RE: %EOF RESET

>From: "Mackie, Roger L. (Precision Press)" <RLMackie@ppress-tc.com>
>Date: Thu, 30 Aug 2001 13:23:33 -0500
>
>Chris,
>
>Before V5R1, SETLL did not reset the %EOF BIF. The only way I found to
reset
>it was to reposition and READxx. But if I read the documentation
correctly,
>in V5R1 SETLL does reset %EOF.

Roger, you're almost right.  There is a behaviour change with CHAIN,
SETLL and SETGT in V5R1.  They reset %EOF(file), but not %EOF with no
parameters.

>From: "Reeve Fritchman" <reeve@ltl400.com>
>Date: Thu, 30 Aug 2001 14:53:32 -0400
> ...
>Otherwise, we have to flood Rochester (not Toronto) with requests to make
>this change available via PTF.  I can't figure out why it's changed "now"
>(V5R1) and not then.

Reeve, it's Toronto, and we don't need a flood of requests.  Just one
APAR would do.  Since there's a behaviour change involved, that could
just barely possibly cause a working program to stop working, we can't
just issue a PTF off our own bat.  We need a "real" customer-reported
APAR.

(The scenario that might stop working is bizarre and hard to imagine, but
we've found that if something is possible to code, there's at least one
program out there that uses it.)

Barbara Morris, RPG Compiler Development, IBM Toronto Lab