m_theologos wrote:

> - ...In optimizing the CONTAINING predicate. Perhaps is necessary to
> have a few words on this:
>
>

I'm afraid you don't understand the "containing" predicate. It is
nothing more than a case insensitive substring search. It is not word
based and can't be indexed.

Full context text search is quite a different beast. It is
straightforward (and quite efficient) to implement as a btree, but
rather than encoding record number, it must carry <table, field, record,
word position> to enable cross table, phrase, and proximity weighting.

Historically, database guys don't "get" search. You're never going to
get any where trying to convince them you have a better way to handle
the index (I don't think you do) until you convince them that there is a
difference between full context word search and "containing". Good luck.