In Oracle, when you
want to update a BLOB, you have to do this first:
it and opens it up for writing. Oracle has a concept of a "locator", which I
guess is similar to a pointer or handle to the BLOB object, and it is required
in order to start streaming data to it. Just setting my_blob_column=null doesn't
do the same thing.
Anyway, I'm trying
to port this currently-Oracle app to Derby and of course Derby doesn't like the
reference to empty_blob(). Can someone put me in the picture regarding the way
in which you initialize BLOBs for update in Derby, maybe posting a code sample
or a link to the relevant docs? I've had a quick scan through the docs and
Google without much success...