Recent US politics aside, it’s unwise to pin your hopes on a single person. But I’m going to do so, because that person is Dave Cutler. Dave is a lead developer for Windows Azure. That means it won’t suck.
Dave is legend: For starters, he built VMS, then built a compartmentalized OS that could run both VMS and UNIX called Mica. This was years ahead of its time.
I had a great conversation about the birth of operating systems with Peter Christy recently, who explained that early on, Digital was very supportive of VMS, forcing engineers to use it instead of older machines. But then they lost their nerve, and weren’t willing to do the same thing for Mica.
So Dave headed to Microsoft.
I first read about Dave in a book called Show Stopper!. It’s a colorful account, but it details how he built Windows NT, dragging Windows from its ancestry as a DOS overlay (think Windows 3.1 and Windows 95) into the world of true operating systems.
Dave bears a grudge: Many folks have told me (though none will confirm) that Windows NT got its name because it’s “VMS+1” (i.e. “WNT”) the way that HAL was IBM+1 in the movie 2001: A Space Odyssey. (I’m hoping Peter can clear this up for me.)
In addition to NT, Dave is also the reason that Microsoft has a consolidated code tree across consumer and commercial platforms. He’s a hardass, but very, very good. And if he’s behind Azure, it won’t stink at all. If it did, I suspect he’d leave. Microsoft knows he’s good, and has given him great latitude to do the right thing in the past.
Jeff Kaplan of Thinkstrategies recently wrote about the plethora of cloud platforms out there, saying Azure has to catch up. Can Azure win?
The short answer is yes.
First of all, it’s got Dave. I may come off as a bit of a fanboy on this one, but you have to admit: The guy has cred.
Second, Redmond has turned on a dime before (though some would disagree) — just look at its adoption of the Internet. Gates made it clear in a 2005 memo that the shift to cloud computing was as significant as the shift to Internet technology that came 10 years before it. Microsoft made good on that promise. While we geeks love our GMail, it’s only got 20 million users (versus 45 million for Hotmail and Yahoo’s 88 million.) Microsoft could snap up Yahoo and have five times Google’s user base (although admittedly, GMail is growing healthily.) Users mean authentication, which means applications, which means revenues.
Finally, there are the developers. Microsoft has a compelling promise to its millions of trained coders: Write your stuff in a .net-friendly language, and we’ll make sure it works on Azure, with all the function calls you know and rely on. That’s a good argument for convincing coders to build for their cloud. It worked to lure PC developers to the XBox, didn’t it?
In my opinion, Microsoft’s biggest challenge is its entrenched sales organization and product revenues: It’s so addicted to up-front software sales that shifting to an on-demand, pay-as-you-go model will mean cannibalizing existing revenue streams. What better time to make a change of revenue streams than during a recession, when you can miss your targets and lay off staffers, and still have money to bankroll the new business?
In the absence of all else, my money’s on Dave Cutler. Which means that for mid-sized businesses, Microsoft will be the eight hundred pound cloud.