Re: Field compiles as Packed but is zoned in file

We actually got a response from IBM regarding the PMR that we raised
with them. According to them this is not an issue.

quote from the IBM email 

PMR 55962 - Zoned fields get defined as packed in DS subfield s

The compiler does change the internal format of the fields described
in your physical file. This occurs in three stages:

1. All externally described numeric fields are converted to packed
format for internal use by the program.

2. Next, the compiler tries to resolve the data types for fields
described in data structures:
  Per the iSeries WebSphere Development Studio ILE RPG Reference
Version 5, SC09-2508-04, page304:
Position 40 (Internal Data Type)
This entry allows you to specify how a standalone field, parameter, or
data-structure subfield is stored internally. This entry pertains
strictly to the internal representation of the data item being
defined, regardless of how the data item is stored externally.
Entry Explanation
(If Position 40 is) Blank When the LIKE keyword is not specified:
* If the decimal positions entry is not blank, then the item is
defined as zoned numeric if it is a subfield.

3. Finally, with all the "basic" definitions nailed down, the compiler
finishes with the definitions that are "LIKE" other fields - your D01_
subfields in this case.

It takes the resolved data type of each field, not the definition from
the DDS, and applies that wherever *LIKE is encountered, so the D01_
fields have the same data types listed above.

  Remember that this is the program's internal format only - whenever
one of the fields is written out, it will be in the format specified
by the DDS. Straight numeric compares will not be affected, but if you
move one of the fields into a character variable or do a compare of
two data structures, the internal format will then affect the results,
and it sounds like that's what you're running into.

-- 
Krish Thirumalai