Alfresco CTO, prolific commentator and Arsenal fan Matt Asay has posted an interesting Tweet suggesting that "Oracle [is] circling Red Hat again". Matt says this is "just a rumor, but seems reasonably well substantiated". Ignore it if you like but as Matt tipped the VMware-SpringSource deal well ahead of the event and has his ear very close to the ground in the OSS community (you might say he has a lot of open sources) that's a rumour worth exploring in my view.

Why would Oracle want Red Hat? Let's count the ways.

1. Red Hat is a market leader. It's the greatest success story in OSS history by pretty much any indicator you care to use: revenue, profit, partners, business user base, whatever.

2. It's still a growth story. Turnover and profit continue to head north and there is some evidence that the down economy is helping to perpetuate this trend.

3. Oracle knows how to make mergers work. Looking at the facts from a purely financial standpoint, no company except perhaps Symantec has done a better job of integrating purchases.

4. Red Hat is not a stack player. Red Hat revenues are derived largely from the RHEL operating system and JBOSS middleware so Oracle wouldn't need to deal with the ton of crossover involved in deals like PeopleSoft and a combination would cater to those who love Red Hat but would like it to offer a broader suite of products (the old 'one-stop shop versus best-of-breed' argument) to reduce the admin overhead of dealing with large rosters of suppliers etc.

5. Red Hat would imporve Oracle's virtualisation and operating system story. Red Hat is making  big push on the KVM front and has the most popular open-source OS distribution in business. Oracle has been a little backwards in coming forwards on both fronts and needs to place its bets on virtualisation. How many CIOs even know about Oracle Enterprise Linux?

6. Oracle has solid experience in open source. Oracle was an early adopter in seeing Linux as opportunity rather than threat and made some of the earliest 'name' acquisitions. 

OK, so what about the flip side?

1. The price is high. Red Hat stock is at a premium and any deal would reflect that fact -- this would be a maybe $5bn valuation.

2. JBOSS and BEA might not blend. There's no doubt Oracle would be doubling up on middleware.

3. Red Hat's value might lie to a large extent in its indie status. This might be the big fly in the ointment: how would Red Hat buyers, developers and partners like to live in OracleWorld? 

4. Er that's it.

So will the deal happen? I think there's a high chance. Here's why.

1. Larry Ellison has balls. A logician (or accountant) might argue that if Oracle were going to buy Red Hat, last year might have been a better time as RHT shares took a tumble but Oracle's CEO is nothing if not an autocrat and he has the capital and the backing of a board and shareholders that is almost wthout peer. This is a CEO who gets to compete in dangerous sailing races and pilot MIGs. He signs off deals to buy hardware companies, pursues hostile bids, grabs companies and swaps out managers. If he wants to push the Red Hat button who is going to stop him?

2. Oracle needs to keep buying. In order to perpetuate growth, the sheer scale of Oracle (and other large software companies) virtually mandates an aggressive merger-and-acquisition strategy.

3. Red Hat's CEO is a pragmatist. Jim Whitehurst is almost unique among software CEOs in that he doesn't appear to have an ego the size of Ellison's ocean liner. Red Hat could use Oracle's channels to market as it builds up its reseller and partner heft and of course it would get a vast amount of sales and marketing budget. Also, with stock concentrated in relatively few hands, this could be a deal that could go through quite quickly.

Oracle to buy Red Hat? It might sound an odd combination but we've seen this movie before, haven't we?