Re: Trigger program buffer lots of stale/ bad info ??

Booth,

The documentation (<https://www.ibm.com/support/knowledgecenter/nl/ssw_ibm_i_73/dbp/rbaforzahftrc.htm>) says: 'offset 48 type BINARY(4)'. Binary(4) is a 4-byte integer; in RPG that is int(10) in free format or '10i 0' in fixed format. A long, long time ago we didn't have integers in RPG, so you had to use data type B, but that is all behind us now :-)

The '4b 0' in your example is plainly wrong; it means '4 digit field with zero decimals'; I wouldn't even know what the compiler would make of it; probably a 2-byte field, which messes up the buffer completely. '49 52b0' would be correct (but outdated).

In general: I prefer positional notation for these buffers, because the locations have to be correct. In your example any mistake in one field (or commenting something out) messes up the rest of the buffer.

Joep Beckeringh

Op 31-5-2018 om 19:32 schreef Booth Martin:

This has frustrated me.  A program I used successfully in 2011 is failing.  As I looked at various examples I find on the web I found the same lines with different values.  This seems like it may be relevant.

....

 *       d OrOffset             7b 0
 *       d OrOffset             4b 0
 *       d OrOffset            10i 0

.....

I also searched with "Jan 2014 to current" and found almost nothing about RPG triggers - just SQL
Probably I need to go the SQL trigger route?