RE: FEOD vs UNLOCK

Hmm, when I use FEOD, my intent is generally to force any BUFFERED data to be forced to disk. In programs where data may have been written (but still in I/O buffer, waiting the buffer to fill). What I'm not sure is whether buffered updates hold their record lock until the buffer is written to disk...

For sure, if a record was locked, but never got UPDATE, then UNLOCK would be used to release...

-Eric DeLong

-----Original Message-----
From: rpg400-l-bounces@xxxxxxxxxxxx [mailto:rpg400-l-bounces@xxxxxxxxxxxx] On Behalf Of Sam_L
Sent: Thursday, December 26, 2013 4:22 PM
To: rpg400-l@xxxxxxxxxxxx
Subject: FEOD vs UNLOCK

It appears that some of the folk who wrote some of our code over the
last 20 year thought that the FEOD opcode would release a record read
for update. I've always used the UNLOCK opcode.

I don't see anything in the documentation that support the idea that
FEOD gets rid of a read-for-update lock. And it appears to have failed
today deep in the bowels of an EDI batch process.

Anyone know where this idea might have come from?