IMO the most sensible thing to do would be to stagnate but not kill mysql while at the same time using code from mysql to build a mysql compatible interface to oracle and a tool for migrating data from mysql to oracle.
They are already doing this. It's called Common Building Blocks. Seems to me like it's mostly used by the smaller players in the market however...
1.79 x 10^12 furlongs per fortnight -- it's not just a good idea, it's the law!