According to Steve Ballmer, our customers have seen a lot more innovation from us than they have seen from that [Linux] community. That’s not completely correct, but then it’s not completely wrong either. Open source is capable of innovative features. And it lags in innovative products.

Innovative products go hand in hand with new markets. And new markets usually take time and money. Time to work with customers and to refine the product. Money for marketing to establish the new market. And both to redirect the product if it started with a dud. Open source products need to start with a near miss – otherwise they never achieve critical mass and become another SourceForge zombie.

But the server market is not innovation driven. IT wants stability, not innovation. And IT is perfectly happy with “a clone of a 20-year-old system” as long as it supports their application requirements (and received 20 years of continued refinement).