Replication and Schema triggers r-code should be stored in the database. The CRUD operations cannot work without them. Currently if your PROPATH is not setup properly if an operation causes a trigger to fire, the code will not be found. If the r-code was in the database then CRUD operations would work regardless of the PROPATH.
This would help Pro2 implementations.
This would help Pro D/L implementations.
This would help any customer that deploys triggers.
This looks similar to another enhancement "Real database triggers and stored procedures (ABL) ". The business cases and votes should be combined (or one of them closed).
I just read "Real Database Trigger" and this idea is different. Real Database Triggers is the notion of having the Server exceute the triggers instead of the client, which the AppServer was created to solve that issue.
This Idea is to store the trigger code in the database. The client still executes it. It's a matter of where the client finds the r-code. Currently it searches the PROPATH to find the r-code, this means the trigger code needs to be pushed to all machines and PROPATHs that are executing against the database. Putting the r-code for the triggers in the database means that this push doesn't have to happen anymore. The r-code could be part of the SAVE SCHEMA command to help startup. Doing this would make sure that any and all connections are able to find the triggers and execute them.
This an interesting idea. There might be some challenges, though; one is that the same r-code might not be compatible with all clients. I think this idea would benefit from my idea ("Real database triggers and stored procedures") so that trigger code would reside in db and it would be run by the server, not client.