would it be possible to integrate an undo step for delete row or delete column or other data operations?
It its quite annoying after minutes of import or hours of work to loose data without undo.
Or ist that something only the snap-shot backup via ZIP can do? The problem is, that this would require
constant backups for each step, so a single step undo could solve that. This issue is probably a non-issue
for database developers, or its hard to keep shadow copies but for users it would be quite convinient. Maybe a
size limit of 1GB could solve the automatic snap-shot issue. Not sure how many people would use it, but what are
the undo-buttons for anyway if they are grayed-out most of the time :-) Also the "backup to ZIP" is not visible in the
File Menu, only via right click (Restor is avail, but not Backup in File Menu).
Undo is not really posible for some types of operations like deleting a database column (IJC field). The column is dropped at the database level and is not directly undoable. The only approaches would be to create a complete database backup, or some very complex mechanism to store the deleted data elsewhere so that it can somehow be restored later, both of which could be quite slow.
Where it is possible to provide undo functionality, such as editing a value, we do enable it, but in cases where its not possible the undo operation has to be disabled.
Your best bet if you are using a local database is probalby to make a backup of the database prior to starting on a series of complex operations like this. To do this r-click on the local database schema node int he projects window and choose 'Backup schema to ZIP file...'. This creates a compelte archive of the local database that can be restored at a later date.
I agree, the ZIP backup is probably the safest way to do. Kind of version control or SVN for the database.