Another advantage of service programs?

[David FOXWELL <David.FOXWELL@xxxxxxxxx>]

Hi,

The thread Running with LVLCHK *NO in production reminded me of a regular difficulty we have :

For development, we have an LPAR which is a copy of our production machine. Developers will modify, test and simulate installations on this LPAR. Now when testing in a test library, you may find that someone else has simulated an installation and changed a file.

Now, if I've modified module M1, I'll run it in a test library of the same name. In that test library is a program TestM1 that I shall first have to bind to M1.
However, if an external bound call is made from M1 to  a procedure in M2 which is compiled on file F1, that another programmer has just modified, then TestM1 will crash if there's a change in level. This is where I'v sometimes been using CHGPF LVLCHK(*NO). If I'm lucky, it works!

This morning I had an idea : if I run DSPPGMREF on TestM1, I see the record level for file F1. ( Everything is bound by copy). If a service program were made from module M2 using file F1, and this service program were used instead of binding M2 by copy, then I would no longer see reference to F1 in TestM1. ...And there'd be no more LVLCHK issues of this kind.

Can someone confirm this?