Citrix drops the word 'XenApp' from its ICA client. Could this mean published apps from XenDesktop?

It's been about a year since I first wrote about Citrix's "app publishing tax" (and how they've tried to address it). Since then they've done countless promotions, bundling options, deals, and repackaging.

It’s been about a year since I first wrote about Citrix’s “app publishing tax” (and how they’ve tried to address it). Since then they’ve done countless promotions, bundling options, deals, and repackaging. But the fundamentals of the problem remain: Citrix XenApp (their Terminal Server-based solution) allows clients to connect via ICA to both full desktops and seamless published apps. Citrix XenDesktop (their single-instance host-based solution) only allows users to connect to full desktops via ICA. Citrix XenApp allows applications to be streamed for local / offline use, but not full desktops. Citrix XenDesktop allows full desktops to be streamed for local / offline use, but not individual apps. (Although XenDesktop does allow you to build individual apps into your desktop bundles, which can be provided remotely via ICA or streamed for local / offline use.)

So far Citrix’s explanation to all of this is that XenApp is TS-based and geared towards task workers, and XenDesktop is single-user workstation OS-based and targeted towards more intense users. But then if you want to stream apps to desktops to run locally, you need XenApp, regardless of whether you're targeting knowledge workers or task workers.

The result is that there’s this really weird overlap / matrix of features and use cases. If you’re talking about apps, you use XenApp. If you’re talking about desktops, you use XenDesktop. Unless your desktop is for a task worker, then you’re talking about XenApp. It’s just a total tangled disaster!

And the problem is exacerbated by the fact that XenDesktop is about half the price per user as XenApp. So XenApp is the solution for the easier task-based workers, except it costs twice as much as XenDesktop?!? Oh wait.. no, it’s twice as much because it supports apps too. Unless you want to use it with XenDesktop, then it’s free, but only for the users who have XenDesktop licenses. (And only if they bought the expensive ones.)

And if you just want to do desktops, why do you have to pay 2x to use XenApp for hosted desktops than XenDesktop for hosted desktops? It’s like Citrix is saying, “Hey, since you can save so much money on hardware by using TS instead of VDI, you might as well give us the money you saved in the form of a more expensive Citrix product!”

The bottom line is that Citrix needs to sort this all out and make a Windows desktop and application product strategy that doesn’t make people want to poke their eyes out when they hear it explained. (I think we can do a Jeff Foxworthy-style thing here. “If you have to have an entire session at your conference explaining how licensing works... you might have a complexity problem!”)

What the community wants is a much simpler model. We want a product that lets us deliver a desktop—local or remote—and deliver an application—local or remote—without having to resort to all these crazy product decisions. We shouldn’t be forced to use one architecture over another because one method is cheaper to license or because Citrix has to protect a sacred cow.

But maybe a bright future is in sight. Citrix recently renamed the ICA Client software to the “XenApp Plugin for Hosted Apps.” But the new beta version (the 11.1.1. version that’s a plugin for the new Citrix receiver) is called “Citrix online app plug-in.” So it’s the “CITRIX online app plug-in,” without the word “XenApp” in it. Is this because XenDesktop will allow online apps too? Is this because Citrix will finally fix their crazy Windows desktop and app product matrix?

Or am I dreaming?

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

When you consider that the VDI arena has no clear-cut top product like XenApp is to the SBC market, adding XenApp features to XenDesktop is a smart move.  Thing is, it could be considered playing catch-up.  Quest's vWorkspace is already there with a full-featured solution, for instance.

I think it hints that a shift is coming in what Citrix considers the flagship product.  In fact, it wouldn't surprise me if they switched pricing around altogether and started selling XenApp for the lower price and XenDesktop (which would include the XenApp feature) at the higher price.  Although I'm sure they'll come up with a way to just raise all the prices :)


Perhaps...  I was told that the name was drop as result of trying to consumerize an enterprise product.   User's don't know what a XenApp is, but they can distinguish that they can run their applications online or offline.



I personally don't see them getting rid of XenApp anytime soon.  Dropping the price, maybe (and it's needed). I do agree with Brian that there is no point in publishing desktops on XenApp anymore since the management and limitations are so great for the end-user.  Why give them a limited, locked down, no multimedia desktop when you can give them something a bit more 'a;ive'?  

I do agree with Gabe though that XenDesktop is the shiny new toy on the block and Citrix is putting a lot of muscle behind it to get it adopted.  Sine they are neck and neck with vmware (depending on who you ask of course) I can't see them increasing the price of XenDesktop either.

