A lot of people have been busy commenting on the EU’s investigation into the competitive dynamics of Oracle’s proposed acquisition of Sun, so I thought I’d join in.
I no longer work for any of the parties involved, nor to my knowledge do I own any shares in any of the parties. I am a citizen of a member state of the EU. I have recently left Sun Microsystems where I was the Chief Technologist in their European sales and support organisation and where I represented Sun on the NESSI steering committee. In this role I was one of the contributors to NESSI position paper : European Software Strategy. I was not privy to any private ‘material’ information while employed at Sun.
The database market, even if narrowly defined, is characterised by a number of players, some of them very large and in the case of IBM & Microsoft, the subjects themselves of previous monopoly investigations. The EU has every right to investigate the proposed acquisition to see if it directly leads to anti-competitive behaviour. Anyone who thought they wouldn’t was being naive. But since everyone seems so free with their advice to Ms Kroes, I shall add mine and state that I hope that the investigation and decisions restrict themselves exclusively to “single market” and competitive/anti-competitive issues. I also hope, that at this time in particular, the Commission does not take the opportunity to act in a protectionist manner on international trade issues.
I tend to agree with those who argue that even the database market is complex and diverse and that Oracle’s leadership is under competitive pressure from established players such as Microsoft and IBM, from open source start-ups including at least two MySQL forks, new appliance vendors such as Greenplum & Netezza, new entrants to the data management market such as the no-SQL players, and from cloud providers such as Amazon, Google & Cloudera. It should be recognised that the most devastating and threatening competition will come from where you least expect it, best articulated in the Monty Python, Spanish Inquisition sketch,
Cardinal Ximenez : “Nobody expects the Spanish Inquisition, our main weapon is surprise…”
In his paper, “The End of an Architectural Era (It’s Time for a Complete Rewrite).” Prof Michael Stonebraker and his collaborators argue that the commoditisation of the relational database reduces its efficiency in doing what it originally did best, i.e. OLTP and enterprise solutions and requires rethinking and re-writing. They argue that they
….predicted the end of “one size fits all” as a commercial relational DBMS paradigm.
It’s another indicator of the competitive advantage new entrants have in the market, and the legacy inertia, both inside itself, and amongst its customers that Oracle has to deal with to keep up.
So the market is large, there seems to be vibrant competition, new entrants are disrupting it, and the established player’s legacy is inhibiting its ability to compete. The new players also show the difficulty in defining a market’s boundaries, is it about relational databases? Databases, SQL databases, or appliances? I think that Stephen O’Grady articulates the market dynamic well in his Blog, Oracle, MySQL and the EU: The Q&A, and also captures some of the best & worst contributions to the debate so far. While the discussions around the ownership of copyright, trademark and licence are fascinating they don’t seem too relevant to anti-competition regulation, as far as I can see. Oracle’s past track record, current public statements and future best interests both economic and in reputation don’t suggest that they’ll trash MySQL to reinforce the market leading position of Oracle RDBMS, even if one accepts that the market in question is an SQL/Relational Database market. Neither does Oracle’s record suggest that they’ll starve it of resources to allow it to die. (It would probably help their case if Oracle were to state this in writing to the Commission.)
Anyway the open source rights are today ensuring that the MySQL code survives and thrives, and yesterday, Amazon announced that they were offering MySQL machine images, I assume they’ve licensed it from Sun in order to get the trademark right to use, but it’s another proof point that the cat is out of the bag.
The Commission has more research resources than I do but I think they should come to the same conclusion I do; I just hope they keep their eyes on the public interest, total market issues and competition policy. By buying Sun, there is no doubt that Oracle ups its game in an ability to compete with IBM, HP and Microsoft, in the broader IT market customers get more choice; at the end of the day, this may not be about the database market, it becomes about the IT supply market as a whole.