I like the following bit of this article in InfoWeek:

Some customers say they aren't sure how they'll benefit. "Open-source Solaris? We're not sure what that's about. What's their market strategy?" asks Kyle Forster, software architect for the state of North Dakota, which is gradually replacing Solaris servers in its data center with Linux servers. The state runs some Java-based applications, but Forster isn't sold on open Java either: "We like vendor support. It's not like we're going to go in there and start changing the code base ourselves."

The trick for McNealy will be to change minds--and convert no-cost public versions of Solaris and Java into not only a sustainable business model, but one that grows. In the 10 years since Java's initial release, Sun is a changed company and a smaller one, having shrunk from 43,500 employees to just under 32,000. The hope within Sun is that open-source versions of Solaris and Java middleware will rekindle interest in the company's underlying technologies and that, as more developers build upon them, revenue opportunities will follow. Says McNealy bluntly, "We've got to make some money, somewhere, somehow."

That last line is so telling. Right now, Sun is flailing around, just throwing stuff at the wall and hoping that some of it sticks. Witness their recent acquisitions - they just spent $4.5 B on stuff they don't need, cutting their monetary reserves by 50%. Meanwhile, they haven't grasped the essential issue with making their core assets open source - there are plenty of smaller, nimbler entities out there who can offer service and support for those assets at a much smaller price point than Sun can afford.

In watching Sun, I'm reminded of ParcPlace-Digitalk in the mid 90's, but with a far larger pile of cash to burn. Same frothing and mindless management decisions though.