Microsoft's universal apps + VDI/VMI = useful "Windows" apps on iPhones & iPads.

Yesterday Microsoft had a big Windows 10 event where they announced all sorts of things, including their vision for Windows 10's "universal apps." The concept of universal Windows apps has been around for a long time (I feel like we've been promised it for a decade now), but I guess it's more real this time?

Yesterday Microsoft had a big Windows 10 event where they announced all sorts of things, including their vision for Windows 10’s “universal apps.” The concept of universal Windows apps has been around for a long time (I feel like we’ve been promised it for a decade now), but I guess it’s more real this time? (Dunno. Maybe a real developer can comment on what this means from a dev’s perspective?)

Anyway the idea is that you write a single Windows app that can run on a variety of devices with varying screen sizes and input methods (including even an XBox). Microsoft showed off a preview of a future version of Office where the same apps ran (locally) on a laptop, a tablet, and a phone, where it was literally the same app code running on all three devices.

The UI transformation to take an app that was written for a large screen, keyboard, and mouse and make it run in a useable way on a phone is the million dollar target here. We’ve been saying for years that that’s the main drawback to VDI on a phone. (In other words the challenge isn’t that you can’t get the remoting and networking to work from a phone, rather, it’s that the app you’re accessing from your phone via VDI is a desktop app with a UI built for a different type of device.)

Microsoft kinda-sorta tried to solve this with so-called “Windows 8 apps” (or “Windows Store Apps” or “Metro” apps or “TileWorld” apps or whatever you want to call them), but the big limitation in the past has been that the API restrictions of those apps were very different than traditional Windows desktop apps. This meant that if you wanted to write a Windows 8 app for a tablet that you essentially had to go back to Square One and develop the app from scratch, and for 99.99% of the world, if they were going to go back and develop a touch-based app from scratch for a tablet, then they were going to build an iOS app for an iPad.

This brings me to the next point. If Windows 10 universal apps are more-or-less “normal” Windows apps with hints that allow them to automagically re-configure themselves for small devices, what’s the plan for getting these on iOS and Android devices?

Microsoft’s demo showed their Windows 10 universal apps running locally on phones and tablets, but of course those were Windows 10 phones and Windows 10 tablets. For these to really catch on in the mobile space, there will have to be a solution for non-Windows devices.

I can see several options.

Option 1: if a Windows 10 universal app really is the same app that runs everywhere, I would assume Microsoft could extend RDSH / RemoteFX / VDI so that if a user connects from a mobile device, the copy of the Windows 10 app running on the full Windows 10 OS in the datacenter could reconfigure its UI so that it presents itself in a way that makes sense on the mobile device.

Option 2: Failing that, I wonder if Microsoft (or Citrix or VMware) will get into what’s now being called the “VMI” (Virtual Mobile Infrastructure) space. VMI is like VDI, except instead of users connecting to back end VMs running Windows, they connect to back-end VMs running mobile OSes. (Due to Apple’s licensing restrictions, all of the VMI products on the market today use Android. Of course this doesn’t mean that the clients have to be Android, so an iPhone user can connect to a remote Android session running an Android app, and that’s probably a fine user experience for them—certainly better than a desktop app.)

Anyway I wonder if Microsoft would support running the mobile version of Windows 10 as a VM which could be used (with RemoteFX) to remote mobile Windows 10 apps to mobile users (again, regardless of their platforms). I would imagine this is a big opportunity for VMware and Citrix as they could add Windows 10 mobile app interfaces into their desktop virtualization suites and deliver them with the same protocols, performance, and policies as traditional Windows desktop apps.

The downside to the first two options is that they both require a network connection to use the app. (Which is maybe not a problem for enterprise apps?)

Option 3: This is a crazy long shot, but I wonder if it would be possible to run a Windows VM (or at least some kind of micro-VM) on an non-Windows mobile device. It wouldn’t necessarily have to be the entire Windows environment, rather, it could just be a sort of translation API which would make the Windows 10 universal app run on a mobile device. (Kind of like what Adobe does with Flex.) Of course it would be up to Microsoft to enable this, and I can’t imagine them wanting to enable “Windows” apps to run on non-Windows phones, but meh, if Adobe can do it, why can’t Microsoft?

All that said, we’ll have to see how easy it will be to port traditional Windows desktop applications to Windows 10 universal apps. (Everything on Microsoft’s site talks about migrating Windows 8 apps to Windows 10 apps, rather than traditional desktop apps.) Still, even if you have to build a new app from scratch, it would be nice to only have to build an app once (if that’s really how it works) for multiple form factors, and Microsoft certainly has the mindshare when it comes to enterprise developers.

