I've been fielding a lot of calls about my previous posting Why InnoDB matters so much to MySQL (really) and now I've had time to think a little bit more about it. I maintain the position that MySQL is probably a little shell-shocked, but is not nearly as screwed as Lisa Vaas implies they are. First of all, I find it doubtful that all of the cool new features in 5.0 are reliant on the InnoDB engine. That would be a very poor engineering and business decision as a software company to base the key elements of your product on something you don't own-it's also something that seems entirely out of character for MySQL who have been extremely prudent in their design and business model-even identifying the risks of the dual license business model.
Odds are this is the just first example of this kind of submarine attack that we will see from big vendors who are feeling threatened by open source products that offer similar functionality for significantly less price. It's a big threat to many other open source companies that include other dual-license open source components in their products. Not sure if you've noticed, but there are a lot of them.
So, the question is how much does InnoDB matter to MySQL? From what I can glean InnoDB sounds like a nice piece to have but is by no means make or break for MySQL.
The more interesting aspect of this whole story is that Oracle specifically went after MySQL after repeatedly saying that the company didn't really matter to them. We shouldn't be surprised. Oracle is just being Oracle, rich, aggressive and slightly obnoxious. Clearly, MySQL was a big threat or else Oracle would not have made a pre-emptive strike to MySQL's technology. Nonetheless, it was a strategic misstep on MySQL's part to let a key piece of it's architecture get out of it's grasp.
Posted by Dave Rosenberg on October 12, 2005 07:55 PM
