Re: RTVDSKINF CCSID 37 vs 65535

[CRPence <CRPbottle@xxxxxxxxx>]

rob wrote:
> Some of the fields in the file generated by RTVDSKINF
> are 37 and some are 65535.
>
> <ed for brevity>
> DICDAT CHAR(6) FOR BIT DATA ... DIOBOW CHAR(10) CCSID 37
>
> I hate having to turn on stuff to translate 65535.
>
> Environment:  Run SQL Scripts in iNav.

  That looks like a defect in the definition for those fields 
[DICCEN, DICDAT, DICTIM, DIUCEN, DIUDAT, DIUTIM] of the model 
outfile QAEZDISK in QSYS.  There should be no reason one should have 
to use the "lie about the data" feature to enable seeing the 
character data [from a non-EBCDIC client]; the DIOBTX is a possible 
exception, but its implementation as 65535 versus something else, is 
a side effect of a larger /issue/.  The true "character field" 
definitions of the file, those intended to have visible character 
results presented, should have a non-hex CCSID defined if the data 
is not binary data.  Those character fields have that intention, but 
miss the mark due to the *HEX CCSID.

Regards, Chuck