Since RHT's acquisition of JBoss last June, ORCL keeps reacting.
As soon as the JBoss acquisition was announced, last April, Larry vented his frustration and unhappiness in the press and made public all of the bad he thinks of open source business models. Then, 6 months after, he announced ORCL was now a Linux vendor: when Larry has twice the same opinion, the only thing you know is that he changed his mind an even number of times.
Truth is that Larry knows very well that Open Source is real and is here to stay. Open Source is one of these radical paradigm shifts that take place on the market and put in danger existing business models that are forced to adapt to these new rules. What are the options? Essentially, there are three.
The first option is to try hard to ignore the shift. That is what most software vendors have been doing until recently, but Larry is too wise and paranoiac to hide behind his finger.
The second option was for ORCL to take the lead of that shift and become the new Microsoft of Open Source. The problem is that the 14B USD gorilla hasn't been agile enough to align his ducks and is now left with a situation that shows how he has been unable to properly execute on an Open Source strategy.
Then remains the third option: If ORCL is not going to lead that shift, they'd better try to slow it down. In case you were still wondering, make no mistake, ORCL is not serious about building a Linux business, they are just trying to impact RHT's growth and margin, hence its ability to further invest and grow the stack in Open Source. This is not about giving value back to the customers, this is about keeping value in-house a bit longer. And in order to prove they are getting some traction, ORCL is willing to slightly stretch the reality.
I bet ORCL is going to announce other b-plans in the near future. As an example, the recent acquisition of Tangosol by ORCL was one of those IMO. ORCL knew they had to find a replacement for JGroups in their Oracle Application Server: in the last months, it must have been quite an embarrassment for ORCL to rely on JBoss to power their Oracle AS clustering and high-availability features.
So Larry, what is going to be your next b-plan?
Update (20070402): Andy Oliver mentioned that it might be useful to refer to a previous related post. On that scale, ORCL would sit between the fourth and fifth categories (i.e. between "anti-strategist" and "head-less chickens")