I know some may be frustrated by even reading the topic's titlte, but WAIT!

I just want to ask something simple (I think...)

I've browsed through all (at least most of them) the posts and topics regarding the subject. I've also extensively reverse-engineered many examples in order to understand them.

Since I am an enthusiast (avatar confirms) I can't afford investing to commercial solutions (BH-A, CNS, DataFileGuard etc). So trying to see if I can make something as close as possible to the commercial ones, I was wondering:

Can an audit Logging system (with changes tracked to a separate log table - not like NightWing's Ultralog) be implemented in a way other than script-triggered when changes occur?

From what I understand, every self-respected developer tries to avoid using triggers in the back-end of the app, so is there any other way for this?

And if YES, can an amateur like me hope that one day (soon enough) be able to make it? I would really want to make it myself, just for the sake of it...

Thank you in advance!

Answer (if you don't want a plug-in; don't want a single field log; don't want script triggers):

Hourly backups which are then zipped and set aside for a year. If you split into data and UI files then you can reduce your backup requirements by backing up data hourly but UI only after schema change.

Edited July 27, 201015 yr by Guest

Answer (if you don't want a plug-in; don't want a single field log; don't want script triggers):

Hourly backups which are then zipped and set aside for a year. If you split into data and UI files then you can reduce your backup requirements by backing up data hourly but UI only after schema change.

OK, straight answer received

Are the known plugins operated on the script-triggered level or not?

I'm wondering why you can't do all edits in a popup (therefore, you're in a script) and write changes to a Log table?

There are lots of ways to do this. For instance you could use the Cologon log but on record commit run a script that writes the recent results to a different table.

Returning to one of my old posts, can someone clarify something on the use of Ray Cologon's example of Ultralog?

According to Ray: "All edits are logged in the table where they reside (so, for instance, portal edits must be logged in the portal table)."

So, if I am using the Anchor-Buoy method in my schema, where do I have to make the needed connections (for the separate LogArchive and LogPivot tables) in order for his example to work?

If one table is to be used in many instances (for different anchors), how do I log it's data into a separate table? Do I need an instance of the LogArchive and LogPivot table connected to every instance of the tables I want to log, or just have simple anchors of those tables with their respective LogArchive and LogPivot connections?

UPDATE: I am talking about the functionality of logging to a separate table, via the relevant script