As for the client rename I think they did it just to make things a bit clearer for the end-user as Joe stated.  The client works for many scenarios on both platforms so I actually see this name change as one of the smart ones from Citrix.  


In my view Citrix is just casuing mass confusion my trying too hard to counter VMWare. What they have is great XenApp product that is capable of delivering apps and desktops cheaply for many use cases without all the added cost and unknown complexity of VDI (storage etc). E.G call centers don't need XD or anything like that. TS or similar with application virtualization is good enough and cheap to implement. VDI is a desktop OS, and that may be important to some or for a small subset of apps that were developed in the dark ages. All that said, the really cool think is that Citrix has a set of products that offer customrs many options for massive flexibility. Therefore my view is that it's obvious that XenApp and XenDesktop should become a single product leveraging the same architecture to make it simply using exactly the same ICA stack. A blended license to do this would also make it easy. One delivery center product please Citrix, it's much easier to understand, and splitting hairs between XD and XA and building different architectures is just dump, confuses customers and adds to costs for no good reason. The XD team should wake up and understand that the real world uses XA, not XD, so they need to become XA like, as opposed to c atching VMWare.


appdetective, I respectfully disagree with most of what you said.  While there is confusion right now the way I look at it is Citrix is doing the smart thing by staying competitive. They can't sit around for another 20 years with only XA under their belt. It just won't last.

"The XD team should wake up and understand that the real world uses XA, not XD, so they need to become XA like"

I could not disagree more with that statement and I consider it (no offense) a little naive.  If everyone followed that sentiment then no progress would ever be made in any industry.  

The fact is that virtualizing the desktop is simply the next step in the whole virtualization niche.  Citrix has held dominance (and still does) in remote access to applications but with VDI starting to show real progress people are seeing how nice it would be to give a user a rich desktop experience instead of just a single application.  I am not a VDI zealot by any means.  I currently make a living off adminstering a large XA farm and I absolutely love the technology and ease of management it brings. However, I also see the value in being able to dynamically provision a full fledged desktop to a user and stream apps down to it 'locally'.  Do I think XA IS going away anytime soon?  Not a chance.  Do I think in the next 5 years we will see a shift away from how we use XA to deliver applications today?  Absolutely.

I honestly envision there to be 'light' versions of desktops given to users that are not quite as restrictive as one we may deliver on a terminal server today but don't consume as many resources.  This would increase user density on the datacenter hardware and at the same time give users a single place to access their applications seamlessly.

No matter what the next technology actually is it seems clear that seperating the application from the OS is the right way to go so there will still be a need for some sort of application delivery mechanism... it just won't be xA in it's current form.  But hey, that's my opinion.



I don't think we are really in as much disagreement as you may think. XA can publish desktops and many people do, it's simple, cheap and all the managment concepts of splitting os, apps, user etc still apply. What I am arguing is that most peoples current implementation including yourself use XA. To introduce a totally seperate XD label on a architecture that is different for limited benefit (copying VMWARE) and adds cost (with current architecture) just to deliver a desktop is naive for Citrix to pretend they are adding huge value and naive for the consumer to assume. You can use both together and build a better virtual everything future.

That brings me to my next point, architecture. XA and XD are nothing more than products that deliver apps and desktops. XA can deliver both apps and desktops from a TS OS (best ROI) with ICA for XA. XD may deliver apps (I hope Brian is right in this article) and desktops on a Desktop OS with XD ICA, at a higher ROI without TS but other benefits of session isolation which results in a more personal desktop with todays technology. So essentially they are doing the same thing, and it makes no sense to make them so different from an architecture perspective including having two ICA stacks). I.e One Delivery Center product to deliver modes of virtual desktops or applications from any OS. As you point out the future is a better management stack. So does that really mean that XA and XD will evole with seperate stacks for online and offline as well......? Of course that does not mean XA in it's current form should not also evolve to provide new capabilities, the larger point being that they (XD and XA) need to work in synch from the same architecture and platform (not today's even for XA) to make it simple to use providing customers choice of deployment mode and protecting exisiting investments.

To Gabes comments, Quest gets this, and I am sure Citrix does also but I suspect they fight internally XD vs. XA (marketing people I suspect who know nothing about real world implementation) which I think adds to customer costs and confuses the buyer. I like the fact that Quest gives options and does not lock you into one way of delivering Desktops unlike VMWare, the expensive, complicated, network storage, lack of management tools and immature technology way. Take note XD team we well and show how you are really different besides ICA.... from VMWare and marketing fluff.

