@NStalnecker
Well, I went and looked at the graph. All numbers are last September to July:
XP went from 71% to 62% a 9% decline. Put another way, for every 8 XP users at the beginning of the period, 7 were still using XP at the end.
Vista went from 18.6% to 14.3%, a 4.3% decline., or for every 4 Vista users, 3 were using it at the end of the period. That would be a steeper decline than XP.
Others went from 4.5% to 5.0% during the period. That's an interesting result. Maybe those are where the mobile oses are hiding? That particular metric has been up and down during the period, but July marked it's highest percentage. November 2009 was the lowest point, at 4.41%
At the beginning of the period, among the people who used modern Windows operating systems, the breakdown was as follows: XP-78%, Vista-20%, Win7-2%. At the end of the period: XP-68%, Vista and Win7 16% each.
Windows 7 official release was at the end of October, so use in the initial period were among RC users.
Apple, which rolled out Snow Leopard at the end of August 2009, right before the period tracked, went from 3.03% to 0.77%, Leopard to SL, to 1.42% to 2.48%. Regarding share of the last two OS X releases among their users, the crossover occurred in April 2010, probably propelled by new Mac sales. At the end of the period for every 3 Leopard users there were 4 Snow Leopard users. I don't have any numbers to compare, but it seems as though there were fewer upgraders than Apple would prefer.
Disclaimer: we don't have any sense of the sample sizes in September, 2009, and July, 2010, but I suspect July would be larger and my description of trends would be accurate only if the sample sizes were the same. For instance, a drop of 71.5 to 61.8 changes if the sample size was 1.1 times larger. And it would be worse if the sample size decreased over the 10 months.
