Re: creating a command with a return value

[Tim Bronski <tim.bronski@xxxxxxxxx>]

That's for the suggestion Chuck, I've added these two options to our 
builds that create commands that retrieve.

On 11/15/2013 9:40 AM, CRPence wrote:
>     Poor REXX, the neglected stepchild ;-)  For completeness, probably
> adding *IREXX and *BREXX too? :-)
>
>     I have always been frustrated by that error when creating commands in
> a somewhat ad hoc manner; e.g. 14=Create within PDM.  Until this moment,
> I had never given much thought to how the command-source compiler could
> be enhanced to be more usable.  Sure would be nice if the default value
> of *ALL for the "Where allowed to run (ALLOW)" parameter of the CRTCMD
> would be treated as a SEV(20) error, with the CPD0281 diagnosing
> instead, that the incompatible environments were dropped.  That is,
> /softening/ the effect, treated almost as if something like a Single
> Value *ALLALWRTN [all environments allowing RTNVAL(*YES)] both existed
> and had been specified, to indicate that *ALL special values for which
> RTNVAL(*YES) is allowed should be included; i.e. IIRC, all but
> *INTERACT, *BATCH, and *EXEC.
>
>     I hardly expect anyone would complain about the former effect if
> changed [considering that is the most likely recovery intent;
> irrespective the recovery implemented].  But I doubt any change would
> ever come, unless implemented more like the latter, as a new single
> value for which CHGCMDDFT would be required, per dev fearing someone
> whining about the change in behavior.
>
> Regards, Chuck
>
> On 14-Nov-2013 15:23 -0800, Vernon Hamberg wrote:
> > You might want to add *BMOD and *IMOD as well.
> >
> > Just in case it matters!!
> >
> > On 11/14/2013 2:46 PM, Gary Thompson wrote:
> > > we just had success creating the cmd with (*BPGM *IPGM) and are
> > > testing now . . .
> > >
> > > Alan Campin Thursday, November 14, 2013 1:37 PM wrote:
> > > > When you did the CRTCMD did you specify that the command can
> > > > only run in a program?
> > > >
> > > > ALLOW(*BPGM *IPGM *IMOD *BMOD)
> > > >
> > > > On Thu, Nov 14, 2013 at 1:32 PM, Gary Thompson wrote:
> > > >
> > > > > We are stumped attempting to create a command which returns a
> > > > > value.
> > > > >
> > > > > when we attempt to create the command, we get compile errors:
> > > > >
> > > > > * CPD0281 3700 30 RTNVAL(*YES) valid only in CL programs
> > > > > or REXX procedures.
> > > > > * CPD0283 3700 10 Default value of null pointer assumed.
> > > > > * CPD029D 3700 20 Parameter with KEYPARM(*YES) follows a
> > > > > parameter with KEYPARM(*NO). KEYPARM(*NO) assumed.
> > > > > <<SNIP>>