To this end, Microsoft has integrated its hypervisor product, Hyper-V, directly into Windows Server 2012; it has made the bare hypervisor available as a free-to-use product (Hyper-V Server 2012); and it has contributed a Hyper-V driver to the Linux kernel so that Linux could run better under Hyper-V -- tactics aimed not just at old foe Linux but at an increasingly formidable competitor: VMware.
By giving away Hyper-V, offering free tools for migrating from VMware products, and positioning Azure as a cloud environment that's technologically compatible with Microsoft's in-house servers, Microsoft is taking aim at VMware's deeply entrenched position in the data center. Microsoft has even turned the tables on VMware's cost calculators, claiming Hyper-V yields better cost/performance ratios than vSphere under specific workloads -- yet another shot across VMware's bow.
For Microsoft, the battle with VMware is at least familiar, albeit with a services twist. Stepping into the burgeoning IaaS (infrastructure as a service) market, however, pits Microsoft against a new breed of competition from the likes of Amazon and (once again) Google. Here, adaptability will be key.
We are already seeing this in action, as Microsoft's approach to Azure has been to make it more than a Microsoft-only playground. Microsoft has added support into Azure for many non-Microsoft languages and frameworks, including Node.js and Java, and Linux-based virtual machines -- which, again, benefit from having Microsoft-contributed code to allow them to run well under Azure. Microsoft has even added support for what amounts to competition for Microsoft's own server-level products, such as Apache Hadoop for Azure.
Coming from Microsoft, this surprisingly open approach to competition within its own walls may be the key to luring converts, Forrester's Rymer and Hammond contend.
"Microsoft hopes to attract developers who don't identify with .Net today in hopes of converting them to the new Windows Platform -- and possibly even .Net -- in the future," Rymer and Hammond say in their report, emphasizing how the new .Net 4.5 Framework is "a combined client and server-side framework," with asynchronous server APIs that are designed to work hand-in-hand with WinRT's client-side parallelism features. It's another sign of how the current RT front end is meant to complement the current and future .Net back end.
Telerik's Sells also notes, "Windows Server and Azure have become wonderful back ends for Web and mobile apps, depending whether you want to use an on-premises or off-premises cloud. As Win8 and Windows Phone 8 grow, they'll attract developers who will make use of Windows Server and/or Azure for their back ends."
Conventional wisdom casts Microsoft in the role of lumbering dinosaur, slouching off into the dusk, coasting on what little momentum it's retained from Windows and Office. The reality is that Microsoft is constantly moving in a number of areas at once: on the desktop, in the server closet, in the services market, in entertainment and mobile tech. They may not always come out ahead (Windows Me, the Kin phone, Windows Mobile 6), but they never stop evolving.
What may be new for Microsoft is the need to better cohere its strategy around an ever-widening array of services and technologies, especially as the breadth of competition it faces widens.
While Microsoft has thus far lagged Apple and Google on mobile, Microsoft would be wise not to let the mobile cart lead the company horse. Yes, mobile tech is king right now, especially as Android and iOS devices eat into PC sales. But those devices are only what they are because of the services they talk to. As the mobile field flattens out and becomes further commodified, Microsoft will be able to better see opportunities to distinguish itself on all sides of the equation: client experiences, service offerings, and infrastructure. Rushing this by forcing the desktop to be more like a mobile device would be a mistake.
Going forward, Microsoft will also have to find a better revenue model for the Surface. Microsoft's other big hardware product, the Xbox, remains a loss leader, but sales of software, auxiliary hardware, and services for the Xbox more than make up for this. The Surface needs a similar ecosystem. Otherwise, Microsoft needs to cut it loose.
RT presents an intriguing opportunity for Microsoft, but if Microsoft is to take RT seriously as a future platform for the company, it must ensure everyone else does, too. This starts with smoothing the path for software makers to port and deploy RT editions of their apps. Microsoft Office on RT isn't bad, but it's missing just enough features (say, macros) to be problematic. We should be able to get 32-bit, 64-bit, and RT-based editions of all our favorite software without feeling like we're missing out. After all, the legacy Win32 world isn't going to die that easily.
Ironic as it may seem, building an open source framework people want to use should also be a key initiative for Microsoft going forward. Microsoft has been accommodating open source products better in their proprietary offerings -- say, Linux in Hyper-V. But it needs to contribute a framework of its own -- a PHP without the cruft or a Ruby without the overhead. TypeScript is a step in the right direction, but there's no question Microsoft can do more if it wants to.