Unicode, %status and *STATUS

We're incorporating more Unicode fields in our database. But in some cases we still have to maintain a 5250 user interface to these data. So we want to signal to 5250 users when we encounter untranslatable characters in a field. For that we use program status; status is 50 when conversion results in substitution.

But we have a problem when we have two Unicode fields, let's call them FIELDA and FIELDB. When we move FIELDA to an EBCDIC field for the display file and FIELDA contains something untranslatable program status becomes 50 , so the program is aware of the problem in FIELDA. But when FIELDB has no problems when it is moved to an EBCDIC field, program status is still 50.

I would have expected, because an unsuccesful move results in a status change, that a succesful move would reset the status. Like file operations do. But apparently it doesn't.

What I did was reset the *STATUS subfield in the program status data structure. Then, when FIELDB does contain untranslatable characters, status turns to 50 again and I can signal the problem. And when FIELDB has no problems, program status remains zero, so I know there is no problem. Sadly, %status function still returns 50. It seems that resetting the *STATUS subfield has no effect on what %status returns.

Is this a bug? Or a feature:-) ? Or is there some other way of resetting program status (apart from causing and catching an error)?

Joep Beckeringh