Oracle questions are fine here since, as you've noted, we're supporting both PG and Oracle with OACS 4.
AFAIK no one has measured the performance implications of replacing the bitmap indexes with straight indexes. It would be good to have some solid numbers. AFAIK aD never measured the impact themselves, at
least I've never seen any numbers published by them.
The impact would
have to be very dramatic to justify the extra cost, IMO. It's cheaper to buy a separate server to use as an AOLserver front-end to offload the box with the db on it than it is to upgrade to EE, for instance. It's
a *lot* cheaper. Oracle SE would have more RAM to play in and more cycles available for its use. Or you could buy more RAM or a faster CPU and pay the increased SE license.
It just seems that there are many better ways to spend the extra $$$ than to do so just for bitmapped indexes. Now that SE comes with the INSO filters for InterMedia the argument for EE is pretty toothless for the ACS.
