Brian Madden made an interesting post today, predicting the end for Citrix XenServer as we know it.
I'm not so sure about the disappearance of Xen from the landscape altogether, but on the disappearance of Citrix XenServer I agree 100%.
I think there's one more aspect worth mentioning that he doesn't touch on... the fact that 99.999% of Citrix admins are Windows admins. Windows admins are about as likely to deploy Linux based solutions as Linux admins are to deploy Windows solutions. Given that the vast majority if x86 virtualisation going on right now involves Windows guests, and there is so much feature parity between XenServer and Hyper-V that it ain't funny, I know which one I'd go with if ESX wasn't an option.
As Brian points out in his post, the value for Citrix in any market has little to do with what's underneath - it's about what they can offer on top. In the virtualisation game, the money for Citrix is in the XenDesktop broker (although currently it's a bit more like a broken than a broker IMHO) - not a Linux based hypervisor.
Citrix knows where their bread is buttered - their position in the market today is based on embracing and extending Microsoft technologies (XenApp for Unix anyone? No, didn't think so). I'm sure it won't take long for Citrix execs to see a gap in the deployment of the XenDesktop broker on ESX / Hyper-V versus the deployment of it on XenServer and wonder why the hell they are even bothering.
[UPDATE] Brian has posted a clarification - turns out I've misinterpreted his original post. Which means I'm going it alone with the prediction of Citrix XenServer's demise (_not_ the demise of the Xen hypervisor), and can claim full credit when it happens :-D