Re: %lookup vs. chain

["Dennis Lovelady" <dlovelady@xxxxxxxx>]

I would argue that the CHAIN would operate more efficiently in general
(provided it's happening more frequently than pages are swapped out).  But
in particular, in this application, it's hard to know with the information
given.

If your array has a large (a relative term, I know) dimension, then the
amount of time spinning through the array to find the same entry over and
over again would seem like wasted effort.  (If you need this particular
entry, why is it in an array in the first place????)

If the record is in memory when the CHAIN operation takes place, it will
likely be more efficient to point to it than to scan an array.

So, as with any other performance question, I can give you an exact answer:
"it depends."

HTH
Dennis





MWalter@hanoverwire.com@midrange.com on 11/22/2002 09:21:53 AM

Please respond to midrange-l@midrange.com

Sent by:    midrange-l-admin@midrange.com


To:    midrange-l@midrange.com
cc:
Subject:    Re: %lookup vs. chain



Since no disk access would be required, and the %lookup BIF, I believe,
does a binary search, that the %lookup method would be much faster.