I feel for the folks at Sun Microsystems right now. This is the end of the road. Make no mistake about it. There will be blood. A lot of blood. While a surprising amount of Sun may just live on under the Oracle logo, there's no denying that a massive portion of the company will soon be out on the streets, looking for work.

Speculating on just what Oracle will keep is an enjoyable passtime, even though it is rather gruesome. There's a lot inside of Sun, and while we did this with the IBM rumors, there are a lot more questions about what exactly Oracle wants here. So let's begin.

First, there's MySQL. Obviously this is the jewel and, really, Oracle paid about US$7 billion out of the total $7.4 billion offer just for MySQL. Watch with horror, or glee, as MySQL slims down, sheds enterprise features and generally morphs into a Microsoft SQL killer. 

Next on the list is Java. I find it hard to believe that Java as a product can be monetized by Oracle, even with its obvious understanding of the corporate sales process. It's not easy to sell a language, especially when much of the language is made up of frameworks and APIs that have competing open-source alternatives. Java's database stuff is really first rate, so I don't expect Oracle to change that side much. As for the Web and Java EE, I'd imagine Oracle will leave them both alone after the Java EE 6 update arrives. I'd even imagine that Java EE 7 won't materialize at all. If the OpenJDK ever arrives, Oracle might just use it as a "We love open source" flagship and be happy not to have to worry about the JCP anymore.

But the real question around the OpenJDK is what happens to the governance committee? They're already far behind schedule. And with the OpenJDK still missing in action, it's unclear if it will even arrive at all. I don't see Oracle being the company to set up a long-term non-profit shepherding organization. I also don't see it wanting to run the JCP. If it were to do the right thing, Oracle would just spin the JCP off into a standards body-like form and give it the ability to run the OpenJDK as well. But I guarantee you that's not what's going to happen.

Remember last year when Oracle announced its first hardware offering? A big old database server rack that pushes SQL into the actual drive controller boards? Certainly an interesting idea, though one that seemingly admitted to the world that Oracle had hit a brick wall as far as concurrency and optimization were concerned. Still, it showed that Oracle was stepping into hardware. Now, I'd expect that Sun's entire hardware division will be molded in this image. The StorageTek stuff won't be touched, I'm sure, as that's the golden goose at Sun. It was a growing arm of the business and one that's undeniably bound for permanent growth. But the actual server-side of Sun's business is doomed. Watch Oracle cast off this area of the business after all of the existing hardware contracts are up. I'd imagine a few Department of Defense contracts may force Oracle to keep a few server areas open, but for the most part, I'd expect this whole section of Sun to be eviscerated slowly over time.

Sun Labs, unfortunately, will be the first thing to go. Oracle has no interest in building media storage boxes or in experimenting with intervals or music search software. On the bright side, this will spark a hiring feeding frenzy, as many of the bearded old geniuses at Sun will be reentering the jobs market. They'll just have to get used to the idea that they're no longer working somewhere that lets them experiment with whatever they want.

ZFS is a tricky one. Oracle's been touting BTRFS as the ZFS killer and has even managed to get the Linux Foundation to put it on the road map for the future. I have to wonder if Oracle will realize that ZFS is ready and BTFS is not, and will thus form around ZFS. Plus, ZFS has a ton of patent protections around it, and BTRFS is only just beginning to get that type of consideration. I'd actually imagine that Oracle will do the right thing here and jump on the ZFS bandwagon. Now, we just need to get some certified people in the field who understand that file system. As it stands now, anyone who uses ZFS in a commercial environment is beholden to Sun if something goes wrong, since they're the only ones who can fix it.

Cloud is another tough one to judge. Sun has been touting its cloud stuff for a few months now, though said stuff doesn't actually exist, yet. On the other side of the coin, Larry Ellison has made it known that cloud is a buzzword, not worthy of his attention. Does that mean the cloud project at Sun will be ignored and left to die? Probably.

Solaris? Who knows. I'd say kill it, but then Oracle might just want to base its servers on it. Nothing's more proprietary than having your own operating system, so Solaris may just end up being the database OS. 

Another thing to consider in all of this is the fact that Sun was purchased for a full billion dollars less than BEA. It's around $3 billion less than Peoplesoft. That's just remarkable.