At 08:40 AM 11/10/2000 -0700, David Berg wrote:

>I think this is the right place. I'm also not sure if IB6 has a "TOP n",
>but I know that we could really use it if it's not there. We're in the
>process of rewriting a bunch of code to break handling up into groups of
>records, and a "TOP n" records option would probably increase performance.

As many of you know, I was at an open source database conference
recently and met senior people at MySQL and ProgreSQL. We agreed
that in general, when extending SQL we would share syntax. MySQL
has a LIMIT clause on select. This is the syntax:

SELECT [DISTINCT] <select list>
FROM ...
[WHERE ...]
[GROUP ...]
[HAVING ...]
[ORDER BY ...]
[LIMIT <n>[, <m>]] <<<---------
[INTO...]

LIMIT <n> indicates that at most <n> records will be returned.
LIMIT <n>, <m> indicates that the rows returned will start at
the <n>th row and include up to <m> rows. The limit is applied
after the projection (distinct), all restrictions, and ordering.
The first row is row 0, so LIMIT 1, 10 will skip the first row
and return the next ten.

My inclination is to do a partial implementation (LIMIT <n>)
quickly and extend it later.

Regards,

Ann