Vista scare stories distort the real picture

Playing the old “I Googled this and got x gazillion pages back” game can be hazardous to the rest of your argument. Today’s case in point comes courtesy of my blogging colleague, friend, and neighbor Marc Orchant, who wrote this chilling headline earlier today:

Vista IE7 issue remains unsolved for many users

The post itself, however, is based on a single anecdote, passed along secondhand. So where's the evidence that this is a widespread and unsolved problem affecting many users? Marc says only, “Problems like this are all too common and are being widely reported and discussed in the tubes.  Do a search and you’ll see what I mean.”

That’s not exactly the strongest argument, and in the post’s accompanying Talkback thread a commenter asks the obvious question:

Why does this article's title say "for many users"? This is one guy's particular experience, and nowhere does he suggest that he has found other people encountering this same problem.

To which Marc replies:

Read my post again. Do a search in whatever engine you prefer for "Vista IE7 hang". It's not me generalizing from one person's experience. There were, at the time I posted this comment, about 362,000 results for that query. Even reducing that number This is not one person's issue as I said in my post. This is a pervasive problem affecting many people.

Oh dear. Whenever I try this argument it rarely ends well. Let’s see how it goes today for Marc. I tried Vista IE7 hangs (without quotes) and got 284,000 results. With Vista IE7 hang I got 368,000. (Does this mean the problem has gotten worse in the last six hours?)

Those are pretty damning Google results, right? Uh, maybe not. If that’s the new metric for product quality, there’s a long list of things you’re going to have to stop using before you get around to Windows Vista. Try a few of these searches for yourself:

Vista Firefox hang / Vista firefox hangs – 828,000 – 842,000 results

Safari hang / Safari hangs – 1,640,000 – 1,730,000 results

Ubuntu hang / Ubuntu hangs – 1,020,000 – 1,040,000 results

Tivo hang / Tivo hangs – 1,150,000 results

In fact, maybe the definitive search is this one:

computers hang – 3,520,000 results

At least in my experience, the smarter a device thinks it is, the more likely it is to hang, with or without the help of a browser. I’ve seen nothing to suggest that there’s a widespread problem with IE7. In fact, if you want a firsthand report to balance out the story Marc passes along, here’s my story.

On December 13, 2006, I upgraded my wife’s Dell Inspiron 6400 notebook (Intel Core 2 Duo, 1GB of RAM, 120GB hard drive) from XP Home to Vista Business. It’s been in continuous service ever since. Here’s what the Reliability Monitor looks like for this machine:

That perfect 10 is an accurate reflection of this portable PC's performance. It’s been rock steady for nearly eight full months, running Outlook 2007 (synced with a Windows Mobile SmartPhone), Word 2007, IE7, and a handful of other programs regularly. I’m the IT guy in this two-person office, and I simply don’t have to worry about this machine. Every other week, Judy asks me to explain some confusing dialog box, but I haven’t had to do anything other than routine maintenance on this system since New Year’s Day. It just works.

Now, I’m not suggesting that these two stories balance each other out or that either one has any significant evidentiary value on the question of Windows Vista’s reliability and performance over a larger sample size. But the two stories do accurately represent two typical buckets of Windows early adopters. Those who tried to retrofit Vista onto older hardware might have mixed results, especially when the target hardware is exotic, like the Toshiba Tablet PC in the example Marc points to. By contrast, if you use well-supported hardware and take the most basic precautions, you can have a nearly perfect experience.