Letter to the EC on the Oracle/Sun Takeover
Dear Commissioner Kroes,
Last week, former MySQL CEO Mårten Mickos wrote you a letter urging approval of Oracle’s takeover of Sun Microsystems ((via Matt Asay’s CNET Blog – Mickos letter to EU: “Approve Oracle-Sun deal”)), asserting that Oracle’s ownership of MySQL (as part of the Sun acquisition) will increase competition in the market.
As a long-time MySQL user, a former MySQL AB staff member ((from 2001 to 2004, I served as MySQL’s community liaison)) and a participant in or consultant to a wide range of other open source and free software projects ((including multiple years serving on Free Software Foundation’s license compliance team, working as a Mozilla Foundation staff member and volunteering for the Open Source Initiative)), I found Mårten’s conclusion to be optimistic at best.
Oracle’s ownership of MySQL will lead to what the commission fears – greater costs and less choice in the DBMS market.
In making this point, I’ll challenge the three key arguments made by Mårten:
- MySQL and Oracle do not compete
- Oracle has many compelling business reasons to continue the ramp-up of the MySQL business
- Oracle’s ownership of MySQL will increase competition in the market
Also, please note that I have no financial interest in Oracle or Sun and haven’t been paid to write this article. ((The founders of MySQL and many of the early MySQL staff are friends, which likely influences my thinking. I don’t know what options or financial arrangements friends and former colleagues who have a stake in Oracle or Sun have in place, but I’m sure that some friends will have significant benefit from a sale of Sun to Oracle.))
MySQL and Oracle Compete
First, it is very clear that MySQL is a difficult and disruptive competitor for Oracle. Imagine yourself Oracle’s position: most of your present and future customers use MySQL at no cost, and the combination of open source communities and the commercial entities backing MySQL work in a distributed fashion to erode your key advantages and populate many niches in the market.
Oracle’s acquisition of Sun would provide them with a way to control this competition.
We don’t have to rely on imagination to see the competition between MySQL and Oracle. Even a cursory examination of the market leads one to the same initial conclusion reached by the commission that, “… Oracle databases and Sun’s MySQL compete directly in many sectors of the database market …” ((EC Press Release: European Commission opens in-depth investigation into proposed takeover of Sun Microsystems by Oracle ))
A volume of evidence demonstrating the heated competition between Oracle and MySQL can be found online, including Oracle’s acquisitions of key pieces of technology licensed to MySQL ((Oracle acquired Innobase OY in 2005, followed by Sleepycat Software Inc. in 2006. These vendors licensed transactional database engine technology to MySQL that allowed MySQL to more effectively compete in Oracle’s space.)), benchmarks ((MySQL.com: February 2002 eWeek Benchmarks)), case studies ((Mysql.com case studies: ThePhoneHouse consolidates its eCommerce Systems on MySQL Enterprise, MySQL.com: Citysearch Saves Over $1 Million Using MySQL (pdf), … )), migration toolkits ((MySQL.com: Introduction to the MySQL Migration Toolkit, Oracle.com: Oracle Migration Workbench)), presentations given by MySQL staff at the MySQL conference ((MySQL User Conference 2005: Migration from Oracle to MySQL, MySQL User Conference 2006: MySQL Migration Toolkit, MySQL User Conference 2007: MySQL for Oracle DBAs and Developers)), attendance of the MySQL User Conference by key Oracle management ((Oracle VP Ken Jacobs has attended and spoken at multiple MySQL User Conferences)), articles in trade publications and recently leaked Sun Microsystems internal documents ((Wikileaks.org: Sun/Microsystems ‘Project Peter’ targets Oracle to MySQL migrations to boost sales)).
Oracle’s Compelling Business Reasons (or, Why Oracle’s Acquisition of MySQL Won’t Increase Competition)
Mårten wrote that, “Oracle has as many compelling business reasons to continue the ramp-up of the MySQL business as Sun Microsystems and MySQL previously did, or even more”, but did not elaborate on what these business reasons would be.
