One interesting thing is that the OS statistics we have available are based on user agent strings from access to participant sites, and this method won’t gather data from offline machines. Statcounter currently shows 3,1% of win7 and 0,52% of winxp, but I wonder if there would be a significant difference if offline machines were counted.
One interesting thing is that the OS statistics we have available are based on user agent strings from access to participant sites, and this method won’t gather data from offline machines. Statcounter currently shows 3,1% of win7 and 0,52% of winxp, but I wonder if there would be a significant difference if offline machines were counted.