Re: Fixing bad data in a numeric field

[Barbara Morris <bmorris@xxxxxxxxxx>]

"Smith, Nelson" wrote:
>
> How about EXTBININT(*YES)?  That's in our standard H-Specs now.  Could the
> same things be said about it?
>

Nelson, EXTBININT(*YES) just says to treat externally-described binary
fields as though they were integers.  So instead of a database field
being internally 9b0, it would be 10i0.

The only time this wouldn't be a good idea is if part of the application
was treating the data as 9i and part of it treating it as 10i.  And then
there'd be a problem if the 10i side wrote a value bigger than
999,999,999 to the field, since the 9B side would get a truncated
version of the value (or an overflow error).

The options that ARE similar (bad) to FIXNBR(*INPUTPACKED) in allowing
problems to be ignored are:
     FIXNBR(*ZONED)
     TRUNCNBR(*YES)

I would code FIXNBR(*NONE) and TRUNCNBR(*NO) in my standard H /copy
file.