Variable corruption

[bvining@xxxxxxxxxxxx]

Loyd,

Just about every case of variable corruption I have seen has been
related to allocating too little storage for a given parameter.  An
example would be for instance allocating only space for CYYMMDDHHMMSS
with output format *YMD when the system would really store an additional
3 bytes (MMM) for milliseconds.  Could you include the actual D specs
for your various variables as used in your program and also the actual
parameter values being passed to QWCCVTDT?  The physical sequence of the
D specs could be important here.

Possibly related, the binary 4 value -252513633, when viewed as character
data, is x'F0F2F29F' so if storage is being overwritten the value being
used is '022' (if that number has any significance to you).

Bruce Vining

>
>Has anyone experienced variable corruption while using APIs in an RPG/IV
>program (non-ILE)?
>
>I wrote a progam to process save files (retrieve info and display).
>Everything worked fine until I started using the QWCCVTDT (Convert date and
>time format) API.
>
>Before using the QWCCVTDT API, nextpos would increment by 172 (the length of
>the OBJL0300 return list). After inserting the API, nextpos would become
>corrupted after the QWCCVTDT API was called.
>
>For example, nextpos would increment from 1196 to 1368 (not real numbers).
>After calling the QWCCVTDT, nextpos was -252513633.
>
>nextpos is used with the QUSRTVUS API, but not with the QWCCVTDT API.
>Has anyone had experience with the variable corruption I'm experiencing? It
>was easy to work around, but this does concern me.
>
>Thanks!
>
>Loyd Goodbar
>

+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---