It's no surprise that key MySQL executives are departing the company not long after its acquisition by Sun. That's just the way it goes when companies merge: The old top brass are the first to get the axe. Often it's a given before the ink on the deal is even dry.
Then again, it's also no surprise that the departures of former MySQL CEO Mårten Mickos and former CTO Michael "Monty" Widenius sent shock waves through the open source community.
As chief executives go, Mickos was a canny, articulate, vocal advocate of open source. And Widenius thoroughly seemed like the developers' developer. The world's only MySQL Fellow, he wrote the first version of the open source database and remained with the company until his company can do without him. Should MySQL be any different?
Mickos' situation is clear-cut. He was a capable CEO, but Sun already has a CEO. Furthermore, the MySQL merger wasn't like Novell acquiring Ximian or Suse. MySQL hasn't become some island of open source adrift in a giant proprietary software sea. Sun is already betting the farm on open source. If he stayed, Mickos' advocacy would be just one more voice in the choir.
And in Widenius' case, it sounds like he simply couldn't put up with the culture of a large corporation. "Sun and I concluded in the end that I have much higher chances of achieving my goals outside of Sun," he wrote in a recent blog post. He has since gone on to found his own software venture, which he describes as "a true open source company" -- take that as you will.
Another MySQL co-founder and engineer, David Axmark, left for similar reasons last year. These are signs of the times -- but if these changes lead to a renewed focus on enterprise-quality software development at MySQL, it could only be a good thing.
... and MySQL, you're no Oracle
While MySQL has been a successful and popular product, there's definitely more work to be done. From the outset, MySQL was never intended to be a replacement for full-featured relational databases like Oracle or DB2. It was a lean, quick database engine ideally suited for light jobs -- Web-based content management systems, for example. As such, its users were willing to forgive a few quirks.
As MySQL's installed base broadened, however, the pressure to add more advanced features grew. Customers began demanding such advanced features as transactions, clustering, and replication -- and the more MySQL tried to accommodate them, the more it drew the attention of competitors in the mainstream database business, many of whom had deep pockets.
Widenius was notoriously critical of MySQL 5.1, the latest production release. But the fixes he proposed included creating "a true open development environment that would encourage outside participation," as opposed to the more centrally managed model for which MySQL is known. Is that really in MySQL's best interests now?
I'd argue that Sun needs to tighten the reins on MySQL development, not let them go. Going forward, MySQL will need to work hard if it wants to stay in front of the low-end database business. It's no longer the reigning speed king; if fast, SQL-based storage is all you care about, SQLite is a better alternative. And competing open source databases like PostgreSQL and Firebird offer better feature sets for exactly the same price as MySQL.
Can Sun get MySQL right?
Sun paid $1 billion for MySQL. It wouldn't have done that if it didn't see a strong market for the database. And even given its financial woes, Sun remains one of the few companies with the R&D muscle to put MySQL development on the fast track, eliminating its bugs and SQL quirks and expanding its enterprise feature base. The question is whether it can pull its act together.
There's reason to be skeptical. Already Sun is up to its old habits. According to reports, Mickos' responsibilities will be handled by Karen Tegan Padir, who becomes vice president of Sun's new "MySQL and software infrastructure group." Said group also includes Sun's SOA technologies, identity management, the GlassFish application server, and the Java Enterprise System (not to be confused with Java EE). Say what? MySQL gets a big shout-out, but the rest gets nary a mention? What's next -- changing Sun's stock ticker symbol to MYSQL?
I jest, but only a little. Talk is cheap, and simply name-dropping MySQL achieves nothing. Sun has all the brains and the resources necessary to make MySQL a leading, truly enterprise-class product. What it hasn't yet done is demonstrate that it can execute a winning strategy. Can Sun stop tooting its own kazoo and chasing its own tail long enough to pull off a win with MySQL?
In a way, this moment is as much an opportunity for Sun as it is for MySQL. Let's hope Sun doesn't screw it up.