John C Dvorak wrote today about the great upgrade upheaval, and argued that Windows 8 is doomed to fail because it’s just going to be too hard to upgrade, and nobody likes Windows upgrades anyway.
I agree on the first point but not the second.
The biggest problem with Windows 8 is that Microsoft finally figured out you can’t force a desktop operating system onto a tablet or phone, so they built a tablet operating system and they’re trying to force it onto PCs. That’s not going to work either. One of the secrets of Windows’ success is that it’s somehow managed to maintain a very high degree of backward compatibility–1981-era DOS 1.0 software may not run on Windows today, but most of it ran just fine on Windows XP, released more than 20 years afterward–and Windows 8 is throwing much of that away in an attempt to become Apple.
But it’s not my experience at all that nobody likes Windows upgrades. While a few people cling for dear life to versions of Windows immediately prior to major changes–versions 3.1, 98SE, XP, and 7 come to mind–not all major changes are failures. When Windows is due for major changes, the fanboys come out with a vengeance. When Windows 7 came out three years ago, I couldn’t say I was waiting a year to install it without inciting a mob. Windows 7 was enough of an improvement over XP that people went ga-ga over it. And we were due–XP was seven years old by then, a ridiculous length of time between viable Windows versions.
Windows XP and Windows 2000 were greeted with similar fanfare. You would have thought Microsoft was paying people to like them. Especially XP, which in its early versions wasn’t any better than Windows 2000. Early versions of Windows 2000 mostly lived up to their hype, but mostly because Windows 98 and Windows NT 4 were so bad.
And the fanfare around Windows 95 was positively Apple-like. As an OS/2 adherent at the time, I just yawned. Microsoft gave me a free, completely legal retail copy of it, and I waited a year before I got around to installing it. Windows 95 didn’t do anything that OS/2 hadn’t been doing for years–Windows 2000 was the first version of Windows that was demonstrably equal to or better than the version of OS/2 I started running in 1994–but everyone thought I was nuts for holding out. I didn’t actually run Windows 95 as my everyday operating system until sometime in 1997, and I triple-booted Windows 95, OS/2 4.0, and one version of Linux or another until late in 1998.
That’s not to say the upgrade experience is always particularly smooth. Frequently it’s not, but hardware support tends to be the bigger problem. I’ve always thought that one reason Microsoft was so profitable in the 1990s was because they’d release a version of Windows, people would buy upgrades and it worked well enough that they knew they liked it, but poorly enough that they’d decide to buy a new computer, so Microsoft sold its most popular versions of Windows at least twice to most of its users.
Of course, Microsoft’s inability to deliver a workable successor to Windows XP continues to bite it. Like Dvorak said, there are a lot of Windows XP boxes out there, and those users have had a long time to lose original copies of critical software. In some cases, a product like PCmover will transfer those licenses over to a new machine–assuming something Microsoft changed didn’t break the software.
Had Windows 7 come out in 2005 or 2006, Microsoft would have a lot fewer problems today.
As it stands, Windows 8 looks poised to take a seat next to Vista and Windows ME. There are millions of happy Windows 7 users out there, just like there were millions of happy Windows 98SE users out there in 1999. Neither Windows ME nor Windows 8 does much to make life better for those users–they’re pretty much just change for the sake of change.
That’s why I can say pretty much anything I want about Windows 8 and perhaps two people will speak up.