Re: Upgrading R&D LPAR from V6R1 to V7R1 TR7

On 25-Nov-2013 18:49 -0800, Steinmetz, Paul wrote:

Three more V7R1 OS upgrade issues.

1) CPYFRMSTMF resulting in incorrect output, probably due to CCSID
of 65535 on to file. <<SNIP>>

The default for the Database file CCSID (DBFCCSID) parameter is the special value *FILE which suggests that if the to-file is program-described and the Data conversion option (CVTDTA) parameter suggests to perform character conversion, then the /default job CCSID/ will be used to translate from the inferred or specified Stream file code page (STMFCODPAG) or Stream file CCSID (STMFCCSID). If the to-file is a database source physical file with CCSID(*HEX), then the copy should fail [would have in older releases] with CPDA08C "Database file CCSID is 65535." rather than producing untranslated text.

The /default/ job CCSID comes from the associated-CCSID for the Language Identifier (LANGID) of the job, if the job did neither resolved a CCSID from the User Profile (*USRPRF) object nor had a CCSID explicitly set with the Change Job (CHGJOB) command. Be sure that either a non-hex [i.e. not 65535] CCSID is set for the user, or that the user has the desired LANGID established, and that the desired values established in the job are set from those, or are explicitly set within the job to desirable values.

Of course whether any of that is relevant is unknown, because "incorrect output" for use of Copy From Stream File into a file with CCSID(*HEX) is too nebulous.