In this editorial I’m going to suggest the possibility that Citrix’s Linux strategy and Microsoft’s Bear Paw strategy are related. Specifically, I wonder if this is a “You scratch my back and I’ll scratch yours” scenario. I wonder if Microsoft’s decision to delay Bear Paw until the next platform release of Windows (probably 2007) is a “gift” to Citrix in return for Citrix choosing not to release a version of MetaFrame Presentation Server for Linux.
Let’s dig into this a bit.
We all know that Microsoft and Citrix are good friends. In fact, they’re even having a breakout session at iForum to talk about how great their relationship is. We also know that friends don’t hurt friends, and we know that Bear Paw would hurt Citrix. Therefore, the logical conclusion would be that Microsoft chose not to hurt their friend by delaying Bear Paw.
It’s been widely publicly reported that Microsoft demonstrated Bear Paw to the entire server group at their MVP conference last April. Then, in August, Microsoft made the surprising announcement that due to “resource constraints,” Bear Paw would not be part of R2, but that it would be part of the next platform release of Windows (which is assumed to be Longhorn Server in 2007).
So why does it take Microsoft three whole years to move from a working demo to a final product? If Bear Paw is only going to be things like seamless windows and application publishing, how hard could that be? And if it’s going to be more than that, then why wouldn’t Microsoft release an interim add-on to Terminal Server and RDP for things like seamless windows? I bet that I personally could write seamless windows from scratch between now and 2007, and I’m just one person.
To me, it seems that there is an ulterior motive behind Microsoft’s decision to delay Bear Paw. I wonder if Microsoft is protecting Citrix. If they are, why are they doing this? Some people suggest that because Microsoft gets revenue with every Citrix license sold (in the form of a Microsoft TS CAL), why would Microsoft hurt Citrix? I argue that Microsoft could get more revenue if they made Terminal Server compelling enough on its own to not cause people to have to buy a $200+ Citrix license for each user.
In response to this, other people point out that Microsoft likes Citrix because Citrix has a sales force that’s dedicated to selling their products, so it’s like a “free” team to sell Terminal Server licenses. While this is certainly true, I think that most of the Citrix selling comes from partners. (After all, this is what Mark Templeton says again and again in his speeches to partners.) Therefore, these partners would be just as happy selling consulting services and CALs around Terminal Server as opposed to Citrix, so I don’t think Microsoft would “lose” anything in this case.
So, why did Microsoft give Citrix the gift of delaying Bear Paw? I think it’s for one reason: Linux.
I think that Citrix could very easily port their MetaFrame Presentation Server product to Linux. In fact, several Citrix executives and engineers have said that they’ve done it in the lab and that they could easily make it into a “real” product, but that they’ve talked to their customers and that there is not a significant demand for it.
To me, this is absolutely ridiculous! MetaFrame Presentation Server for Linux is one of the most popular things that people ask me about. I honestly wonder where Citrix gets this idea that customers are not interested. Citrix executives say things like “Linux is not ready for the desktop.”
I believe that MetaFrame for Linux would have huge advantages. Consider this: In today’s world, many IT managers talk about wanting to explore Linux for desktop applications. However, the commitments to entry are too high. What IT manager is going to rip out Windows for Linux desktops? (Like that old saying goes, “No one ever got fired for buying IBM—I think the same thing applies today for Microsoft.) The reason that no one uses Linux on the desktop is because it’s too much trouble and too much of a commitment to something unknown.
Now, imagine for a moment that Citrix had a version of MetaFrame Presentation Server for Linux. Then, a customer could build one or two Linux servers to serve Linux applications via MetaFrame and ICA. They could integrate Linux applications into their existing Web Interfaces, Program Neighborhoods, Published Applications, and Secure Gateways. They could test the waters with one or two Linux desktop applications without “throwing out” all their current Microsoft investments.
From a Citrix standpoint, that would make their product more compelling. I don’t think Citrix should try to compete with Tarantella here. Instead, Citrix can push MetaFrame for Linux to their existing customers. Most enterprise Citrix environments contain multiple farms from multiple locations—why not bring Linux applications into that mix?
Why not indeed. I believe this would be devastating to Microsoft. The last thing that Microsoft wants is an “easy” way for customers to integrate a few Linux applications with their everyday applications. After a while customers might chose more and more Linux applications, and MetaFrame would be the management and access infrastructure issues as easy to manage as Windows.
To me, this sets up a very plausible scenario where Microsoft agrees to delay Bear Paw as long as Citrix agrees to not release MetaFrame Presentation Server for Linux.