Sun has been in trouble for a number of years. While they make good product and have good services, the market and value of their core product has declined significantly in relation to Intel servers. All of the Unix market has declined, but other players that had other businesses to prop the Unix segment up (notable IBM and HP) were able to cope better. Sun’s strategy for a long time was to avoid caving in to commodity hardware and operating systems. Java, N1 and thin client (to name a few) were strategies to drive the decision away from the platform where they could compete better. As were strategic relationships with the likes of Oracle. And Sun arguably did a better job than any other company in capatalising on the Internet boom, which cave them a significant war chest of money and customers.
Sun has been able to garner fantastic loyalty in its customer base, particularly amongst the technical bods. Complexity of operation gives techo’s a huge investment in their knowledge and all important bragging rights with their peers. Extremely high levels of service ensure the complexity does not hurt customers and ties Sun into the daily operations (which is future sales goodness). And the mythos of the company made it geek cool, especially with the skill and antics of its demagogue leader, Scott McNealy, particularly the denigration of anything Windows.
With Sun announcing it is going to ship Windows servers all of these benefits are now gone. The geek skill prestige of operating these servers is gone, the ability to over-service declines under cost pressure from HP, Acer and Dell, and Scott McNealy resigned a year ago. Consider the cost of maintaining a processor that is now slower than a mid-range Intel; and also the cost of a proprietary Unix against Open Source alternatives. These costs make Linux on Intel a good move, and this has been successful for Sun. Windows probably seems like an economic good move, there is easy opportunity in their customer base. It is very bad for the Brand though.