Oracle owning MySQL would be worse than the proverbial fox in the henhouse.

MySQL has come a long way and it has a bright future ahead of it, but only if its continued development is in the hands of someone who truly wants to compete with Oracle. Not Oracle itself.

There's a million things Oracle could do with MySQL. Databases are a huge field, they can promise to spend a lot of money and do a lot of work in all sorts of areas. But there are two things Oracle will never do:

• seriously improve MySQL in ways that hurt Oracle's own high-margin business
• aggressively sell and promote MySQL to customers as an alternative for (not all, but ever more) database purposes for which Oracle is traditionally used

No company in history has "cannibalized" its market-leading high-priced cash cow with an aggressively priced alternative. It always takes another vendor to compete fiercely with the overpriced incumbent. Once the two are under the same roof, it just doesn't work anymore.

An Oracle-owned MySQL would have no more teeth. It would be steered in ways that might affect others in the market but certainly won't hurt Oracle. This is even worse than the proverbial fox in the henhouse. The fox will eat the hens because he's hungry. But the hens are no threat to the fox, while MySQL is the single biggest threat Oracle has faced in its corporate history. It wants to get rid of it, not in a simple straightforward way but in a disguised way that would still have essentially the same negative effect for customers.

Therefore, antitrust regulators should not allow Oracle to simply acquire MySQL as part of Sun. Promises for what Oracle will do or will not do over time don't really help. They may look like a solution but they aren't. There's no substitute for having a truly motivated player in the market who owns the related intellectual property rights (IPRs) but has no conflict of interest due to another database business.

The simplest and most reliable solution would be to require Oracle to sell MySQL to a suitable third party. Then the market still has a serious competitor that customers can turn to for everything related to MySQL, and on that basis, Oracle can have the rest of Sun. The second-best approach would be to let Oracle buy all of Sun including MySQL but to ensure that other companies have a reasonable (not only a theoretical) opportunity to provide MySQL-related innovation. In order for third parties to be able to do that, they would need to be given access to the existing MySQL code and future releases on a basis that enables them to serve, independently of Oracle, the entire MySQL ecosystem, not only a subset of it.

Innovation always takes the combination of two things: the means and the motivation. Oracle would certainly have the means: it has deep pockets and knows the business. But its motivation would always be to keep MySQL limited to the web and the low end. Based on the views of MySQL that Oracle has expressed repeatedly over the years, there's no doubt that that's what they want it to be. It's just not what it should be in the interest of the market.

So there has to be someone else who has the motivation to succeed at all levels and in all areas. Ideally, a serious competitor gets to buy the MySQL-related IPRs to continue to grow that successful business independently. In the second-best scenario, there has to be a fundamental change to MySQL's licensing rules in favor of independent innovators so they get, in addition to the motivation they have, the means to succeed. Those new entrants wouldn't have the famous MySQL brand and they would need to build a whole new business, but at least there could then be another significant competitor (or several of them) after some time.

Anything less won't do. Promises for future behavior always come down to "trust me, trust me" (against all economic logic).