Cue the obligatory wordplay about sunsets, falling stars, or fading sunlight. Sun Microsystems, the former darling of Silicon Valley, is dwindling fast. And I must say, it’s about time.

What value does Sun provide, really? I mean apart from the not-insignificant employment of tens of thousands of clever people. I haven’t been able to make a decent business case for Sun’s existence for years. The company has gradually edged into irrelevance and insignificance, succumbing to the very forces that led to its creation. There’s some poetic justice in that, I suppose. Almost like one of nature’s epic cycles.

First, the headlines. Sun announced recently that it’s whacking and whopping 6,000 jobs. That’s in addition to the 7,000 job losses and eight restructurings it’s already done in the past few years. Clearly this is a company in contraction and it has nothing to do with recent short-term changes in the economy. No, sun’s been flaming out for quite awhile. While other computer makers were hiring Sun was already on its way down. Sales of its server computers are down 27% from just a few months ago. That’s a precipitous drop in a short period. Can you imagine your own income dropping by one-quarter in a quarter? (Sadly, I suspect some of you can do more than just imagine it.)

Brighter minds than mine have wondered about Sun’s place in the world. The company’s management, its board of directors, and its (remaining) employees have all expended a lot of ergs cogitating on Sun’s future. I hope they find a solution but frankly, I don’t see it. Sun’s products just aren’t all that interesting, nor are they sufficiently differentiated from their competitors’ offerings. That’s a recipe for failure no matter who you are – or were.

In The Beginning, Sun created powerful and interesting workstations. Based almost entirely on open industry standards and commercial components, early Sun workstations used Motorola’s 68000-family processors, the Unix operating system, Ethernet networking, and standard TCP/IP stacks. Everything about Sun reeked of commonality, interoperability, and community. It was a welcome (even heart-warming) alternative to the closed, proprietary boxes from IBM, Mentor Graphics, Apollo, and others. It showed that the good guys could win without resorting to old-school tricks that locked customers into a single vendor. And it was good.

But then Sun got uppity. The 68K processors gave way to Sun’s own SPARC chip designs. At first, SPARC was new and interesting. At the time, RISC processors in general and SPARC in particular, looked like they might take over the world and topple Intel’s dominance of the desktop microprocessor market. But in pursuing SPARC, Sun took its eye off the ball. Except for a very short period in the 1990s, SPARC was never a performance leader. It was slow even by RISC standards. Year by year, SPARC’s performance slipped further and further behind that of rival chips, even – most embarrassingly – Intel’s Pentium and Itanium processors. Today SPARC is among the slowest (yet most expensive) microprocessor chips you can find.

On the software side, Unix gave way to Sun’s own Solaris operating system. Once again, the company gave up on industry standards in favor of an in-house alternative. The GUI also morphed into a Sun-only product. The network protocol stacks became Sun property and the silicon was brought in-house.

Microprocessor design, operating system development, and ASIC design are all notoriously expensive undertakings. Yet these high-dollar developments added nothing to Sun’s value proposition. What good did any of this do Sun’s customers, who were paying $10,000 for an “engineering workstation” when a good Windows PC cost less than half that and provided a faster processor and more memory? Benchmark after benchmark would embarrass Sun’s hardware while “mere PCs” from Dell and Gateway pulled ahead of the supposedly professional-grade systems.

The doomed game plan was foretold in Clayton Christensen’s book “The Innovator’s Solution,” which describes how cheaper, inferior products displace better, more expensive ones. In this case, relatively underpowered PCs improved faster than expensive workstations did, to the point where PCs were as good as – and ultimately better than – the supposedly superior workstations. It was a moment of cognitive dissonance: a cheap-looking Acer PC was outperforming an impressive-looking Sun workstation. Sun’s salespeople whistled past the graveyard while their customers wondered aloud what exactly separated a workstation from a merely overpriced computer.

Java was Sun’s late-game Hail Mary attempt to reinsert itself into relevance. “If you don’t like our hardware,” the company seemed to say,” let us distract you with this shiny software bauble.” Java was another Sun-owned property masquerading as a broad industry standard. Standard, that is, as long as you did things Sun’s way and made sure the company collected its payment.

Longtime readers know I’m no fan of Java. It’s miserable technology cloaked in brilliant marketing. The language and its runtime environment is slow, it’s interpreted, it’s slow, it’s resource-hungry, and it’s slow. Which of these features were programmers clamoring for? Experienced programmers know it’s also not very portable, thus neatly defeating its entire reason for being. Java was simply a distraction from Sun’s larger issue of dwindling hardware sales due to an ever-worsening price/performance ratio. The emperor was slowly losing his clothes.

There was a time when Sun was “the dot in dot com.” When companies were buying Web servers like popcorn its products were the de facto standard. But now that drunken-sailor spending is out of fashion and IT managers have a bit more time to evaluate server products before buying a container-load of them, Sun’s star doesn’t shine so brightly. Server racks increasingly sport Dell logos instead of Sun’s.

Still, old habits die hard. It’s tough to admit you’ve been overpaying for computers or that that image-enhancing workstation on your desk is really just an overpriced and underpowered PC with a funny operating system and increasingly marginal support. Engineers, programmers, and IT managers with a sentimental attachment to heavy metal will continue to buy Sun products for as long as they’re available. Better get your orders in soon.