RE: Keyed PF

I have never *noticed* anything but arrival sequence. This also applies to a 
non-unique LF, and probably also after keyed or unkeyed RGZPFM, but of cause 
not if deleted records are re-used. Also: Why should anybody in Rochester code 
the internals in a randomizing way?

But then again I clearly remember having seen the same information "IBM doesn't 
guarantee anything about the order", so when the order is important I have 
coded it in a way that guaranteed the order. This makes it likely that I would 
not *see* anything else, even if it happens. However I *have* a thousind times 
compared a DSPPFM (arrival sequence) with a list from a print program (reading 
keyed non-unique).

Henrik

> date: Wed, 12 Feb 2003 12:44:42 -0800
> from: Joel Fritz <JFritz@xxxxxxxxxxxxxxxx>
> subject: RE: Keyed PF
> 
> IIRC, IBM doesn't guarantee anything about the order.
> 
> > -----Original Message-----
> > From: TechSupport@xxxxxxxxxxx [mailto:TechSupport@xxxxxxxxxxx]
> > Sent: Wednesday, February 12, 2003 12:10 PM
> > To: midrange-l@xxxxxxxxxxxx
> > Subject: Keyed PF
> > 
> > 
> > Here's a simple question that I can't remember the answer to. 
> >  If I have a
> > PF keyed, non-unique, in what order will duplicate keyed records be
> > accessed?  I'm guessing arrival sequence but could be wrong.  Thanks.
> > 
> > Paul Fenstermacher
> > Bass Pro Shops
> > 417-873-5424
> > paulf@xxxxxxxxxxx