Um, what? In a weird report, Gregg Keizer at IDG News says that upgrading in-place from Windows Vista to Windows 7 "could take over 20 hours to complete, according to Microsoft." This sounded suspicious to me, so I took a few seconds to see if this could be accurate.
Turns out it's another example of sensationalist journalism, and you just gotta love that he chose the "20 hour" bit for his headline. (Actual headline: "Microsoft: Windows 7 upgrades to take 20 hours." Really nice.)
The Microsoft blog post Keizer got this info from concludes with the following:
From the testing we have done, the results show that Windows 7 upgrade time is faster or equal within a 5% threshold to the Vista SP1 upgrade time.
Sounds OK, right? And the point here, of course, is that the Vista-to-7 upgrade experience is excellent. In fact, it's possible to upgrade to Windows 7 in as little as one hour and 24 minutes, again, according to Microsoft. (I'm curious why the headline isn't, "Microsoft: Windows 7 upgrades to take as little as 90 minutes." Wait, no I'm not.)
So where does the 20 hour thing come from? Well, in its bid to be as inclusionary as possible, Microsoft tested a wide range of upgrade scenarios. And on a so-called 32-bit "medium system" (really a "super user" somehow constrained to "low-end hardware"), yeah, it can apparently take all day. (Go figure, but the same PC/person configuration in 64-bits takes half the time, at 10 hours.)
UPDATE: Microsoft updated its post to include this info about the "20 hour" scenario:
The "super user profile" is not a normal user; rather, it's the user profile that represents the extreme power-user who's working with an enormous data set (650Gb of user data) and a large number of installed applications. This user profile is not representative of what most "regular" users, who typically have a much smaller data set and would therefore experience a much, much shorter upgrade time.
What bugs me about this is that it wouldn't be hard to, a) get an actual average of the upgrade time based on Microsoft's figures (hint: Nowhere near 20 hours) and, b) get an average of the upgrade time for scenarios that are in fact realistic for actual customers. In my experience, the Vista-to-7 upgrade takes a couple of hours on actual PCs that have really been used by real people. But I find the Microsoft data here to be overly conservative (i.e. "taking more time than in the real world") for that and another reason: I've performed clean installs of Windows 7 in 15 minutes, 20 minutes tops, and that's true even in virtual environments. But a clean install could take 27 to 47 minutes (according to Keizer, anyway). Seems to me that Microsoft is just covering its bases so it won't hear any complaining from those who try this stuff on borderline PCs.
And its seems like Keizer and/or IDG News are going for the cheap-shot headline. Let's leave that stuff to the blogosphere, eh?