On 12/3/2011 5:11 PM, David Holland wrote:
On Fri, Dec 02, 2011 at 11:31:48PM -0500, Thor Lancelot Simon wrote: > I can't see why. One in 3,333 times, the test falses positive, > the machine reboots, and comes back up running normally. You mean: one in 3,333 new users who installs NetBSD discovers that it panics while running the installer, doesn't bother to try again, and goes around saying "oh, I tried NetBSD, it doesn't work". We really don't need that.
Yeah, that seems like quite a high false positive rate when you think of it across all (potential) NetBSD users, rather than a specific machine. I admit that I've only been glancing at this thread, so perhaps this has already been explained, but why is the false positive rate so high? What about just running the test twice in a row and not declaring a failure unless it fails both times?