This is the usage share on January 2014
While Windows 8/8.1 increased its market share by 0,12% (10,56% ---> 10,68%), Windows XP increased its market share by 0,23%, almost double than the latest OS
Not really-really, whatever the source of that data is, there is no way on earth *anything* can measure market share with an approximation smaller than 1 or 2 % (at the most).
However, BOTH Vista and Windows 7 seemingly reduced their share , "share" is a moving target (the real numbers should take into account the new machines delivered to customers and the old machines the customers got rid off in the month), but in the hypothesis that the total numebr of PC's remained the same (new deliveries=dumped hardware), i.e. a "perfect" replacement market., those data:
Windows 7 47.31% - 47.46% = -0.15%
Windows Vista 3.10% - 3.30% = -0.20%
Windows XP 29.53% - 29.30% = +0.23%
Windows 8 6.38% - 6.62% = - 0.24 %
Windows 8.1 4.30% - 3.94% = + 0,36
can also be read (since what is "mainstream" is 8.1 now) how:
0.24% of share upgraded from 8 to 8.1
actual new deliveries of 8.1 are 0.12% of share <- which makes a nice trend of an increase in share of around 1.5% per year
0.15% of share (WIndows 7 users) upgraded to XP
0.08% of share (Windows Vista users) upgraded to XP
0.12% of share (still Windows Vista users) suddenly threw their machines in the litter from sheer desperation
Of course those numbers make no sense whatever, the only way to read them is as "nothing changed in market share between January and February 2014".