Re: Command Parameter Question

[Douglas Handy <dhandy1@xxxxxxxxxxxxx>]

Mike/Nelson,

>a product manager for a product which front ends every function with a command

I do the same thing; I think commands are typically under-utilized on the 400.

>It would be possible to translate a parameter prior to the
>CPP, if you did the processing in a prompt override program,

You wouldn't want to use a POP for this, as a POP is *only* called when the
command is prompted.  If the command is run without prompting, whether from the
command line or as part of a CLP, etc, the POP is *not* invoked.  Thus while a
POP can be convienent for substituting real values for special values like
*SAME, the CPP cannot rely on having the POP run prior to the CPP.

>or in a
>validity checker, though with the latter, IBM do state that it is not
>guaranteed that any values changed in a CVC will be passed to the CPP
>correctly.

Historically, IBM never guaranteed they would, but many people wrote VCP's which
altered parameter values and expected the CPP to see the updated values.  In the
past the process worked because the arguments were passed to the VCP by
reference.

However, this creates a potential security exposure so at QSECURITY level 50 a
VCP receives only a copy of the parameter values.  Any changes made in the VCP
are thus not reflected in the value received by the CPP.

Even below level 50 it is not guaranteed to work.

>I guess we can also add to the list the new command exit
>programs available.

They are now the preferred way of changing parameter values, if you can't handle
it in the CPP itself.

>I would say though that doing it in the CPP would be
>the most straightforward way.

I agree, whether the CPP is written in CL or RPG.

Doug