I am all for virtual everything, so respectfully I am a huge fan of Virtual Desktops, Apps and beyound, make a living doing both but am frustrated by a seperation that makes no sense and alientates customers that have in good faith invested in Citrix over the years. Citrix is perfectly postioned to bring XA and XD together as a simple to use product that does not alienate exisiting customers and enables flexibile deliver of Desktops and Apps anywhere. That's my view anyway......and hopefully paying Citrix SA continues to make sense vs. making a switch. Not trying to offend any posters, vendors a different story :-)



You are right, we really don't see things too differently so my apologies :)

My point is that if most enterprises want to get to the benefits of VDI  (arguably easier to manage, keeps things in the datacenter) then they are likely going to do that with most of their end-user desktops.  If you are already virtualizing the desktop then where is the value in having a whole TS/XA infrastructure behind the scenes to deliver seamless apps? User density?  Most people are already starting to virtualize TS/XA so your user density per hypervisor host is still going to be limited.  And you are also going to have to have the infrastructure for the desktops anyway.  Having both is just too costly.

I do agree they seemed to rush into the market with XD but as I stated previously they are trying to keep Citrix on the 'tongues' of those in the enterprise (ie stay competitive).  Once VMWare gets View into the organization they then tack on ThinApp and there goes a lot of functionality from XA/App Streaming.

At any rate I am going to continue brushing up on my XD skills and if I had the ability at the moment I would do the same with VMWare View.

Good discussion!



For the Virtual Desktop World, the value of XA is in delivering apps in different modes. So you can present seamless apps if you don't want to present another Desktop Shell to confuse monitor realestate as an example, (users often have more than one machine), you can take apps offline, load balancing is there, hopefully desktop os pusblished apps is coming, etc.  All of this adds to greater density, together XA + XD are better, and complement exisiting software distribution.Think about Dazzle with Merchant server and this whole shift towards User driven SAAS, Green dynamic XA for power mgmt and XA is begining to enable the single image goal, and add capabilities that XD can take on as well. Again I will say I like the fact that Quest is already well down a single management framework path is great. Citrix needs to do the same to avoid duplication and avoid cost by making XA and XD run from the same platform accomplished by using a single architecture. It's obvious 101 stuff to me at least, and ICA makes a huge difference to address many use cases.

Also for the record I am also a VMWare customer, and just don't think the OS vendors can loose the Hypervisor battle. Getting VMWare and MS to work together on real world poblems has at least in my experience been a nightmare. So while ESX is the clear leader today, I just don't see how they can continue and hence they are moving up the management stack, and pushing so much marketing that is all over the place, but it is generating push back from MS, which makes me even more worried about whose Hypervisor I should continue to invest in when the Desktop OS of choice is from MS........  

Finally my Linux friends will also tell me that RedHat with KVM embedded in the Kernel will enchance performance, and in parallel to my OS vendor is Hypervisor king assumption will tell me that RedHat will win that battle for Linux. Perhaps IBM will buy RedHat and then what happens to ESX and Xen. My view, ESX dies, and trys to survive by attacking everybody and marketing fluff (see articles on and moving up the management stack (they are not multi vendor today....), Citrix will never make money out of the Xen Hypervisor, but will use it to open up new options faster rather than waiting for MS and continue to build upon their core strenght which is delivering stuff (desktops and apps) to the Windows platform which helps MS, and becoming broader on where the deliver center goes, i.e MAC and varied devices.

Agree good discussion, glad there is a place where we can come to have the discussion!


Why not just bundle everything and call it.....wait for it.....


I agreee that there is currently too much overlap and confusion. Especially when it comes to licensing. I fear because of this many people will take the cheapest way out rather than analysing the requirements and selecting the best product.



There are a few comments in the above posts that I'd be interested in your feedback.

Reading into some comments it seems the TS model is seen to be only suited to "process workers" (I hate the label personally) due to the TS "limitations " and that a knowledge worker needs more!!

The whole XA and XD conversation.

I'd suggest that we are at a point where we hope to empower everyone with tools that have very few limitations. To some that would mean XD to all but I'm not so sure.

It may be worthwhile re-visiting your "use case" post of a year or so ago and attempt to map the product sets to the users. I am personally of the view that we are starting to hand the "use case" decisions back to the users with a combination of XA, XD and Streamed applications and it may be hard to map it specifically to a "plumbing chart". Maybe handing back those decisions to the users on how they want to consume applications and desktops is where Dazzle can help!