Windows Vista's 30-day report card
It's been a month since Microsoft released Windows Vista to manufacturing. That same week, I upgraded three heavily used machines in this household to the final build. So, how have they done? Each upgrade has a different story to tell, as it turns out, and I'll share the details in three separate follow-up posts. But here are some overall comments.
First, the third-party development community hasn't just been standing still. In the past month, I've downloaded several new drivers and installed at least one new application (Nero 7) that was specifically designed to work on Windows Vista. Earlier versions produced compatibility errors and lacked some basic functionality.
The most important driver update for me was from Realtek, who appear to have finally fixed the problems with their AC97 audio drivers. Up until yesterday, one machine in my lab played audio that was literally unlistenable: scratchy, static-ridden, and distorted. The final update, posted to Realtek's website on November 30, appears to work beautifully. (Realtek has an update for its High Definition Audio drivers as well.)
Most of the bugs I've run into are obscure, and I have yet to see or hear of anything that has data-destroying potential. More common are design decisions that don't become obvious until you really try to use the feature instead of just testing it.
Of course, during the last month I've been absorbed in Windows Vista as Carl Siechert and Craig Stinson and I delivered the final manuscript for Windows Vista Inside Out. Last week, to test some upgrade scenarios, I installed a copy of Windows XP Professional on a relatively modern test computer and was struck by how slow the process was. It took over an hour and a half, and there were a couple of points along the way where setup stopped and awaited user input from me. By contrast, installing Windows Vista on the exact same hardware took roughly 20 minutes, with all required user input at the very beginning and the very end. Those fast, mostly unattended setups made it very easy to install and reinstall Windows for specific test scenarios. We also benefited from being able to run Vista in virtual machines, where we could try out test scenarios without worrying about downtime if an experiment went wrong.
One observation about Windows Vista worth noting is how quickly it came together over the past few months. In mid-summer, I was skeptical about the chances that Vista would be ready for its scheduled launch, and I wasn't alone. But somehow all the pieces actually fit. In fact, Microsoft's own Reliability and Performance Monitor is one of the tools that proved to my satisfaction that Vista wasn't rushed out the door after all. I'll post the results from this tool for three different machines in the follow-ups to this post.