In reviewing RDBMS and related licensing I found something I hadn't expected: I knew that Oracle's strategy is to shift the revenue base to applications and so ultimately position itself to give the infrastructure products away - but I hadn't realized that Sybase and IBM have pretty much resigned themselves to selling only to their customer bases.
And yet, it appears that both are transfixed in the headlights of impending change: desperately focused on continuing to do what they've been doing:
- so they make noises about competing with Oracle, but new money seems to come mostly from services to existing customers, not new licensing;
- and they talk a great game about competing with SQL-Server in the Windows market (mostly by trying to buy market share by acquiring niche products), but that too is a business on a collision course with extinction as more and more of the dollars come from fewer and fewer loyalists.
In the short term the adaptive failures at Sybase and IBM suggest an opportunity for Oracle: just licensing all the infrastructure tools at small, fixed, percentages of the hardware cost as an interim step on the way to open sourcing the products would let them scoop up a lot of new business - perhaps even enough to avoid or significantly postpone having to go all the way down to free.
In the longer term, however, this stagnation is bad news for customers of all three companies, not terribly revelant to Microsoft, and good news for open source - because the defectors have to go somewhere, and for most of them that somewhere isn't Microsoft.