RE: SQL vs. traditional I/O?

["Joe Pluta" <joepluta@xxxxxxxxxxxxxxxxx>]

> Dan:
> Is this still true?  I thought well-designed SQL will perform as well
as,
> if not better than, well-designed native I/O, for most applications as
> of one of the V5 releases.


> Rob:
> As I said before, a dyed-in-the-wool traditionalist got better
performance
> out of his SQL (he decided that it wasn't fast enough to justify
changing
> his programming paradigm).

This is EXACTLY the kind of question the IAAI is going to address.  And
I think this particular question of SQL vs. native I/O should be one of
the first questions on our agenda.  My last tests showed SQL to be quite
the dog in INSERT and UPDATE, and still a laggard in record-at-a-time
processing.  It's time to reimplement those tests and others.

Anybody who wants to get involved in this issue, please come visit the
IAAI website once it's fully operational.  (And you bleeding edgers are
of course welcome now <grin>.)

Joe