It was almost exactly one year ago that Sun Microsystems agreed a "$1 billion deal" to buy MySQL and I have followed with some interest the turbulent aftermath of their delayed release of MySQL 5.1.

Initially slated for April 2007, MySQL 5.1 finally emerged at the back end of 2008, touting many new enterprise features, such as table and index partitioning, row-based replication, event scheduling, and so on. In terms of overall strategy and positioning, the story sounds similar to that guiding the evolution of Microsoft and SQL Server, the aim being to improve performance, scalability, and general manageability, of the database software, in the face of ever increasing numbers and sizes of web (sorry, web 2.0) applications. Sun also seem to have pinched a trick or two from Microsoft in terms of restricting some of the most appetising new features, such as the Query Analyzer, to "Enterprise Edition only".

However, there the similarities end. When SQL Server 2008 was released, it was to general applause. It was a better, more usable tool, with many improvements to existing features and new features that were largely driven by the requests of the ordinary user. And, most importantly of all, it was stable and reliable.

Compare this to the furore surrounding the MySQL release. Amid rumblings of discontent about the new bureaucracy at Sun, and community-requested features not making it through to release, the euphoria bubble of release was almost immediately burst by the original creator of MySQL, Monty Widenius. He advised deploying MySQL 5.1 only with extreme caution and, ideally, only if you didn't plan to use any of the new features yet. He cited 20 known and crashing bugs, 180 serious bugs, and many more bugs that "should have been fixed before release and weren't".

It made me wonder whether more than a few web 2.0 developers, whose natural data platform would have been MySQL, might be peering over the SQL Server Express fence and wondering if the grass doesn't look slightly greener over here.

It is easy to criticise Microsoft but one should also give due credit. Judging by the bloat of SharePoint, the debacle of PerformancePoint, the confusion surrounding LINQ-to-SQL, entity framework, and so on, Microsoft are still sometimes laughably inept at really helping people to use, manage and understand their data, throughout their business. However, in terms of knowing how to nail down the database platform itself, they are leaders, from whom Sun et al. have a lot to learn.

Cheers,
Tony.