Sun took an existing asset, Solaris 10, and made it open source for purely defensive purposes. The resulting OpenSolaris project was a backfire set to absorb developer interest and energy before Linux could race through the customer base. Sun scorched the earth before Linux got there.
Sun was always more susceptible to Linux than its big server competitors because Sun dominates the low end of the Unix market. Instead of watching its leading asset be commoditized, it commoditized it itself. In the process, Sun has gained access to one of its natural markets, the powerful, low-end servers being built with dual-core Intel and AMD processors. With Solaris moving broadly onto these servers, Sun may have an expanding opportunity to sell its portal or identity-management software suites.
There are those who doubt this strategy will work, but for Sun, it was one of the few strokes left to it. As of Jan. 31, Sun has passed the four million mark for registered downloaders of Solaris 10. Two-thirds of them are for Solaris on x86. As IBM, HP, and Dell see the rapid uptake of Solaris on Intel and AMD servers, all three are saying they will ship Solaris on their Intel and/or AMD lines. That's a big plus for Sun. That generates a new market into which it may sell its previously stalled Java software. When your strategy forces your competitors to endorse your position, that's a successful plan. HP, IBM, and Dell do not willingly promote Sun Solaris. The strategy may not by itself pull Sun out of the loss column. Nevertheless, a risky maneuver has been executed by Sun President Jonathan Schwartz, and it will be studied in the future as showing how to make a defensive open-source move.