NQ Logic

Technology | Strategy | Consulting

Oracle and the Sun-set

Ironically Sun Microsystems Inc. is, at the present time, on the verge of bankruptcy because of an open source software.

In its last quarterly announcement, Sun had a 25% revenue drop, and reported a net loss of $120m for the first quarter of its 2010 fiscal year. The company reported a net loss of U$2.2bn in 2009, compared to a net loss of U$403m in 2008. Sun is supposedly "losing U$100m per month" and has announced another reduction of its workforce by 10% or 3,000 jobs, after a previous 6,000 reduction (Nov, 14, 2008).

Sun Setting

Sun had been in trouble since the beginning of the financial crisis, due to tight credit conditions, revenue dependency on high-end corporate servers, large recent acquisitions and a drop in technology spending overall from its large multinational clients.

A few players were interested (Cisco, HP…), but IBM was the most resilient of all potential buyers. Unfortunately, the discussion failed over the price tag, some post-merge financial guarantees, and other cultural differences that could have resulted in a successful business acquisition (April 05, 2009). Days later, Oracle jumped on the occasion and proposed to buy Sun for $7.4bn in cash, or $5.6bn net of the company’s cash and debt (April 20, 2009). Although the deal was unanimously approved by Sun’s board (July 16, 2009), and cleared by the US’s Department of Justice (Aug 20, 2009), a crucial point was itching the European Commission (EC).

The Commission has objected to the deal, pointing out that the acquisition of Sun’s MySQL database product (an open source database) by Oracle could drastically impact the competition in the database market in the 27-nation European space.

"The Commission has to examine very carefully the effects on competition in Europe when the world's leading proprietary database company proposes to take over the world's leading open source database company." – Neelie Kroes, Competition Commissioner, EU (Sep 3, 2009)

Since then, neither Oracle user group's public statement (IOUG), nor US Justice Department's antitrust division official reinforcement declaration (Nov 10, 2009), nor vocal political pressures from 59 US Senators (Nov 25, 2009) have influenced the European Commission's position on Oracle’s potential conflict of interest in the acquisition.

The initial reaction from Oracle was to broadcast in various press releases that it would “decline to divest anything” (and especially MySQL) from the transaction (Oct. 7, 2009), that European regulators have a “profound misunderstanding of both database competition and open source dynamics”, and that “Oracle and MySQL are very different database products” (Nov 9, 2009). After some pressure, Oracle has finally agreed to divulge its plan for the free product and that it intends to “continue to develop and provide the open source MySQL database after the transaction closes”.

The "statement of objections" from the Commission still remains on the table and is subject to intense discussions between Commissioner Kroes and Oracle President Safra A. Catz. The statement doesn't necessarily mean the Commission will block the deal, but it has a strong sense of command to satisfy European’s regulators. For example, last May the European Commission fined Intel U$1.5bn for engaging in “illegal” competitive practices, and last month Microsoft said the Commission had tentatively accepted its proposal that could end their decade-long battle over web browsers.

So in a last attempt to convince European officials to approve the deal, Oracle had requested another audience on Dec 10, 1.5 month prior the Jan 27 final deadline.

Sun Portfolio

Sun Microsystems, Inc. (which stands for Stanford University Network) was founded in 1982 by McNealy, Khosla, Joy, and Bechtolsheim. The company has today over 29,000 employees around the planet and is headquartered in Santa Clara, California. Over the years the company has ventured in many different business directions, and stabilized itself recently around three large components:
  1. Hardware business with a x86-based, a SPARC-based and a Motorola-based systems
  2. Storage business, with the Sun Open Storage platform, and the High-Performance Computing (HPC) for TOP500 systems and supercomputing centers
  3. Software business, with an Unix OS (Sun Solaris), a free OS (OpenSolaris), a Java Platform (Java, JES, Open ESB), a free office suite (OpenOffice), a virtualization product (VirtualBox), a tape storage, a database backup, a Software-as-a-Service (Xembly), a firewall (SunScreen), a cloud computing offer (Project Kenai), and finally MySQL, an open source database.
With a highly-respected, high-end hardware server solution, combined with a strong and stable UNIX-based operating system and coupled with a very large software offering, Sun was a probable target for Oracle, the world leader in enterprise software, complementing perfectly with a unique product portfolio ranging from hardware to the application layer. Oracle could finally be in the same league as IBM and HP.

