A lot of folks in thetech media, including yours truly, have speculated about the fate of Sun Microsystems. As someone who's dealt with Sun on many levels over the past 12 years, I think it's safe to say that Sun is not going to just roll over and go under the horizon anytime soon–it's just not their style, and they have too much cash in the bank to go out with a whimper.

The problems Sun faces are nothing new–they're just magnified by the current market climate. Somehow, the company managed to generate enough forward momentum during the dot-com boom to grow its business dramatically. But the company has never done an effective job of selling software. Now, with so much of its future riding on software,that's become a real big problem.

I drank the Java Kool-Aid a few years back when I took over a Java developers' magazine. J2EE did it for me; it was my belief in 1999 that the desktop was becoming more and more irrelavent from a business software point of view, and that Java offered the best path to the Web for business developers. But I never really completely bought Java as a desktop technology outside of a limited cross-platform development domain, and I didn't hold out much hope for Java on mobile devices and cell phones either.

Scott MacNealy firmed up my thinking for me when he dissed software as a business. “You don't buy software to control your right turn signal,” he said at a Gartner conference. Sun software execs winced.

MacNealy has never “gotten” software. He's a hardware guy. Java wasn't a product–it was, in the automobile industry's language (which MacNealy is so familiar with) “content”–another feature that would help Sun sell servers.

For all its work on Java, Sun never successfully produced a good set of developer tools for Java on its own–because, quite frankly, its development tools have always sucked (as anyone who has used Sun Workshop dev tools can testify to). This is why most Java development in the first few years of Java's history was done in vi (or, for the wannabes doing Java on Windows in the early days, Notepad). Sun acquired developer tools from Netscape, Forte (and Forte's tool is pretty good), but then it didn't know how to market them well.

To put it bluntly, as one Sun employee once said to me, “This company couldn't market its way out of a wet paper bag.” While it captured developer mindshare with Java, it failed to turn that mindshare into a lasting commercial relationship; instead, it has to rely on Java licensees to generate most of its Java bucks.

Then came XML, which, despite Sun's claims of intellectual authorship of XML, it failed to capitalize on XML or web services quickly. Theopen source community (and IBM) did much of the grunt work of making XML and web services connect to Java successfully. Software politics, intellectual property concerns, and plain old intrangisence managed to keep Sun away early and lock Sun out of the