Re: Named activation groups vs *CALLER (was: H ActGrp('QILE') vs H ActGrp(*caller))

Hi Gary,

Indeed I did read your article (quite recently; the magazine goes from desk
to desk, which can be a pain, especially in the holiday season).  Somewhere
in the back of my head was this nagging question: "Now, did I promise to
react or only to read?  I'll have to look it up".  I just looked it up and
indeed, I did promise to react.  Oops.  OK, I'll admit it: I'm a lazy bum:-)

Now for your article:  I have a question for you:  For whom did you write
your article?  The title ('So you think you understand file overrides?')
suggests it is targeted at people who have at least some interest in file
overrides.  If that is the case, it is a fine, detailed, technical
description of how overrides work.  But it has something in common with most
of IBM's documentation: it explains in great detail HOW things work, but not
WHY things work that way and not which option to take in which situation.
Also, you mention in your introduction that 'an ill-advised job-level
override coupled with modifications may introduce application problems in
the future', but you don't explain what that problems might be.

In my experience (well, maybe I should say 'in my company') most programmers
are more interested in getting the job done than knowing all the technical
details.  Those programmers would probably be better served by an article
that uses a number of practical situations where overrides are needed as a
starting point, and then explains how and why to apply which override.  (Of
course the 'Further reading' section would name your article :-).

As for your examples: they were so hilariously complicated that they would
scare anyone away from using any overrides at all.  In fact I did not even
follow your instructions to try and determine the effect of the combined
overrides.  I said that I'm a lazy bum, didn't I?

I found your analogy with the car engine quite a strong case for using job
level overrides, by the way: Offer an owner of a car with a fouled spark
plug to replace the engine without charge and he'll probably happily agree.
If using a job level override always gets the job done, programmers only
need to remember to delete the override after they are done.  Rest assured
though:  I don't use job level overrides (and I get bitten once in a while).

Joep Beckeringh


----- Original Message -----
From: "Gary Guthrie" <garyguthrie@home.com>
To: <rpg400-l@midrange.com>
Sent: Wednesday, September 26, 2001 12:40 AM
Subject: Re: Named activation groups vs *CALLER (was: H ActGrp('QILE') vs H
ActGrp(*caller))


> Close, Joep. In addition to the call-level or activation group ending
> and explicit deletion of the override, an override that is in effect can
> be replaced.
>
> For a definitive look at overrides, take a peek at "So You Think You
> Understand File Overrides" on page 55 of the July 2001 issue of
> NEWS/400. Or if you're a professional member of the AS/400 network, the
> following link will take you to the online version of the article.
>
>
http://www.as400network.com/resources/artarchive/index.cfm?fuseaction=viewar
ticle&CO_ContentID=10447
>
> In this article, I lay out the details of overrides and how they
> function, including information on call level, activation group level,
> and job level overrides as well as the ILE and OPM environments.
>
> By the way, Joep, did you ever read that article? I was looking forward
> to your feedback as I recall.
>
> Gary Guthrie
> Senior Technical Editor, NEWS/400