Re: OVRPRTF question

[CRPence <CRPbottle@xxxxxxxxx>]

On 2/2/11 7:58 PM, jmmckee wrote:
> Is there a difference between a WRKQRY with an 80 character output
> line and an RPG III program with an 80 character line?  The first
> one, sent to a Xerox remote outq, prints portrait.  The second prints
> landscape.

  WRKQRY uses the printer file *LIBL/QPQUPRFIL.  The RPG program would 
be [nearly] unable to use the same printer file.  Review beyond just the 
PAGESIZE() for the FONT(), CPI() and PAGRTT(), as suggested by Gary.

  Note also that the Query/400 has a "feature" which, in response to a 
PAGESIZE [which RUNQRY incorrectly kept as pre-GA naming: FORMSIZE()] 
with "wide" output, the Query/400 run-time will open the printer file 
with a specific FONT() setting; i.e. irrespective of what is defined in 
the PRTF for FONT(), the open is set IIRC as though FONT(*CPI) had been 
defined.  The PRTF QPQUPRFIL having the default page rotation of 
PAGRTT(*AUTO) combined with a default CPI(10) can not "fit" across the 
page when printing in portrait, so the landscape mode will be 
"automatically" chosen to effect "best fit" according to that CPI.  An 
override to include FONT() and\or PAGESIZE(), or a smaller CPI(), can 
control that nuanced behavior.

> This is an old RPG III program, used as a print service program. It
> reads from a file and prints data to selected printers. Hard to test
> the thing without impacting users.

  Appropriate use of CHGSYSLIBL, CRTDUPOBJ, CHGPRTF, OVRPRTF, OVRDBF, 
and even QRYFILE() and OUTFILE() specifications on the RUNQRY should all 
help to enable testing without impacting the users.?

> A second question, related to the above:  Original compile time
> table holding the skeleton OVRPRTF command only includes the first
> parameter on PAGESIZE().

  Presumably meant: The first element of the PAGESIZE parameter, such 
as PAGESIZE(66) or the same as PAGESIZE(66 *N *N) having been specified?

> I put a second OVRPRTF after that with PAGESIZE(66 80). No change.

  The first override with the same scope takes precedence, unless 
secured, or the override to FILE() was replaced entirely at the same 
call level.  AFaIK the precedence is at a parameter level, not to the 
level of the elements of any one parameter; i.e. the first of the 
following two invocations having been performed at a lower caller level 
than the second, will show in DSPOVR after both invocations, as having 
PAGESIZE(44):

    ovrprtf x pagesize(44) ovrscope(*calllvl)
    ovrprtf x pagesize(66 80) ovrscope(*calllvl)

> Moved it in front of the original.  Still no change.
> Output is always landscape. The skeleton does not have the SECURE
> parameter. Which is why I am wondering if there is a difference
> between WRKQRY and RPG III.

  Seems probable, for different printer files, each with specific 
parameters [besides just PAGESIZE()] for which rotation is occurring for 
the spooled output created by the RPG program.

> Because of the nature of this old program, I was reluctant to change
> the skeleton.  But, I'm not clear on why the original override works,
> whether it is first or second, and the added override does not appear
> to work.
>
> Obviously, I am missing something.

  Unless I missed something, I think this is answered by the 
"precedence" described earlier; i.e. the first override at the same 
scope to the same parameter takes precedence.

Regards, Chuck