If one initiates a rollback on another's question, then obviously to undo an invalid edit. Subsequent edits may again be invalid as the following examples show.
Yesterday I did three rollbacks for complete deletes on the following questions. Two of them were again modified to change the question and render answers meaningless.
- Implement my own maplist in Prolog? (1 rollback)
- Prolog - Multiply a list with an element results with a weird tail? (2 rollbacks)
- Prolog - get the factors for a given number doesn't stop? (2 rollbacks)
A notification about an edit after the first rollback would help to find such cases.
Or ... is there another way to handle these?