Ok, here’s my layman’s take from being in the general vicinity at the time - let me know if this has any merit. Linux was billed as the Microsoft killer when it began to gain prominence, but it never made real inroads in the desktop market. It also turned out that orgs who had implemented MS Server products were not culturally or technically great candidates to shift their back office to a Unix-based OS. However, companies that had invested in top of the line Sun boxes (at a top of the line price point) were very intrigued at the prospect of running something on wintel architecture that may give them 70% of the benefit at 10% of the cost. Those companies were already invested in a Unix-based environment and when Redhat became popular and offered a corporate wrapper to Linux, it was the beginning of the end for Sun. The big data centers also realized they could just string together a bunch of wintel machines running linux and get the same performance as a huge Sun box at a fraction of the cost. The reason I find this ironic is because Sun was one of the largest proponents of open source software and the arguably most successful open source software ended up killing their market.
As an aside, my wife worked in M&A for Sun during the height of the dot com bubble and I attended some amazing acquisition parties all around the Bay.
