🚀 go-pugleaf

Mike Ramundo <SYSMRR@ALBNYVM1.BITNET> Writes:
> ... When we bring down
>vmsecure and run test exec ( which builds and issues a call to a
>MODULE ) everything works per specs.  When the test exec is built
>into the exit, and accessed by vmsecure - constant return code = 4
>We have tested under both versions of vmsecure and using both stacks
>and files to pass/retrieve parameters/data.
>
>Does any of this ring any vague bells - or could it be forest-and-trees
>and I'm missing something obvious, possibly added restrictions/tricks
>needed when exit calls a module ?

We don't run VMSECURE (DIRMAINT is much more versatile  :-)), but since
it fails (If that is what "rc=4" means) EACH time you run it INSIDE of
VBMSECURE then you may want to check that when the test exec is being
used INSIDE VMSECURE that the passwords are NOT converted to UPPERCASE
(or visa-versa).

Or, since it fails all of the time, check that inside VMSECURE that the
parameter sequence is not being altered in some way (eg. "userid, passwd,
newpswd" becomes "Password_Routine_Name, Userid, passwd, newpswd").

Or maybe inside VMSECURE, if the console stack is being used, there is a
"Push/Pull ^= Queue/Pull" type of conflict and so the routine is simply
telling you that "the stacked line that is NOT the one with the password
in it" is the one who's positional argument (that should be the password)
is not a valid password (ie. 2 lines are being stacked in reverse order).

We have DIRMAINT here using a similar exit.  If you do not parse the
parms passed correctly, etc. you (like we did a few years ago) will have
troubles getting the exit to work.  The key is that *ALL* passwords are
considered invalid.  Hmmm, do invalid passwords also get rejected?  Maybe
you simply have a decision reversed.

Ron,,,