Consider the market. WebLogic remains one of the technically best J2EE servers available, but the capabilities of J2EE servers have rapidly been commoditised. For those who want adequate functionality at an unbeatable price, there's Open Source JBoss. For those who want the guarantee of commercial support, there's JBoss with commercial support. For those whose paranoia drives them to pay any price for the reassurance of a large corporation behind their app server, there's always IBM's WebSphere. And now with Apache's Geronimo getting ready to roll, the market is hotting up, even in the Open Source segment.
Who in their right mind will buy WebLogic anymore?
I'm sure this idea has been kicking around at BEA for a while, but as long as people were still buying and paying for WebLogic licences, there wasn't a strong reason to consider it seriously. Now that the license market is shrinking, BEA will have to do something dramatic to stay relevant.
Open-source WebLogic - that's what I think they will have to do. They'll have to make their money through support and professional services. It may not buy as many Ferraris, but continuing with the old model will have them walking to work at the current rate, assuming there is some work left. They really have no choice.
I'm beginning to think the proprietary software model is itself a fly-by-night model. Make your money on licensing while you can, because the window closes as soon as Open Source hits the market. Move on to the next segment undiscovered by Open Source and milk it for all it's worth while that window is open, etc.
[I also think Sun will Open Source the Java VM, tools and core libraries this year, but that's another story.]