Oracle will be the only company that can engineer an integrated system - applications to disk - where all the pieces fit and work together so customers do not have to do it themselves.” – Oracle CEO Larry Ellison (Apr 20, 2009)

In this acquisition, the Java Platform could have been a major issue. With more than seven billion electronic devices using Java around the world, including computers and mobile phones, the recently declassified open source (May 8, 2007) is still under constant pressure from developers' communities and other IT experts (SAP CTO Vishal Sikka). But since IBM and Oracle are heavily involved in the development of the Java EE (a platform for server programming in the Java programming language) the European Commission has considered that the two software giants have a common interest in keeping a free Java alive.

The concern lies in MySQL, and Oracle’s intention of keeping this open source software active and accessible to the large majority of companies around the world that are already using the software.

The Database Market

MySQL is a Relational DataBase Management Software (RDBMS) used in free software projects (such as WordPress or phpBB) and is one of the four key components of the overly popular LAMP stack (Linux, Apache, MySQL, PHP or Perl). It is also used in very high-scale internet companies such as Flickr, Facebook, Wikipedia, Baidu or YouTube.

MySQL was previously created, owned and sponsored by a single for-profit firm, the Swedish company MySQL AB, before being acquired by Sun for U$1bn (Jan 16, 2008).

MySQL AB was considered at the time to be the leader of the “second generation” of open source companies, in implementing what is called today the dual-licensing distribution model; [1] the software was given as a free copy under the GNU General Public License (GPL), but [2] also could be purchased under a more traditional license when the intended use was incompatible with the GPL license (e.g., reselling a software using MySQL).

Database Market by Revenue, Source: IDC 2007

The total revenue of MySQL in 2007, which barely passed the U$40m bar (0.2% of the market), is not to be compared with the U$44.1bn Oracle made, just on the RDBMS segment only (44% of the market). But since MySQL is open source and is therefore free, it would be misleading to compare a very expensive RDBMS like Oracle, or DB2 (from IBM) with a much less expensive software such as SQL Server (Microsoft), and more so with a free product like MySQL. A more appropriate reference would be the number of instances installed in the various servers around the world to see usage. There again, due to its free copy to download and install, it is quite hard to estimate the real number of instances around the planet since nobody reports to the company when an open source software is installed, but JoinVision has managed to collect some numbers about the database market.

Market Share, Installation Number
Source: JoinVision E-Services GmbH, July 2006

After the likely merger, a combined Oracle-MySQL will top at 52% of the market, leaving Microsoft behind at 24% and IBM at 10%, with the remaining 14% being split widely among the other open source software and small dedicated players, but most significantively, the RDMS market will change drastically.

Can Oracle Kill MySQL?

If the merger occurs, Oracle would have total control of MySQL’s source code and intellectual property, and could change the licensing agreement (e.g., fewer distribution deals, much higher price), restricting therefore the ability for other companies to develop their own products. Another possibility is to downsize the product by “re-allocating” internal resource investment (through natural attrition attributable to culture clash for example, or lack of prioritization of talent acquisition), dismantling the intangible knowledge cultivated over years and throwing the MySQL brand into a downward spiral. This plausible scenario has happened already with InnoDB, an open source storage engine for MySQL, acquired by Oracle (Oct 07, 2005), and merged into the Oracle Open Source initiative. Since that acquisition, Oracle (intentionally or unintentionally) only moderately supported the functionality and performance development of the product; and a competitor project was kicked off in 2008 by Percona out of the open source community's frustration. In the same vein, Oracle has been offering for years a product called Oracle Unbreakable Linux, a clone and competitor of Red Hat Enterprise Linux, fragmenting even more the open source operating system market. Rightly so, the open source community doesn't have a large amount of trust in Oracle, and its first failed attempt to acquire just MySQL from MySQL AB in 2006 is certainly fueling the current suspicions from the industry.

Could a Successful Open Source Database Emerge?

Not for years.
Among the many open source database contestants, the second and most likely competitor PostgreSQL Database Management System is licensed under the BSD (Berkeley Software Distribution) scheme, which inherently prevents the viral effect of free software from happening, unlike MySQL's GNU license. This non-copyleft feature had stunted software propagation velocity, developer community growth and partners ecosystem density, putting the PostgreSQL project years behind in terms of market penetration. MySQL is an open source product, whereas PostgreSQL is an open source project, with many limitations for business usage. Users will not have a viable, alternative open source database for years if Oracle decides to put the brakes on MySQL development.

Can a Successful MySQL ‘Fork’ Emerge?

