MySQLobotomy
Oracle acquires InnoDB.
It's easy not to like the limitations of MySQL if you're used to more comprehensive db platforms, but you do have to respect its ubiquity, ease-of-use, and suitability to WORM tasks (write once, read many times).
When it comes time to pick a database for a project, I like it just fine in the right circumstances ( esp. for ISP-hosted applications).
The biggest reasons I don't like to recommend it for non-hosted projects have been:
- MySQL AB's somewhat aggressive interpretation of GPL (if you make $$$, we make $$$)
- that such core functionality as a transactional persistence engine was a bolt on, and not even owned by MySQL AB.
I didn't actually know that Innobase was a separate company, just that its storage engines are pluggable. Other people view that as a feature, but it has always seemed like a good way to introduce subtle bugs and inconsistencies to code that otherwise claims to support 'MySQL'.
But with the most strategic storage engine now owned by a enormous competitor, one who wouldn't feel too bad if, in the interests of supporting open source, InnoDB was released GPL-only and MySQL's product revenue flatlined, it's easy to say I'm even less likely to pick MySQL for a behind-the-firewall application.
Josh Berkus has an interesting writeup of how the Postgresql team dodged this bullet once, and consequently now has a more distributed, highly available, fault tolerant employment strategy. :-)
Aside: The Oracle press release is an good example of surreal business doublespeak -- replace s/Oracle/Microsoft/ and s/innodb/ext3/ and see if you find it as amusing as I did.
8:33:42 AM 
|