We know that Oracle will seek to use MySQL to provide maximum value to their shareholders, but this is not the same as a “compelling business reasons to continue the ramp-up of the MySQL business” nor is it the same as, “(increasing) competition in the database market.”
Oracle has a large and successful business. Its net income for the first fiscal quarter of 2009 was reported at 1.1bn USD ((As reported by BusinessWeek: Oracle’s Earnings: Summer Doldrums Set In. Note that this was a rather weak quarter, as previous recent quarters reported nearly twice the income.)) The direct commercial value from licensing and services that it would be able to extract from MySQL would be trivial compared to this (and would likely be at a much lower margin than services and licensing it is accustomed to.)
Some reasonable tactics and strategies for an Oracle who has acquired MySQL would include:
- Using MySQL to price-target customers, ensuring that each customer pays as much as possible to Oracle. In the past, Oracle has reduced prices on a case-by-case basis to help retain customers who have “defected” to MySQL. The reduced competition in the marketplace will give Oracle more control, especially over large institutions who currently rely on MySQL Enterprise.
- Using control of the non-software MySQL assets (such as domain names, documentation, trademarks, conferences, …) to manage competition in the MySQL space.
- Managing the rate of innovation in the MySQL product, so as to ensure that price-targeting can be effective.
- Continued support of the open source version of MySQL, so as to ensure that other open source competitors do not become prominent enough to challenge Oracle’s business.
Oracle will engage in strategies such as these for as long as it has a compelling business reason to do so, but certainly not out of concern for or in service to the market.
Closing
Commissioner Kroes, I won’t presume to advise the commission on the best path forward, however I do hope that you have a clearer view of the facts.
Open source and the market forces that supported MySQL’s rise to prominence and allowed it to compete with Oracle will exist regardless of what the commission chooses to do.
If Oracle acquires MySQL, then the market will be hindered for the next three to five years. Customers will pay higher prices. The open source community will need a few years to route around Oracle’s control. Current MySQL customers will be faced with challenges as they decide whether or not to stay with an aggressive vendor who now has much more control of a database that they often rely on to serve the online market.
If Oracle does not acquire MySQL then it will still have significant influence, as it controls a key MySQL resource in the form of InnoDB. This is something that the open source space is still wrestling with, as various engines and forks attempt to deal with the problem.
If the commission truly wants to foster competition, a middle road would be to allow Oracle to acquire MySQL on the condition that the database, engines and documentation are released under a permissive open source licence, such as the New BSD license ((http://www.opensource.org/licenses/bsd-license.php)). This would allow Oracle to make the acquisition that it so desires without having to spin off MySQL and would foster a great deal of competition in the market, as no single party would be able to control the integration of MySQL with other products.
Posted on Thursday, October 15th, 2009 at 12:53
You can follow any responses to this entry through the RSS 2.0 feed.
You can leave a response, or trackback from your own site.12 Responses to “Letter to the EC on the Oracle/Sun Takeover”
Leave a Reply
Comments are moderated. I delete obvious SEO attempts and other content-free comments.
October 15th, 2009 at 13:21
Hi Zak,
You say “Oracle’s acquisition of Sun would provide them with a way to stifle this competition,” but don’t provide any detail about how they might do that. I think Marten’s letter argues that the users of MySQL are so much more powerful than the owner of the MySQL assets that the acquisition actually does not give Oracle any way to stifle the competition.
JD
October 15th, 2009 at 16:17
Hey JD,
I’ll address how and why I would expect Oracle to do this.
Thanks!
–zak
October 15th, 2009 at 23:29
I agree with Marten. We all know that for quite some time Sun hasnt been the one distributing the most powerful version of MySQL, that there are plenty of professional consulting alternatives and that Oracle stands nothing to gain from dropping MySQL. I am not worried at all.
October 16th, 2009 at 12:55
I don’t understand all the fear surrounding MySQL being acquired. If Oracle doesn’t do a good job someone else will fork the code and do it better.
