We've looked at WordPress, Drupal and Joomla in previous articles. All three are written in PHP and all three use the MySQL database. In the open-source web services world the application stack of choice is the LAMP stack: Linux, Apache, MySQL and PHP (or Perl or Python). Lots of open-source projects and lots of commercial sites such as Flickr, YouTube, Wikipedia and Digg run on MySQL. MySQL is the leading open-source database in the world. It was started by three Swedes in 1995 who also started a company around it: MySQL AB. MySQL was released under a dual-license model: one could freely download the GPL (Gnu Public License) version of it, or buy a commercially-licensed version with a support contract. 

MySQL AB grew over the years to be a truly virtual company with almost 400 employees, $22 million+ annual revenues. Life was good, but Oracle (the 500lb gorilla in the database world) had its sights set on MySQL. Oracle tried to buy MySQL around 2005. The Swedes didn't want to sell. Oracle then tried an end-run. The version of MySQL that they were working on at the time, version 5.0, was a quantum leap in functionality over the 4.0 release. To achieve this functionality MySQL used other open source tools from a Finnish company, Innobase OY and an American one, Sleepycat Software. After MySQL didn't sell out to Oracle, Oracle bought Innobase in October 2005 and Sleepycat in February 2006. Some would say these moves were designed to cut the legs out from under MySQL. After the acquisition, if you went to what was the Sleepycat home page, you'd be redirected to an obscure page in Oracle's Technical Network site. Nonetheless, MySQL continued and they were purchased by Sun Microsystems in January 2008 for $1billion ($200 million in cash, $800 million in stock). Some would say that the acquisition was overpriced, others would say that the road since has been rocky with the release of MySQL 5.1 with some known data corruption bugs. 

So now Oracle has bought Sun and at last has its hands on MySQL. What will they do with it? The could enthusiastically develop and promote it and all would be well. History argues against this. Oracle hasn't been a terribly active proponent of Open Source in the past. Also one could look at IBM's acquisition of Informix in April 2001. They promised to continue to support, sell and maintain Informix. Notice anything missing there? How about 'develop'? IBM's upgrade path for Informix customers was DB2. One would suppose that Oracle could follow the same path and maintain MySQL, while providing an upgrade path to the current Oracle platform. Or Oracle could just kill MySQL the way they effectively killed Innodb and Sleepyct. 

One of the odd things about Open Souce software is that its life isn't so much in the code as in the developer and user communities. As long as they're alive and using the software, the project lives. Today Michael Widenius (Monty), the original developer of MySQL and one of the original Three Swedes let the shoe drop.  He's taking a part of his buyout money from the Sun sale and starting up a new MySQL support and development shop, in hopes of keeping the development and support team that's been with him through the years together. What about non-compete with Sun? My guess is that as soon as the Oracle deal's done, those non-competes go out the window. 

What's my bet? I bet Oracle won't be interested in fostering a competing product to their flagship offering and will  let MySQL langush, converting their commercial customers to Oracle 11i. My guess is that Monty's new company Monty Program Ab will be carrying the MySQL flag two years hence. 

Read Monty's blog post yourself.