You've spent 23 years at Microsoft, like Bob Muglia, and now top management says you're out. What do you do next? Your problem is the same as the company's. For many years it's clung to the wrong set of assumptions, as the ground shifted around it. The cloud is ascendant, and Microsoft is woefully unprepared to adapt to it.

Bob Muglia played as strong a hand as he could as head of the servers and tools division at Microsoft, building it into a $15 billion-a-year business. The problem isn't that Muglia isn't smart or has lapsed into becoming a representative of the past. On the contrary, he is among the most flexible and modern of Microsoft executives and one of the few with the general skills to potentially replace Steve Ballmer as CEO.

His problem, if he had one, was subscribing to the culture he operated in, one that sees the world as operating system-based and Windows-centric. In fact, technology's arena of play has moved up a level. The cloud delivers services based on software; Microsoft still wants to deliver the software, or have a new service based on only its software in its Azure cloud.

Definitions of the cloud vary widely, but at its core, the cloud is a changed relationship between a computer server on the network and an end user. The end user may be farther away from the server than ever before, but can still dictate programs to it and elicit a sought-after response without an authority figure stepping in to grant permission. The server, in turn, can react and come up with a longer or deeper list of services than before, tailored for the end user. In the cloud, where servers, such as a Google or Amazon Web Services data center, are huge, the client/server relationship paradoxically has become more equal. As the client shrinks into an iPad or smartphone, it's gained more control over the cloud.

Where is Microsoft in this shift? The operating system has little to do with it. Windows remains strong on the desktop, despite early cloud-based efforts to displace it, but Microsoft is struggling to get its share of the new small device market. The Windows Phone 7 is a respectable example of a client device for use with cloud servers. It shines at working with Windows applications -- although it didn't include the copy/paste function for moving things around when it arrived last November. But it needs to gain more adherents to become a broad-based cloud device. Perhaps the main question is whether the Windows Phone Marketplace will ever compete with the Apple App Store.

The overall community of Windows developers remains large and potentially ready to supply new apps to a young market. But Microsoft had to re-jigger its Windows Mobile operating system to get to the kind of operations it wanted on Windows Phone 7. It's a very late entrant in a field crowded with Android, Symbian, PalmOS, and other mobile operating systems. Even the phone's name is conventional and tired. Is the new operating system worth it for developers to learn its details?

Microsoft had previous chances but its early phone entrants never got out of the starting blocks, including the Kin One and Two. The firm fumbled its approach to a market with different set of priorities than the desktop. What would you wager on the prospect that a tablet running Windows will be a big hit soon?