Database trigger Issue in V11

Reposting with formatting so I can read it. 

Hi I am currently implementing V11.2 from V102B and experienced the following different behavior. I have the following for each loop with a database trigger on the child table:

for each parent no-lock, 
  each child no-lock where child.id = parent.id: 
  'do something' 
end. 

A database trigger on the find event:

TRIGGER PROCEDURE FOR FIND OF child. 
if child.type = "deleted" then return error.

The trigger code should skip child records that have a type of 'deleted'. This loop works fine in V102B but in V11.2 it does not finish properly. It works for a while then it drops out of the procedure and does not reach the end of the for each loop to continue. If I use 2 for each loops, one for the parent and one for the child, that works correctly.

Has anybody experienced the same behavior and has maybe a workaround. Thanks for any help. - See more at: https://community.progress.com/community_groups/openedge_development/f/19/t/14989.aspx?Redirected=true#sthash.GUGhUMUz.dpuf

My work-around would be to do away with the FIND trigger and handle the exceptions "properly".

Just realised my repost with formatting sounded rude. Not meant that way, sorry. The formatting will have been lost by the community software. A bug I gather is being looked at.

"If I use 2 for each loops, one for the parent and one for the child, that works correctly."

I'd recommend doing that if you're looking for a quick workaround as it should require only minimal code changes.

And then plan on phasing out the FIND triggers.

Thank you for your help,  James

I was hoping for a quick fix in the trigger code but in the long run it is better to deal with it properly. I am most likely to add another line in the for each loop tor the child record to do what the trigger code is doing.

That's a good solution I think. DB Triggers can be really useful, but from experience FIND triggers are a nightmare, not least because it's not transparent in the slightest. You end up with stuff happening and it's impossible to track down why.

I completely agree and I personally never liked the db trigger for that reason. At least when it is dealt with in the code, everybody knows what is happening.