I am assuming you get that from this:
Shipments do not equal Deployments:
Example: I purchased One large server w/out an OS and put ESX on it. That one server now runs several Dozen Linux server. Depending on what version of Linux I would run none of those would be counted as 'shipments'. So one Server I purchased which runs upwards of 30 Linux servers would count for *zero*. We have already seen this week that CentOS is being used by ISP's, those were boxes purchased without Operating systems (or with windows) and wiped to have Cent on them again not counting towards 'shipments'..
Even if this article was not two years old: Long story short if Windows ships with 70% of sold server operating systems they are likely far beneath that in actual deployments..
So keep post two sentence sound bites from articles that are nearly two years old, without any actual understanding of enterprise architecture (which you display almost daily) your statements are even more meaningless than your sources..
Trying to pad your stats with virtual servers is lame, that's still just one actual Linux server from the hardware standpoint. Windows being preloaded on ~70% of server shipments is consistent to this day as well, if you think you have stats that prove otherwise then let's see them. Give us a link this time, instead of just more hot air.
