Tom Foremski, the former reporter for the venerable Financial Times, posted a pretty interesting tip on his Silicon Valley Watcher blog this Sunday. According to his sources, Salesforce.com approached Oracle to check out the possibility of a sale—at $75 per share. Shares of SFDC closed Friday at $50.87. You can do the math on that one.

While it's always been in the back of my mind that Oracle would acquire NetSuite for its on-demand capabilities—Oracle CEO Larry Ellison & family own a major stake in NetSuite—a Salesforce.com union makes possibly even more sense. Salesforce is the undisputed leader in the SAAS (software-as-a-service) sector. And Oracle likes to buy winners. Here's the other thing. Despite Oracle's long-standing history of hosting applications, its "on demand" business is just that: hosted software. It's not real on demand if you're going to take the stance that on demand means multitenant, built from the ground up. Oracle's got nothing going on, on that front. Even with Siebel On Demand.

And here's another really compelling reason Salesforce and Oracle would make a good pair: While competitors and naysayers denounce the whole necessity of another proprietary programming language (and maybe they're right; time will tell) Salesforce has built a pretty compelling offer with its Force.com platform and Apex on-demand programming language. And it's tightly linked to Oracle's database, in a good way.

I sat down with Salesforce co-founder Parker Harris shortly after eWEEK first broke the news about Apex in October 2006. Here's how he explained the company's Apex and the connection to the database.

My question: The comparison to Apex has been to Java, and maybe that's not accurate. But if so, how is Apex different? And why wouldn't a developer just use Java?

Parker's answer: What's different about Apex is that it's an on-demand programming language. By that I mean that if you're using our sales force automation [software] you obviously can get up and running, and you don't have to buy the hardware. The same thing is true with Apex. If you're a developer you can start programming. What's different about Apex is it's running in the database basically (it's actually running on application servers). But by saying it's running in the database, it's very tightly bound to the database. And one of the key things you can do with Apex that you can't do with Web services is you can have transactional control of that database. Web services doesn't offer that. If you look at the Web services world, they're trying to solve that with various technical ways, but you can't open a database transaction and have that database wait there for whenever you decide to call back later and issue a close statement. So that's a huge hurdle that we've overcome with Apex.

But the other thing, in terms of a language, is it's really, really easy to use to write the logic, the business logic that needs to happen within the service. The power of having a super clean language like that right at the database level is super powerful. You can create these really complex applications really quickly.

Foremski brings up a good point in his blog: that Salesforce CEO Marc Benioff would be a good replacement for Ellison, who has, it seems, no actual succession plans in place. At least none that are public. Benioff, like many Silicon Valley CEOs and execs, is an Oracle alum. When he left the company Ellison provided cash to get Salesforce off the ground. As the head of Salesforce, Benioff has been really accessible—the complete opposite of Ellison. Let's hope that doesn't change should this prediction ever come true.

Here's one more point that Foremski makes, to borrow liberally from his blog:

"Mr Benioff needs a new challenge, he appears to be losing interest in Salesforce, or at least reducing his financial interest in his company at a rapid daily rate. He has been selling 10,000 Salesforce shares every single day since 21 August 2007. Before then, he sold 20,000 shares every day since 14 November 2006. Prior to that date, Mr Benioff sold thousands of shares every day in variable amounts since 31 July 2006."

Pretty compelling reasons all the way around. Oracle, the ball just might be in your court.