It had very little presence as a software player, beyond providing Solaris Unix operating system technology for its own boxes. And if it is not careful, the company could soon end up back in that position.
Through the late 1990s, when it basked in the glory of Java and its association with the Web, the company used the rapid rise in its sales to fuel a vast expansion of both its systems' hardware "footprint" and the level of recognition of its software technology. During this period it acquired a whole raft of Java application development, Web and middleware technologies, database software, and server and storage system hardware.
There is no doubt that Sun has a formidable research organisation at its heart. However, through its acquisitions, it has built an internal culture that is schizophrenically split between an "old guard", which is keen to build business primarily on Solaris and Sparc processor, and "new blood", which is happier with an open approach to hardware platforms and operating systems and wants to concentrate on delivering value-added software solutions for customers.
The problem is that the company oscillates between these two positions, and often its products and customers suffer.
Sun's attitude to Java solutions is a case in point. At the start of its foray into solutions, its iPlanet alliance with AOL/Netscape naturally led Sun to act as a market-maker for others, rather than selling Java products itself. The implied rationale was that iPlanet would act as Sun's software champion. But for some time iPlanet was publicly positioned as an independent software supplier.
The old-guard personality within Sun wanted to be seen as an impartial technology lab, while the new-blood people at Sun wanted the company to influence the path that iPlanet took in order to build solutions credibility. The result? Throughout the whole of iPlanet's lifetime, some really great software products languished and Sun failed to capitalise on the potential of its Java innovation.
A couple of weeks ago Sun announced a new entry-level server running Linux and based on an Intel Pentium processor "due to customer demand" - a very new-blood move.
However, the old-guard influence puts Sun in a difficult position. If it does not offer Linux when IT directors ask for it, then they will naturally talk to the likes of IBM, which has announced a migration program to transition Solaris users onto Linux. But if it does, then it risks being forced to try to compete with companies such as Dell and HP/Compaq, which are much more geared up to sell cheaper systems at higher volumes. That way lies trouble.
A full-on software and services business could save Sun from this dilemma, but unless it diverts serious management energy to banging heads together internally, its software and services organisations will continue to be confused and under-valued.
Whatever happens, Sun is unlikely to disappear completely - but if it does not sort out its corporate personality disorder, soon, then you should start asking your Sun representatives some really hard questions.