Jesper Thorhauge wrote:
> Okay, thanks alot for the quick answer! I was wondering, do you know how
> much more efficient it could be done using the binary format you are
> mentioning? I'm especially interested in decreasing the amount of memory
> used to get large bytea values...
Um, rough estimate would be about a factor of 5 (assume text
representation of bytea averages about ~2.5 characters per raw byte, and
characters are 16 bits on the Java side)
-O