Anyway, I think fueling this debate to give the EU reasons to block the acquisition will do more damage to MySQL users than any plans Larry might have.
October 16th, 2009 at 19:21
I disagree with your representation of Marten’s argument and think your evidence is weak. For example, I think the e-week benchmark is silly and I chuckle every time I read about it in MySQL PR. I am wary of any benchmark that is won by using the query cache. Besides, the e-mark benchmark was published 7 year ago.Where are the recent results?
But even if your evidence is strong, you claim there is significant competition between Oracle and MySQL and that was not the main point in Marten’s letter.
The points were:
1) Oracle has reasons to do right by MySQL
2) The majority of MySQL users/consumers don’t have much at risk.
The second point is the focus of Marten’s letter. GPL is fine for the majority of MySQL users, therefore a fork of MySQL is good enough for them. They may have to pay to support fork contrary to the current model.
October 19th, 2009 at 14:40
Hey Lukas,
Thanks for commenting!
I agree with Mårten on any number of points. However, the assertion that Oracle’s acquisition of MySQL will help the market is over-the-top (unless Mårten is predicting that ownership of MySQL by Oracle will force the market to adopt solutions that they have better control of.)
However, I hope that I’ve stated this all clearly in the now finished letter.
October 19th, 2009 at 14:43
Hey Mark,
Thanks for commenting!
The results of the benchmark aren’t important. The important bit is that a major IT publication saw fit to compare MySQL and Oracle quite some time ago, when they were much less similar.
As for Oracle having reasons to to right by MySQL, neither you or Mårten supported this point.
I do agree that most users have little risk and won’t care.
Cheers!
–zak
October 19th, 2009 at 16:34
Zak,
This letter is better than the first version, whether or not I agree with it. This letter has 3 points at the top and sections for the first 2 points. Where is the section for the third point?
October 19th, 2009 at 17:07
Hi Mark,
I’m glad to read that you find the (nearly) final version better than the draft. That is how things are supposed to work.
Thanks for the note on the mismatch. Points two and three blur together – I’ve modified a title to reflect this.
October 20th, 2009 at 07:06
Hi Zak,
I’m somewhat confused by your recommendation:
“…allow Oracle to acquire MySQL on the condition that the database, engines and documentation are released under a permissive open source licence…”
As it implies there is somehow something anti-OSS about the GPL license.
I’ve always understood the adoption of the GPL by MySQL to be our best attempt to capture the “quid pro quo” belief of the MySQL founders and ensure that what they open up does not become closed.
October 20th, 2009 at 12:14
Hey Anonymous,
The issues are orthogonal. The GPL is a fine licence for many cases, however it’s use in dual-licencing schemes can be troubling. More on this in a moment.
The commission is concerned that Oracle’s acquisition of MySQL would lead to greater costs and less choice in the DBMS market. One of the key levers that Oracle has over the market is the exclusive right to control proprietary use of the database. BSD licensing would remove this lever and level the playing field. Oracle would still own trademarks, have the opportunity to retain talented Sun/MySQL staff, etc. It would balance the power that they acquired. Frankly, I’d be stunned if the commission did this. The suggestion is mostly in place as a way to help the commission think about the dynamics of the market.
As for the GPL being anti-FLOSS, I think that is stretching what I’ve said rather far. The restrictions of the GPL can work well when no party has an unfair advantage over the other. Take Linux as an example – many parties have a stake in the code base, but no single entity controls it.
In the case of MySQL, the dual licensing scheme gives the owner of MySQL financial incentive to encourage non-GPL extensions, to push customers away from GPL use and to not use other GPL-licensed code, for fear of losing its proprietary advantage.
October 23rd, 2009 at 11:22
[...] writing their own open letters, recommendations and editorials on the matter. Zak Greant’s Letter to the EC on the Oracle/Sun Takeover, Lukas Kahwe Smith’s MySQL needs to be saved from Oracle?) and Brian Aker’s RMS, GPL, [...]