full_refresh on_schema_change

[ma1ster]

Describe the feature

Right now, none of the on_schema_change options allows to backfill the values in newly added column. The only solution is to run the model in full_refresh mode. In my case models are running in CI/CD pipeline and in case of change in schema I have to provide flag to CI/CD for this one specific model (only technical user that dbt uses have write privileges). It would be so much more convenient if it would be automated providing option on_schema_change: full_refresh.
I know this discussion popped up in the past and back then developers said that incremental materialization needs to be cleaned up to avoid duplicates. Since cleaning has been done in 1.2.0, would that be now possible ?

Who will benefit ?

• those who run dbt in automated environment
• those who doesn't have write privilege ( in this case technical dbt user can only write )