VDI, VDI+, VDI.next... Really, it's all about all desktops. (And so are we!)

In a conversation about what is or isn't "VDI," I wrote about how the term "VDI" was evolving to what I called "VDI+", but that in five years it wouldn't be called "VDI" or "VDI+", it'd just be called "desktops."

Last night I was thinking about something I'd written last week. In a conversation about what is or isn't "VDI," I wrote about how the term "VDI" was evolving to what I called "VDI+", but that in five years it wouldn't be called "VDI" or "VDI+", it'd just be called "desktops," and everyone would know we were using combinations of local/remote/online/offline/streamed/whatever.

In the desktop world of 2013:

  • Some of us may use multi-user remote desktops (formally known as terminal server)
  • Some of us may use remote desktop connected to single-user Windows instances (formally known as VDI)
  • Some of us may stream disk images down that run physically or natively on local client devices
  • Some of us may stream disk images that run in VMMs or bare-metal hypervisors on clients

And this also applies to applications.

  • Some of us may provide remote access to seamless single-windowed apps running on remote multi-user systems (TS RemoteApp, Citrix XenApp, Provision Networks, Ericom)
  • Some of us may provide remote access to seamless single-windowed apps running on single-user systems (Provision Networks, Ericom)
  • Some of us may stream apps to run locally instances of Windows (App-V, XenApp Streaming, ThinApp, Symantec SVS, etc.)
  • Some of us may remotely push software packages to Windows instances to be installed in the traditional way (SMS / System Center Configuration Manager / WSUS)

What do all of these have in common? While they each leverage very different technologies, they all provide access to applications and Windows desktops.

Microsoft is calling this the "Windows Optimized Desktop." They're looking at all their products--Vista, Terminal Server, App-V, MED-V, VECD, System Center, Live--and saying "we can deliver a Windows desktop that's optimized for the hardware, connection, application, and user that needs it. (Of course this is more of a "goal" at this point, as several of these products need to be bolstered by third-party add-ons.) But the messaging is right, and I like the fact that a "normal" desktop engineer (i.e. someone who hasn't been doing terminal server for 15 years) will start to understand "Hey, I can do things differently now!")

So what?

Most people reading this are probably thinking, "Yes, of course this is true. What's the point?"

I'm still thinking about Robert Hammersmith's blog post from last week where he talked about VDI being just a small part of the overall desktop picture (and a small part of what Citrix XenDesktop can offer). Of course I (and most people reading this) agree 100%.

What's different for me, though, is that I think my efforts to change the definition of "VDI" are misguided. I've been saying, "No.. No.. VDI is more than remote desktop connections to single-instance hosts. VDI is about offline, and local, and app streaming, and ...." I even went so far as to invent my own term. (VDI+)

But really I think that's the wrong approach. Instead of trying to redefine "VDI" to include all future types of desktop deployment, let's just called it "desktops."

So yeah, "desktops" in 2013 will be VDI, and offline, and local, and streamed apps, and installed apps, and everything else we need.

BrianMadden.com: Making desktops cool again

To that end, we need to evolve how we define the space that we cover on BrianMadden.com.

We started out covering server-based computing in 2003. We naturally expanded into the SoftGrid / app streaming areas, and then after that we went into VDI. Last summer, Gabe wondered whether we should expand our focus to include general virtualization? We decided not to (even as many other sites did), and to instead keep our focus on Windows application and desktop virtualization. This doesn't mean that we don't ever talk about general virtualization--it's just that when we do, we'll do so in the context of desktops and applications. The same is true for The Cloud. We don't want to cover the cloud per se, but we'll talk about the cloud as it related to delivering Windows desktops and applications.

So BrianMadden.com of 2008 was all about desktop and application virtualization.

But as the definition of "desktop virtualization" expands and evolves into just the way all desktops are deployed, I wonder whether we should think about the broader desktop and application picture?

There are a lot of technologies for desktop management, lockdown, and provisioning that really are important to us. For example, how can we cover Microsoft App-V in the context of delivering apps into VMs, but ignore the fact that many people us SMS / System Center to deliver applications directly to physical desktops? Can we really have complete conversations about using XenApp remote applications versus App-V streamed apps without also considering System Center for deploying MSI's?

I'm not suggesting that we change the site to be a hard-core SMS or desktop management site, but I do think we have to slightly expand our focus to think about ALL ways that desktops and applications are deployed and managed, and that might mean paying a bit more attention to things like System Center and companies like Symantec.

What are your thoughts?

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

Server virtualisation is getting to the point where is just something you can do, moving main stream. Desktop virtualisation is still not done yet and I think the harder task as it involves more components.  However I think it is an important focus as it is the point at which most users interact with systems and has the potential to have the most influence of management costs.

