BugToaster does not report operating hours by OS. The number of PCs reported running Windows 2000 SP2 increased from 1,451 at noon on October 7 to 1,556 at noon on October 15. During those 192 hours, the number of reported crashes increased from 4,915 to 6,200. We don't know when the 105 individual PCs were added to the database. If all had been added at the beginning, the numbers would reflect 1,285 crashes in 298,752 operating hours, or one every 232.5 hours.
They used the data above to calculate the time that the OS were presumably running in order then calculate the mean time to failure. That is not a very scientific method. They kill the argument for using BugToaster data with this sentence "Bugtoaster does not report operating hours by OS". So, 'extrapolating' this data isn't going to persuade me.
Another thing to consider, which I don't know if BugToaster factors in (after a quick skim), is the people using BugToaster are most likely the people who experience more crashes than the norm (average) which, would skew the numbers.
Slashdot Mirror
There is even an American Life episode about the informant that helped bust ADM. http://www.thislife.org/pages/descriptions/00/168. html
BugToaster does not report operating hours by OS. The number of PCs reported running Windows 2000 SP2 increased from 1,451 at noon on October 7 to 1,556 at noon on October 15. During those 192 hours, the number of reported crashes increased from 4,915 to 6,200. We don't know when the 105 individual PCs were added to the database. If all had been added at the beginning, the numbers would reflect 1,285 crashes in 298,752 operating hours, or one every 232.5 hours.
They used the data above to calculate the time that the OS were presumably running in order then calculate the mean time to failure. That is not a very scientific method. They kill the argument for using BugToaster data with this sentence "Bugtoaster does not report operating hours by OS". So, 'extrapolating' this data isn't going to persuade me.
Another thing to consider, which I don't know if BugToaster factors in (after a quick skim), is the people using BugToaster are most likely the people who experience more crashes than the norm (average) which, would skew the numbers.