[Dimensio]

I remember that study. When you analyzed the reports you saw that most of the Linux exploits were local ones (ie, someone had to be physically present at the machine or at the very least have some level of user access beyond just being able to see the machine on a network) while the majority of the Windows ones were remote (ie, any schmuck with the right tool could break in). Also, I think that they counted a number of individual security alerts from various sources (Red Hat, SuSE, Mandrake, etc) as individual exploits even though the different companies were sending an alert for the same problem, so one exploit in Linux would get counted three times.

Is that the study you're referencing, or am I remembering a different one?

[Gorons]

Its the same study, its all about perception and what distros ship with installed versus uninstalled, etc...

The point is all computers have security issues, no vendor is immune. ;)