Not for years.
But many have tried. OurDelta (started by Arjen Lentz, an ex-MySQL developer), Drizzle (a fork approved officially by Sun!), and MariaDB (started by Monty Widenius, a MySQL co-founder) have not yet attained the level of trust of the original MySQL product. The MySQL brand is strong and will be difficult to match, even with the same code and same development and management team. All these fork developers-cum-vendors face the same challenges: building a new network of partners and a new client base, replicating the MySQL documentation solely owned by Sun, overcoming the question of longevity in a challenging business climate, or living under the potential shadow of an expensive legal battle over trademark patent. MySQL has managed to create a solid brand, backed by an enthusiastic developer community, and with strong ecosystem ties. This is the only plausible reason why Sun, and Oracle, wanted to buy MySQL in the first place, rather then 'fork'ing themselves.

Hefty open source software alternatives are important for the economy because it helps to [1] define global, common technological standards due to its participatory method and inclusive group attitude, [2] keep prices low and thereby lower the barrier for innovation beyond the database space (e.g., development of blogs, open source CMS's), and finally [3] pressure commercial competitors to innovate in their current market. I want to believe that that is the true motivation behind European Commissioner’s objection to the merger, although rumors suspect Oracle's competitor pressure (SAP, IBM) and personal political agenda.


MySQL’s future depends uniquely on Commissioner Kroes’ will to force Oracle to spin-off just the MySQL product from the deal, since there is no doubt that any other outcome will see an eventual death of the product. But the fundamental question is much broader than the current acquisition approval intricacies; the question is:

What should be a viable open source organization in a corporate world?

Since its inception in the 80’s, the Open Source movement has gained significant traction and tremendous business pressure on large corporations with competitive products. Time is long gone when puzzled looks were given to software developers when they mentioned their unusual pastime.

"Who can afford to do professional work for nothing? What hobbyist can put three man-years into programming, finding all bugs, documenting his product, and distribute for free?" Bill Gates - General Partner, Microsoft (1976)

Over the past 25 years, many large open source softwares have emerged as viable alternatives (Linux, Apache, Firefox, to name a few) to commercial products and moved their way up in the disruptive technology scale, by proposing lower functionality at cheaper Total Cost of Ownership (TCO). And as time passed and as development gained more support from growing communities, a previously low-functionality, disruptive open source software could soon lead a market with a rich, innovative product on par with a competitive product, whether open source or commercial.

Source: Christensen and Joseph L. Bower’s HBR article
Disruptive Technologies: Catching the Wave
(January–February 1995).

The fundamental difference that exists between Free and Open Source Software (FOSS) movement and corporations is essentially in their ownership of the organization, not their purpose or outcome. A 'corporation' is owned by shareholders, who demand a growth on their financial investment. On the other hand, a FOSS is owned by a community, who demand growth of the community itself. Each time an organization crosses the line (as MySQL AB did), its mission statement becomes unclear and inevitably, conflict of interest arises.

MySQL AB’s shareholders and board members, in accepting to be sold to Sun in 2008, crossed the invisible line to join the corporate world, with different rules of engagement and bloody consequences, one of them being a victim of a hostile takeover by competitor Oracle (has MySQL AB thought of a poison pill?). Some would argue that they had already crossed this line long before, when they secured their series B in 2003.

The future will continue to have a wide range of businesses and volunteers that use and need to improve open source software. FOSS has proposed an original and unique solution in decentralizing the resources (both time & expertise) to communities but is still struggling on the cost side to achieve a viable equilibrium.

One solution to become more economically sustainable for open source entities may be to become a completely open and democratic alliance, including the community and all the stakeholders. As it is for many FOSS today, a transparent, independent, small, dedicated team could lead the main development and drive that progress. This team could be a for-profit organization but not necessarily, as long as the software copyright and distribution ownership resides in the hands of the entire alliance. On the basis of being a completely open, participatory and democratic entity for the good of open source, every major decision (e.g., CEO appointment, technical standards) should be done through a majority vote in the alliance, which includes developers, users, partners, distributors, regulators and even commercial competitors.

To fulfill the financial needs, the alliance could generate revenue by [1] the dual licensing as well as premium subscription scheme to partners and clients, [2] offering consulting services on the open source product itself, and [3] soliciting sponsorships from those within the alliance, potentially large corporations. Sponsors should not have any incentive in the alliance except to benefit from a working product (as was done for Linux by Novell, Firefox by Google, or Apache by IBM, to name a few). However, some corporations can use sponsorship of open source as a defensive strategy towards its competitor in another product market (e.g., SAP sponsored MySQL AB previously for it to compete against Oracle).

This dialectic between a corporate model type of organization and a cooperative model type of organization has recently added an extra victim in the name of MySQL, but is a central argument in our more technology-driven, more complex and more globalized world. The Free and Open Source Software movement is a great social and technological experiment; and this latest rhetoric with Oracle and Sun (and MySQL and the European Commission) will impact the advancement of not only FOSS but also the society at large.

<< Home