RE: Numerics PACKED, ODD format -- Always recommended?

["Walden Leverich" <walden@xxxxxxxxxxxxxxx>]

Booth,

In packed data the 9th character is free from a DASD point of view. That is
one of the reasons for the argument of going to the odd lengths. Think of it
this way:

8,0 field with value 12345678 is stored as x'012345678F' on the database (5
bytes)
9,0 field with value 123456789 is stored as x'123456789F' on the database
(still 5 bytes)

The argument is: since you need the last nibble to store the sign 'F' and
since all data must be byte aligned (at a minimum) you are simple wasting
the high order nibble going with a even length field.

-Walden

-----Original Message-----
From: mcsnet!midrange.com!midrange-l-owner@Mcs.Net
[mailto:mcsnet!midrange.com!midrange-l-owner@Mcs.Net]On Behalf Of Booth
Martin
Sent: Thursday, March 12, 1998 12:36 AM
To: MIDRANGE-L@midrange.com
Subject: Re: Numerics PACKED, ODD format -- Always recommended?


Bob Crothers wrote:
>
> Karen,
>
> Yes, it is best (from a performance point of view) to use odd
> length packed numeric data in your files.


Doesn't db/400 ignore the packed vs. zoned definitions now?  I thought
all numeric fields were packed regardless of what one writes in the dds.
About the odd vs even issue: if you need 8 characters, then you need 8
at least so you can't go to 7.  My understanding always was that the
cycles for 8 are the same price as the cycles for 9, so the 9th
character is a freebie, excepting dasd.
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---
uucp

+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---