Where the Sun Never Shines

Once upon a time not so long ago, IBM was considered the fuddy-duddy old computer company reeking of odorous mainframe contrivances and a business model several decades out of date. Sun Microsystems (News - Alert) in the 1990s, however, was looked upon in awe as the visionary company of the future, with its promotion of Java and powerful Netra servers (such as the Netra i and Netra j models) which ultimately became the SPARCstation and Sun Ultra series systems bundled with server application software, the “must-have” equipment for every web-based dotcom company. I remember those well – toward the end of the 1990s, I was the principal consultant on the world’s first all-Flash website, town24.com (which became defunct when the tech bubble popped). 

Town24.com insisted on flying me to meet with a sort of integrator/consultant group in Santa Clara, California, to okay the selection and purchase of equipment to power the site. The meeting went something like this:

 

Yours Truly: “Do you have authority to buy the biggest available Sun servers?”

 

Them: “Yes.”

 

Yours Truly: “Okay use those.”

 

Them: “How much memory should we use? They can hold four gigs.”

 

Yours Truly: “That’s fine. And install the biggest, fastest disk drives. Max them all out.”

 

After the tech bubble went pop, Sun faced an identity crisis. They began to gravitate more toward software, and then toward open source software. But making a lot of noise about “opening up” the Solaris operating system, and then acquiring and promoting the open source phenomenon MySQL (with its 11+ million installations), which should have been a coup, never really revitalized the company. The problem with open source is, if programmers don’t like you or your company, they can create their own “fork” and work on the software elsewhere (which they did). However, to their credit, Sun does offer a nifty 110 Terabit per second InfiniBand switch.

 

Meanwhile, with the appearance of computer languages such as Python, Objective C and Ruby (on and off Rails), Java is threatening to become the COBOL of the 21st century.

 

And that brings us to the on again/off again soap opera involving IBM’s (News - Alert) negotiations to buy Sun. Both companies have considerable overlap. There’s a slight chance you could combine Solaris and AIX and the respective companies’ server technology, but how about Glassfish and Websphere? Both user groups would end up getting angry, unless one were phased out. Same thing with trying to integrate Sun’s SPARC processor into an IBM world increasingly populated with POWER3 and descendant processors that implement the full 64-bit PowerPC architecture. It just doesn’t seem plausible, unless IBM simply wants to kill it and confiscate their customers.

 

Yes friends, it’s no coincidence that many industry pundits are comparing Sun’s impending doom to that of SGI (Silicon Graphics Inc.), the remnants of which were recently snapped up by Rackable Systems (News - Alert) for $25 million, just after SGI filed for bankruptcy.

 

But before we pack up Sun and ship it off to Palookaville (come to think of it, Gregory “Seth” Gallant should do a Palookaville comic about Sun, as they “generally dwell on lost, lonely characters searching for meaning, often reaching back into the past,” according to Wikipedia) perhaps Sun should try breaking up the company and spinning off its components. I wouldn’t mind having a desktop version of Solaris running on a desktop UltraSparc.

 

But that’s just wishful thinking – not quite as wishful as what’s been going on at Sun these days, but close.