As already mentoned, XP mode (or something alike) have no place in corp. world. Personally I would go futher, suggesting that even alleviating solutions (aka. band-aid) have a fairly limited traction. My *wild* guess is that corp. continue to use legacy OS & apps as long required by many rationale. Well, that's how things usually go. No?



Legacy Apps yes OS no.  


I think that 'compatibility' is the least significant question for the broader future of applications. (Bear with me, I know it sounds bold.)

A few things are clear now:

- applications become more and more 'consumerized' at a rapid pace that has not been seen before.

- this consumerization is driven by two new platforms: pocket sized mobile computers + web browsers

- applications are written by application developers using highly sophisticated tools (IDE) that have evolved continuously over the years.

- applications are purchased by consumers and businesses alike to solve their unique problems

The 'application war' is not only fought on the platform layer (Windows/NET/Azure vs. iOS vs. Android vs. vmgooforce) but more so on the IDE layer and the purchase/distribution layer.

What I mean is, that the company that can offer the best value in all of these three layers will be the most successful company.

Back to my original statement. Yes compatibility does matter, but it's only a necessary hardship that Microsoft obviously is much more struck with than the other three guys (Apple, Google, VMware+salesforce) because Microsoft has been around the longest and has created way more legacy platforms.

BUT there is ZERO innovation in legacy platforms. That why they become legacy platforms! :)

Innovation happens in:

- new platforms (this is where the compatibility is a problem)

- IDE layer (look at Visual Studio 2010 and its automated testing and debugging enhancements!)

- the distribution layer (this is where we see disruption with the sprawl of application stores)

Alright, to sum it up:

Compatibility is no big deal for consumers. Consumers can take advantage of new platforms and new distribution models very quickly.

Compatibility _is_ a big deal for businesses because they need to protect their investments in existing applications. This is where desktop virtualization and application virtualization will play an even bigger role in the future than it does now.

Microsoft is uniquely placed in the application space for businesses, because they can offer all the three components. Platform = Windows (incl. Azure), IDE= Visual Studio, Distribution Model = SCCM + BPOS

The weakest link for Microsoft is the distribution model, because their SaaS offering is still playing catch-up with Salesforce.

But Microsoft has the killer IDE! One ring to rule them all.

I don't see Windows going away anytime soon!

Will Windows change in the way it provides platform services to applications and ultimately how we interact with applications? Absolutely! Windows being overtaken by Apple or VMGooForce? Hardly.


In my opinion XPMode is because MS doesn't want to pay penalties to the EU if they go back on their argument that IE6 is part of the OS.


I posted on my LinkedIn profile some time ago about how I thought that simply hosting a Windows OS in a remote datacenter does not qualify as 'cloud computing'. You're simply finding another location to host a monolithic, flaky and old-fashioned OS. Whether thats on your servers, virtual or physical, or in a datacenter at the other side of the world makes no difference.

Why the hell do we need Windows? Its legacy. Do you need Windows to run Facebook? No. Do you need it for anything thats truly cloud-based? No.

Until we unshackle our apps from a clunky, monolithic OS, we will not have true cloud computing. Its like needing your own PABX just to have a single telephone line at your house.

And then there's the blatant stupidity of the same OS being both a desktop and server OS (Windows 7 / 2008 R2 Service Pack 1 being the same file? Oh please).