Throwback Thursday: Windows “Vienna”

Most people don’t know or forgot, but Windows 7 actually went through a couple of codenames—Vienna first, then Windows Seven—before Microsoft settled on its final name. But the most notable thing about the start of this project, perhaps, was that it happened as a new team was taking over a Windows, ushering in a six-year period of Soviet-style secrecy.

Consider the timing and the background. The ill-fated Windows Vista first shipped to business customers in November 2006 and then to consumers in early 2007. Originally codenamed Longhorn and in development for several years, what became Windows Vista was actually hastily cobbled together in about two years after then-Windows head Jim Allchin finally conceded that Longhorn had “crashed into the ground.” That miscue cost Allchin his job—he left Microsoft right after Vista shipped—and triggered major changes within the Windows organization.

Without getting into the politics too much, Microsoft CEO Steve Ballmer handed control of Windows over to Steven Sinofsky, who previously ran the Office team. Sinofsky proceeded to oust most of Allchin’s team—many ended up on Windows Phone and elsewhere in the company—and put his own team in place. And unlike Allchin, who was open and perhaps too honest about his intentions, Sinofsky enacted a veil of secrecy under which the Windows team would only promise what it could actually deliver and would do so on schedule.

(This sounded good on paper, but Sinofsky’s over-reaction was as extreme in one direction as Allchin was in the other, and I’d argue that the more temperate Terry Myerson, who runs Windows today, strikes the right balance here.)

Steven Sinofsky officially became the president of Microsoft’s Windows division in July 2009. But Sinofsky was working on Windows two years before that, and he was responsible for Vienna. Indeed, my first peeks at the successor to Windows Vista happened in early 2007, just as Vista was heading out to disinterested consumers.

Windows Vienna was originally part of a now-dead release cadence—instituted first by the Server team—of major/minor releases, where each major Windows release (Vista) would be followed by a minor release (Vienna). So Windows Vienna was originally seen as a minor update to Windows Vista, one that would correct its performance and compatibility problems. This is of course exactly what Windows 7 was, too, but Sinofsky refused to think in terms of major and minor releases. So Windows 7 was simply described as “a release.”

Microsoft responded to my first Window Vienna report by issuing a statement—now expunged from its corporate web site—stating that it was (publicly) focused on the present, not the future.

“The launch of Windows Vista was an incredibly exciting moment for our customers and partners around the world, and the company is focused on the value Windows Vista will bring to people today,” Kevin Kutz, a Director in the Windows Client group at Microsoft said on February 13, 2007. “We are not giving official guidance to the public yet about the next version of Windows, other than that we’re working on it. When we are ready, we will provide updates.”

My first Vienna report noted that the product was due “in about two years”: Windows 7 shipped in October 2009, about two and a half years later.

And Vienna was “expected to include several features that didn’t make it into Vista, as well as a new shell that’s being developed by the people who created the Ribbon UI for the Microsoft Office 2007 system.” Which of course it did.

Vienna was likewise expected to include “some form of the ‘Hypervisor’ (Windows Virtualization) technologies that will ship shortly after Windows Server ‘Longhorn’,” I wrote. That was delayed to Windows 8, as you may recall.

(You can chuckle at this line, too, given recent events: “Microsoft, meanwhile, has promised that future Windows versions will be delivered more quickly.” That never happened. Each release has taken about three years, and as Sinofsky himself noted later, “three years is about right for each release.”)

Of course, before Windows Vienna/7 could ship, Microsoft needed to fix Windows Vista. And it did, though few credit it for this, with Windows Vista Service Pack 1, which, by the way, was codenamed “Fiji.” Fiji and Vienna together represent the last instances of the Windows team using place names as codenames, a practice that dated back to the 1990s.

One last bit: While screenshot fakes were particularly rampant during the Longhorn days, Vienna wasn’t really around long enough to benefit from a cavalcade of fakes. But I did debunk one particularly bad fake in March 2007. Here it is.

My last Vienna tip came in late March 2007, courtesy of some internal slides. But Microsoft’s veil of secrecy was working: Unless I’m just missing it in my archives, I didn’t get another serious whiff of Windows 7—Microsoft spent much of 2007 fixing Vista, getting Longhorn Server ready, and working on products like Windows Live OneCare—until late in the year, when news of “MinWin” componentization efforts were revealed.

Side note: news of MinWin triggered a mini-battle between me and Microsoft Fellow Mark Russinovich, who was then leading the MinWin efforts. (Mark is the smartest guy I’ve ever met and is the definitive expert on Windows internals. That he left Windows for Azure during the early Windows 8 days says a lot, I think.) But regardless of the details, it’s fair to say that a lot of the “One Windows” stuff Microsoft touts now comes directly from these efforts to “carve off a bottom chunk of Windows” and make the system more componentized.