Since sometime last summer a company called Southeastern Asset Management has been both accumulating Sun shares and hawking them to various big players. They got nowhere with HP, nowhere with Microsoft, nowhere with the Koreans, and nowhere with IBM - but recent political change has changed that: bringing IBM back to the table with no anti-trust concerns between them and a kill Sun agenda.
Thus this week's carefully calibrated leaks to a select group of trusted friends in the media are intended to do more than just raise the price for IBM: they're also intended to remind the refusers of what they've lost and, perhaps, bring an alternate with the money and ego to play back to the table.
Sun's major shareholders, and particularly Southeastern, will, of course, win if this goes ahead - but it would be horrible for everybody else: just as Microsoft buying Apple would end the desktop competition and thus raise everyone's costs while weakening American economic productivity, this deal would kill server hardware competition while weakening American economic productivity.
It's easy to see what might drive IBM to this decision: they've got the money, anti-trust in the Obama era is just a dollar issue, Sun's Unix licenses predate Novell's and have more validity, and right now IBM can't begin to compete with SPARC/Solaris on price, performance, reliability, or innovation.
Thus the bottom line for IBM is that buying Sun is a cheaper and more permanent way of killing it than competition - meaning that you can hope that IBM would treat Sun the way they have Lotus, but they're a lot more likely to treat it the way their proxy, Exxon, treated Zilog or the way they themselves treated Informix.
So is there a good side to this? Well, maybe: I think you can be assured that pretty much the entire Sun brain trust -the people who drove hardware and software innovations from NFS, SunView, and BSD Unix to dTrace, ZFS, Crossbow, and Rock would leave the company at their first opportunities - and some of them will go on to form new companies doing new and wonderful things.
(And, for the wildly optimistic, it's possible to argue that, because both SPARC and Solaris are fully open source, IBM could end up killing off nothing but the most ineffective server marketing organization ever deployed internationally.)
Meanwhile there's a wild card: every significant security installation in the DOD uses Sun servers with Sun Rays - and IBM has a Lenova hangover. Thus predictions on whether this deal will ultimately go through are intensely political in nature because they depend on the extent to which you want to believe that any deal obviously bad for both the American economy and American national security will attract high level political support.
Personally I think the Weathermen will love the deal - but your views may differ.