Part of our release process is generating a complete database schema. We can then compare the generated schema to the existing production schema. This enables us to catch any accidental “drift” between what the production schema should be and what it actually is.
The generated changes are created by comparing to an empty db. As such we do not need to track unique ids nor ‘authors’. The author tag and id prefix all cause spurious deltas between runs. Filtering out the spurious differences is time consuming and has resulted in important changes being missed.
Currently, a changeSet looks like this:
We would like to force the ‘author’ tag to be a constant value. We would like to eliminate the id prefix (in this example: 'How can we do this?