What Oracle's purchase of Sun likely means for MySQL

[chipotle]

Nothin’ much.

• Small companies and open source projects weren’t going to buy Oracle anyway. Killing it doesn’t get you any extra business from them.

• Companies that really want Oracle really want Oracle. The number of places that could afford Oracle but chose MySQL instead are frankly pretty minimal.

• MySQL is a big player—far and away the biggest—in the web database market. Oracle rarely shows up there except at companies already using Oracle. Again: not much overlap.

• MySQL’s biggest “competition” in the open source space is PostgreSQL. PostgreSQL, though, does treat Oracle and other “enterprise databases” as competition, and has a lot of “enterprise-level solutions” that have far more direct analogues to The Oracle Way. In other words, MySQL isn’t directly going after Oracle’s customer base—but PostgreSQL is. Even if Oracle doesn’t particularly care about MySQL, pushing users to PostgreSQL is not in their best interest.

• Last but not least, if you use MySQL, there’s a pretty good chance you use the InnoDB storage engine because it’s the one that sucks the least. Oracle has owned InnoDB since October 2005. Back when they bought it, people said, “Oh, they’re just gonna kill it because they hate MySQL.” I don’t think that’s become any more true in the last three and a half years, though.

This isn’t to say that this isn’t something to keep an eye on—MySQL could find itself in a zombie state like the old FoxPro database after its acquisition by Microsoft, where updates and even major releases kept coming for years but there was always a vague sense of it being treated like a bastard stepchild. But that doesn’t seem likely to me; MySQL is far more visible and is a market leader within certain segments. Oracle stands to gain far more by continuing the model that MySQL AB already had developed and that Sun continued: an open source “community” build and a commercial build.

Comments

[chipotle.livejournal.com]

Monty did leave MySQL just a couple months ago, apparently -- I knew he was working on a new storage engine, "Maria," which is apparently supposed to be a fully ACID-compliant replacement for MyISAM. There's a couple other MySQL companies about as well -- one fork of the database I know of called "Drizzle" (which doesn't seem to really be ready yet) and a couple companies making their own higher-performance distributions. I suspect they'll probably help keep things on track.

I used to be mildly anti-MySQL; I learned SQL back in the mid-90s on Microsoft SQL Server, and at the time current versions of MySQL didn't support things that seemed to be really basic to me like subselects (I'm not sure it even supported inner joins then) -- and that was also before InnoDB, so it wasn't transaction safe and ignored foreign key constraints. MySQL 4.x fixed most of the things that made me cringe, though, and I've found myself using 5.0 even for personal projects. There are still ways in which PostgreSQL is a better design, I'm given to understand, but it's a lot harder to get up and running. And there are some pretty nice MySQL-only design tools lurking around out there.

I admit since I'm a Mac user, I haven't looked at Winamp in years. iTunes on the Mac tends to be fairly good, but a lot of the software Apple produces for Windows tends to be... uh... less good. :)