In the last episode (Oct 06), Richard Jacobsen said:
> I've got some large tables, ~160GB each with indexes, which are only
> used for read only access. Since obviously all table space doesn't
> fit in memory, Will using the compress engine speed reads/seeks on
> these tables up at all, or will they slow them down considerably?
It won't speed up seeks, but it will increase the chances that two
nearby records are on the same disk block, potentially saving you a
seek to fetch the 2nd one. Full table scans may be faster or slower,
depending on your disk speed and CPU.