Don’t believe any of the studies of the relative merits of Microsoft vs. non-Microsoft products until you study the metrics carefully. It’s strange how often many so-called studies are (deliberately?) biased in favor of Microsoft. For example, consider a recent study of the relative security of Linux by a “government security group”. On the surface, the study seems to show that Linux has more security issues than Windows. But Desktop Pipeline editor David DeJean says
The “government security group” turned out to be US-CERT — United States Computer Emergency Readiness Team — an often-quoted organization that you’d like to feel like you could trust. But when you went to look at the source material on the CERT Web site it was instantly obvious that these guys either (a) were actually on Microsoft’s payroll, or (b) shouldn’t be working with numbers higher than 9 without adult supervision.