Microsoft CEO Steve Ballmer has done his best over the past year-plus to try to dampen expectations around Windows 7. He’s doing it again this week during his pre-launch European tour, telling press, analysts and others there that he doesn’t expect Windows 7 to provide a sudden and miraculous boost to the PC market.

But I’m more intrigued by a related comment Ballmer made, as I’ve thought about this very scenario myself in recent months. Ballmer pointed to Vista as an example that tester feedback may not always be the best measure of the success of a new operating system release. From an October 7 Bloomberg story:

“’The test feedback (on Windows 7) has been good, but the test feedback on Vista was good,’ Ballmer, 53, said in an interview last week. ‘I am optimistic, but the proof will be in the pudding.’”

It feels like a long time ago when testers were assessing the many Longhorn/Vista builds that Microsoft issued both before and after the “reset” in 2004. Before the reset, Microsoft officials heard from testers that there were some deep-seated problems with its next planned version of Windows. As a result, the Windows team went back to the drawing board and rejiggered it. Then there were lots more builds. And finally, in the fall of 2006, Microsoft released Vista to manufacturing.

I’ve been trying to recall if there were any early warning signs about the problems Vista had when it first came out. Were there any major outcries from the hundreds of thousands of public and private testers about Vista/Longhorn being bloated; slow to power on and shut down; and including such an onerous number of security prompts that many users would just shut off UAC?

Full article