I have no idea how this will shake out, but I know for sure that (1) Windows 10 will be huge, as will Windows 10 apps, and (2) this is a great opportunity for Citrix and VMware to extend XenDesktop/XenApp and Horizon into the mobile app space (beyond just EMM app and device management). Regardless of what happens, it’s kind of exciting and it will be fun to watch!

What do you think? What am I missing?

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

They already have universal app for Windows and Windows phone:

which is based on WinJS (wrapped withinin browser, like PhoneGap). Microsoft open sourced it and made it work on cross-platform:

I bet this universal app is based on WinJS instead of virtualization.


Assuming that this conversation is all about new apps:

If the goal is to create new cross-platform apps, then we have plenty of ways to do that web, hybrid, and even mostly native-ish apps. (Just look back at my app development platform articles.) So in that case, my reaction is "meh." I mean, who's building new native windows apps? People build web apps, plus iOS and Android apps.

Now if there's a reason why an app should be remote via VDI or VMI, then great. In that case, Windows 10 really is the ultimate platform for VDI/VMI apps. Much better than having VDI on Windows 7 and VMI on Android, actually. So yeah, that's really cool!

So then the question is, do we get to a point where we decide that VDI/VMI is the default way we want to build an app in general, and/or the way we want to build cross platform apps? That's the next conversation we need to have. (Brian and I have talked about this a lot, so I guess it's time to write the article version :)


By any chance Microsoft taking the wraps off of this new type of application virtualization?


It seems to me that the term "Universal App" is mostly the new name for the "title-world apps" rather than a true designation of universality. Not only are these apps limited to Windows, they are, in fact intended for Windows touch. Not even for "regular, old" Windows. At the beginning of his demo Microsoft's Joe Belfiore explicitly stated that they will "continue to evolve their *full* Win32 versions of Office". If the new Universal Office is so universal, why would they need to do that?

Also, note the emphasis on *full* - to me this means that the Universal Office, at least to begin with, will be less than full. This would not surprise me given that the same set of executables need to be small enough to fit in a phone.

RFX does support multi-touch, and the whole point of these apps is that they can adjust to most any screen size, so I definitely see them as good candidates for remoting, from the technical perspective. Whether anyone will need to remote such apps from the business perspective will remain to be seen. If they are, indeed, remoted, given they can run on Desktop Windows, why would you want to use VMI instead of VDI?

In general, I'm with Jack on this: meh. If you want true universality then just go Web.

One thing that does excite me is the possibility of getting XBox games to run on the other Microsoft platforms, including mobile platforms.


This doesn't make any sense to me:

- VDI does not have a future in the mobile era where apps are delivered from appstores, apps run sandboxed and OSs are maintained from the cloud. It simply doesn't solve any problems in this world

- Users are on iOS and Android because that is where the apps are. Application developers target iOS and Android because that's where the users are. Nobody is going to build new Windows apps to target iOS and Android via complex VDI infrastructure

- End-users are looking for native experiences. That's why we have Facebook Apps, Instagram etc. on every major platform

- Don't forget that Windows is the contender in the mobile space. Android and iOS are the majority

We have to face the fact that the destiny of VDI is bound to legacy Windows applications and when they fade away, so will VDI. In a way VDI is the terminal emulation software of this decade.


Great article and discussion here.  It sparked a good discussion within Reddo and we ended up posting a blog on the topic -  

In short, the Universal App strategy is very attractive for enterprise IT organizations looking for practical approaches to mobilizing existing enterprise applications.  There are huge economies to managing a single application infrastructure that has multiple UIs built and optimized for different form factors and operating systems.    HTML5/Hybrid apps are by far the most practical approach for building enterprise apps with positive cross platform user experience that can be managed within a single infrastructure.    

A huge challenge is the large number of existing enterprise applications built on Windows-desktop technologies.   Existing VDI approaches are not solutions for delivering acceptable cross-platform user experiences. This is where Reddo focuses - extending applications hosted on VDI with the flexibility of HTML5-driven cross-platform user experience.


Interesting opportunity.

Wanted to comment on the post above from Redenos. I completely disagree! VDI doesn't go away when applications go away. (And I don't think they are ever going away!)

Yes, end users are looking for native experiences, but more important they want the same experience they are used to on laptops and PCs. Apps simply can't provide that level of experience IMHO.


Karen J. Bannan, commenting on behalf of IDG and VMware.