MLex Comment: Competitive dynamic between Oracle, MySQL could lead to Statement of Objections

Author: Dafydd Nelson
26 Oct 09 | 13:56 GMT+1

IN BRIEF
In its review of Oracle's purchase of Sun Microsystems, the European Commission could be exploring the possibility that MySQL actually competes with Oracle's proprietary database offerings. Should such a theory be substantiated by hard evidence there is every reason to believe that the regulator could issue a Statement of Objections in the coming weeks, if not days.

In its review of Oracle's purchase of Sun Microsystems the European Commission could be exploring a possible theory of harm based on the removal of MySQL as an actual competitor to Oracle's proprietary database offerings.

This could be the case no matter which market definition is arrived at as evidence points to MySQL being adopted by both large (enterprise), and small and medium-sized businesses (SMEs) alike.

The regulator has been seeking to substantiate whether there is one overall market for databases, or whether the market should be divided into several sub-markets according to business size. Furthermore, if sub-markets exist, the commission must decide whether Oracle and MySQL directly compete.

Should such an actual removal of competition be substantiated by hard evidence, which has been collected over the course of the past few months, there is every possibility that the commission could issue a Statement of Objections in the coming weeks, if not days.

This could force Oracle into considering whether to placate these concerns by offering remedies, either of a behavioural or a more drastic structural nature: this is something the company, in public at least, has seemed very unwilling to do.

Competition concerns may stem from the role MySQL is claimed to play in constraining prices charged by proprietary database makers like Oracle.

Even in markets where several strong competitors would remain post transaction, the removal of what could be deemed a 'maverick' price-setter is likely to set alarm bells ringing.

The commission has used this thinking before on several occasions, for example, in last year's review of StatoilHydro's deal with ConocoPhillips.

In referring to StatoilHydro's purchase of JET branded petrol stations in Sweden and whether JET had acted as a 'pricing maverick', the regulator said that it needed to determine if "the company has played a role in the market which is greater than its market share would imply at first glance" (see here).

In assessing the level of competition between Oracle and MySQL, as well as MySQL’s strength, the regulator will probably take into account a wide range of evidence from various interested parties including competitors, customers and the companies themselves.

- Database migration -

One way of assessing the competitive relationship is by checking for instances of substitution between the two products.

In tough economic times businesses attempting to save money may look into switching from expensive proprietary database solutions to free open source alternatives.

There is past evidence of large organisations migrating from Oracle to MySQL, indicating that a competitive interaction at least existed some time between the two.

For example, MySQL itself points to the decision by the Food and Agriculture Organization of the United Nations to switch over from mainly Oracle-based systems to open source in 2006. 

Indeed, database products used by SMEs are said to be essentially the same as those sold in the enterprise segment, only that those products produced by proprietary makers aimed at smaller customers are artificially limited in, for example, data capacity with a restriction also placed on the number of users.

MySQL does not place these sorts of limits, and can therefore be used by companies of all sizes and at a far lower cost.

Switching from proprietary to open source can result in significant cost savings, an option the commission would most likely be eager to protect.

- The enterprise segment -

MySQL's exact presence in the high-end enterprise segment of the database market has been a hot topic in determining whether it actually competes with Oracle. 

Despite the doubts cast on MySQL's high-end capabilities, many instances of its adoption by large organisations exist. Construction materials company Lafarge, for example, selected MySQL to operate "the real-time management of the different sales and dispatch phases of its Aggregates & Concrete business", according to the parties announcing the deal. 

And in April this year MySQL version 5.4 went into 'alpha' testing. This product is said to be more suitable for companies active in the enterprise segment.

MySQL 5.4 uses 'InnoDB', a powerful transactional storage engine. This technology is licenced by a Finnish company called Innobase, an Oracle subsidiary which it bought in 2005.

MySQL and Innobase have a long history of collaboration, leaving plenty of room to speculate why Oracle hasn't cut off MySQL's use of InnoDB if it has been the leading force behind MySQL's drive into the enterprise segment.

It could be that Oracle genuinely does not see MySQL as a threat; alternatively, Innobase could be locked into a licencing agreement with MySQL and the easiest way of dealing with the rising threat is for Oracle to buy MySQL.

And it can be argued that MySQL is Oracle's main competitor when considering that both are used on the Linux, rather than Windows, operating system.

This is another area in which the commission may be considering defining separate markets given that Microsoft's SQL database only runs on the Windows platform, whereas both Oracle and MySQL run on all operating systems.

MySQL's main installation base appears to be on Linux, where Oracle is said to be the number one.

- Open source rivals -

A further point of contention could be the ability of 'competing' open source products to step into MySQL's shoes should Oracle's acquisition be approved on an unconditional basis.

Of those currently available, the general purpose database PostgreSQL is seemingly in the best position to offer a competitive constraint if MySQL disappears into Oracle's hands.

This database's commercial impact, however, has been relatively insignificant to date, and it is thought that it could take years to approach the level reached by MySQL which is claimed to be far easier to use.

If Oracle’s purchase is deemed to remove a strong competitive force, without another open source product ready and able to take the role it plays in the market, it could mean that Oracle will have to make some tough decisions over the next couple of months on whether to offer commitments should a Statement of Objections be issued.