Key to the issue is to look at Sun's strengths. It's best at the big iron and that's where it's major markets lie. Mission-critical, large multi-processor systems are Solaris' meat and drink. At this level, absolute faith in and good relationships with your operating system supplier become as important as the software itself and Sun has built up a tidy stock over the years.
Now Sun says that Solaris is to go open source.
From a strategic point of view, this could be the start of a very critical period for Sun. It's too early to say whether the system software market has reached an inflection point, where the process of making it has become as important as the product. Yet Sun and open source are old friends. The company already has open source software on the market, including OpenOffice, components of the Gnome desktop, and the Tomcat servlet container. Of the important software products, only Solaris and Java remain closed.
However, even though its OpenOffice could be deemed at least a moderate success, Sun appears now to believe that, if it doesn't at least clothe itself with the glory that would accrue from joining the open source movement body and soul, it could lose out. Why else would it bother?
It's hard to judge what the true motives are. One source within Sun suggested that there remains a huge degree of uncertainty within the company about the move. And no details have yet been draped over the skeletal announcement: we know neither if the whole OS is to be opened and to what degree, nor has Sun divulged a timescale for the move. Whatever, reactions have been very mixed, with some analysts saying that it doesn't matter, some customers alleging that they want an all-in-one solution, while others welcome the move.
More importantly, we don't yet know if open source means true open source as in the GPL (General Public License). The last statement Sun made, following more goading from IBM on the issue, was about Java; chief operating officer Jonathan Schwartz said that opening the Java source would cause compatibility problems and that open source encourages forking adding that this meant the creation of many different versions of the same product, like Linux. In other words, you'll be able to get a copy of Solaris from a variety of sources, each of which is likely to be just a bit different from the others. It is this that Schwartz says he is most keen to avoid.
However, by retaining the degree of control that will be necessary to avoid forking, most of the benefits claimed by the open source community will be denied to Sun. The company would be left with a chimera of open source and Linux would continue to eat into Sun's market share, as it has done steadily over the last year or two. What's more, shareholders are unlikely to be convinced by Sun's apparent giving away of its crown jewels.
It seems that, unless Sun is ready to bite the bullet and do it properly, the halfway house it is proposing could well be the worst of both worlds, satisfying neither the one-stop-shop buyers, nor those clamouring for Solaris' source code. And consequently, the markets will be even more unhappy.