Sun’s culture placed too much emphasis on engineering stature. Every time I’d talk with a Sun engineer off the record, they would tell me stories about how good ideas were often eviscerated by weighty luminaries with strongly held opinions. This problem is at the core of Sun’s final years. It was a company guided not by solid economic reasoning, but by gut feeling and opinion. From JavaFx to Netbeans, you could tell there was no solid strategy in place for a platform which had every reason to become the dominant player of the last decade.
Now that Oracle’s acquisition of Sun is complete, we’re seeing a wave of high profile resignations. Now we see James Gosling trying to figure out “what next?”. Gosling hints at this in his blog post, it appears that Oracle just didn’t pony up enough money to convince him to stick around, here’s a quote from Gosling’s blog:
Oracle’s acquisition methodology is making job offers that involve a (often significant) salary cut.
This makes me feel a bit better about Oracle’s strategy for Java (no offense to Gosling). Toward the end of Sun, I wasn’t convinced that the company was making rational business decisions. (While others would radibly disagree), they bungled the open sourcing of Java and alienated anyone paying attention to the JCP. Each JavaOne would bring yet another announcement that didn’t really connect with the audience. (How many developers were chomping at the bit to author Interactive BluRay? What about Project Looking Glass? Remember the UN/”phones for developing nations” announcement?) There was so much missed opportunity in the Java market, you can really start to resonate with one of the commenters on Gosling’s blog:
While we should be talking about the dominance of Java in the mobile market, we’re running scared from Apple’s iPhone platform. While we should be having discussions about the boundless possibilities for Java-based web applications, we’re talking about PHP powering the most popular and dynamic web sites?
Sun’s failure was as much about overhead as it was about bad technology decisions. Schwartz could never tame operating expenses at Sun. The fact that Oracle isn’t afraid to reduce overhead convinced me that they are serious about investing in Java. I’d feel less confident in Oracle’s Java strategy they had left the same cast of characters in place. There was something very dysfunctional about Java leadership at Sun, it will be interesting to see how this changes at Oracle.