Re: is this the correct understanding for commitment control running in ILE?

On 2024-07-25 9:26 p.m., Javier Sanchez wrote:

I got a handout of one of her sessions when I attended the RPG Summit in
the fall of 2015. There she summarized what the 7 deadly sins are and also
has a link to the document but it's not reachable.

Summarizing, though, here's the 7 sins:

1. Run ILE programs in the Default ILE Activation Group (DAG)
Use ILE named Activation Groups instead (such as QILE)

> ...

I hate to disagree with Susan, but I think using QILE should be listed as another deadly sin.

QILE is the activation group that you get with DFTACTGRP(*NO) when no specific activation group is specified. (Or it could be QILETS if you have STGMDL(*TERASPACE), but my argument also applies to QILETS.)

If you choose QILE, you are using the same activation group as all those programs written by people who didn't bother to consider what activation group to use. Those programs might be adding overrides or commitment control that you don't want to be part of. And if the programmer of the QILE program didn't consider what activation group to use, they might also be a bit sloppy about exception handling, with the increased risk of a program ending with an unhandled exception, which could cause the activation group to end before you want it to.

I can't imagine any scenario where it's a good idea to choose QILE as the activation group for any production program. So unlike *NEW, where it's only a sin if it's unplanned, I think QILE is always bad.