Re: Contiguous memory and the *INKx indicators

>From: bmorris@ca.ibm.com
>Date: Wed, 19 Dec 2001 16:29:25 -0500
>
>Joe, there are no guarantees, but I imagine that any RPG compiler would
>have those indicators contiguous.  Regarding the current ILE RPG compiler,
>the changes you'd have to make could be done a hundred times over before
>we'd be finished making our changes (just thinking about it gives me THE
>SHIVERS).  So, I think you're pretty safe.

Joe, I take it back a bit.  I'd forgotten about *INKO, which doesn't exist
(O as in MNOP).  It's possible that some RPG compiler might have an
internal
array of 25 indicators, so the index would match the letter.  (Currently,
we have *INKN at index 14 and *INKP at index 15, but it might be reasonable
to have *INKP at index 16.)

So, while I can't see the ILE RPG compiler ever changing this (brrrr),
_SOME_
RPG compiler might not work that way.

If I understand you correctly, you generate new RPG source.  If the RPG
compiler ever did change how it handles the indicators, you'd have to
regenerate the source to make the correction, right?  (That is, the
indicator
moving stuff is done in the RPG source, not in your routine.)

Barbara Morris