Microsoft has done well offering online applications to small and midsize business and enterprise departments -- Exchange, SharePoint, Dynamics GP, or Great Plains. It's moved the capabilities of its Visual Studio tools into its Azure cloud, it's instituted data synching between on-premises SQL Server and Azure SQL, all innovations for the dedicated Windows community.
But it's missed a much larger shift toward services flowing out of cloud. The innovations in large-scale dataset handling, such as Cassandra, CouchDB, and MongoDB, are coming from open source and small companies, not Microsoft. The cloud is extraordinarily hospitable to open source code, since it is often built atop it and needs flexible licensing for an elastic response to demand. Linux was the first operating system made available in Amazon Web Services EC2, not Windows.
Ballmer's legacy of foot dragging, posturing, and tooth gnashing at the sight of open source ought to be reviewed more skeptically than Muglia's accomplishment list. Microsoft belatedly decided it must cooperate with open source developers in what it described as -- this was news to Microsoft -- its customers' heterogeneous environments. Its discovery that it wasn't the only child on the playground came as an apparent shock, and late in life to boot.
Now the challenge is getting customers' environments able to work with different clouds. Microsoft has worked assiduously at getting them to work with its Azure cloud. The multi-cloud landscape will use different hypervisors. Microsoft has acknowledged only a slight possibility that not everyone will be using Hyper-V in the coming years. In short, the old rules that Microsoft sought to establish for uniform Windows environments and customer lock-in don't transfer well into the cloud. They've been displaced by a new set of rules that put user priorities higher on the list.
Muglia could be as aggravating as any other Microsoft executive when it came to asserting the company's priorities were correct and executed only in the customers' best interest. But he could also travel to a hostile camp, as he did to the OSCon open source conference in San Diego, and represent Microsoft's stance, while listening and engaging in a dialogue. Occasionally he could concede some ground. He tended to have a calming effect on warring factions, not an inflaming one.
Microsoft may not need Muglia's specific skill set and track record any longer, but it needs to understand that the cloud shifts the balance of power between provider and end user. It moves customer priorities up the ladder and customer lock-in down. It's a hard lesson for some to absorb. Getting rid of the guy who made the best of a bad hand won't necessarily convince everybody that Microsoft is learning how to do it.