Seeing more articles area about this would be good.


I think your post sums up an important point that the Desktop experience/mgt/support is more important than any one vendor's description of virtualization method (remoteapp,vdi,etc).  However, all of the solutions that are currently maturing which you listed at the top all have one thing in common and that is that they are more progressive and better than the old SMS,CA, way of doing things.  These vendors are know that the old way is dying and are licensing appvirt solutions (landesk w thinapp, ms buys softricity, sym w appstream).  So my question for you is why would you want to waste time expanding the scope to cover those products dying on the vine, and vine is a good word to describe how they've managed to stay in place at some companies.  I can see why you could look at managing these new technologies through an SMS framework or something like that but please save your fastest clock cycles for newer and better technology.


Totally agree.

How about the name "dynamic desktop" which is the super set of all those choices you listed above. Also, dynamic since we're exiting the world of the static, fixed configurations of today's application and desktop environments?


Yes and no....

When you're talking about SMS/SC, Altiris, LANDesk, what are you really talking about?  Repackaging MSI, Transforms files, MSP's and customization wizards?  Why not throw in Citrix Installation Manager which essentially is just the same thing.

So are you going to focus on different management consoles and agents, or the various sub-components that each vendor offers like PC Backup, Change and config management, patch remediation, and asset management? - then please NO!  There are a hundred other sites that focus on just that (just like virtualization sites)

I think that falls more into the category of desktop installation and automation.

I dunno,  brianmadden.com was always a great resource for application delivery and in "delivery" I mean on-demand, instant, and scalable.


Is it going to be about managing "desktops" in the enterprise in 2013 geared more towards managing a user's identity..


Opps.. Typo.. Is it going to be about managing "desktops" in the enterprise in 2013...OR... Geared more towards managing a user's identity..



I am pleased to see you moving in this direction.  I have been of the view for some time now that the real killer app in the network is the user.  What is most important to the user are the applications and services they need to work productively and profitably for the business.  What most often stands in the way of the user is the complexity of the network infrastructure.  

In this vein your recognition of Desktop, which I refer to as Infrastructure, recognizes the need for change we face in the network service offering today.  

There was a moment a decade or so ago when the first Fortune 1000 organization in the world created a "Terminal Services" team within the network management schema.  That moment crystalized the value proposition of application delivery as a "need" within the network.

As a pre-sales consultant for RES Software I speak to large globally based corporations on a recurring basis, organizations that are struggling with the very dilemna you reference in this article.  What is this new world?  How will we manage the complexities of the "Some of us" configurations you mention?  How are we to go about the work of justifying the soft costs associated with the duplicitous service management for the "some of us" who need the service here and in this way, while "some of us" need it here in that way.

In your article it is my feeling that you have found the first half of the answer we will see in the network of the future.  While today we have a Desktop team, a Terminal Service team. Application virtualization team, Platform Virtualization team, all of this complexity must be winnowed down to something far simpler.  

I believe your "Desktop" team is the first half of this simpler world.  It works from a business point of view.  We are able to quantify hard costs on a per device basis in this "Desktop" world, and this is good for business modeling, resulting in a more dynamic work environment for technology.

The second half of the equation is User Workspace Management, a term I notice you (and Gartner) are using a bit more often these days.  Not to turn ourselves into Marchitects (far from it!), but User Workspace offers an opportunity to manage the user experience in such a way that is decoupled from underlying "Desktop" complexities.

User Workspace at its best allows for the management and delivery of application service configuration, resource access, policy and security delivery from a single management console.  UWM decouples the complexities associated with the "Desktop" environment you mention in your article.  UWM focuses on the context of the user (in the building, out of the building, and so on) for the dynamic rendering of user session space (no scripting, significantly reduced reliance on profile retention).  

UWM achieves the essential work of configuring services once, over the entirety of the "Desktop" environment, for the purpose of productive and profitable consumption of network resources by the only killer app that counts, the user.  

The future is a team of "Desktop" and User Workspace.


We are seeing convergence with EDS taking on the intergration of say app virtualisation, and other elements... but I would suggest that you stay away from covering EDS in detail, if you can help it.

There are other forums for that.


I think that we can't help but give at least some consideration to ALL forms of delivery. With so many methods of delivery available (SMS, streaming, SBC, VDI, Cloud), we must remember that the end goal is providing the user with a consistent and expected set of tools.

In order to meet the demands of anytime / anywhere computing, I can see many of these methods converging (along with availability and resource auto-provisioning). However, the reason I come to this site is to gain insight on that original core topic: "How do I get my applications to my people". The delivery is the means to that end.